Задайте вопрос стоматологу! Полезная информация для пациентов на сайте доктора Бакалова

StomFak.ru

Результаты 145 - 164 из 2382

Dental and oral surgical procedures are some of the most frequently performed minor surgical procedures in the United States. Because these procedures rarely occasion admission to hospital, either for the initial procedure or for the care of a complication, data on incidence rates for procedure-related infections in this setting are scanty. Maxillofacial surgery is more commonly performed in an inpatient setting, especially surgery for reconstructive purposes after trauma or that involving major restructuring of bones for cosmetic surgical reasons. Therefore, more information exists concerning risks of procedure-related infection for maxillofacial procedures. Because of the recognition of human immunodeficiency virus (HIV) transmission in one dentist's practice, national attention has been focused on infection control practices in dentistry (1). This chapter discusses the infections seen in these settings, their recognition, and measures for their prevention. ETIOLOGY Infections after surgery to the gums or teeth or involving mucosal incisions made in the mouth are caused by a combination of the aerobic, facultatively anaerobic, and anaerobic microorganisms found in the saliva and the gingival crevices (2, 3). The number and variety of bacteria found in the oral cavity of each person increases as he or she matures, the dentition erupts, and the flora of the gingival crevice establishes itself. Cross-sectional surveys have suggested that a few anaerobes are present in the mouth of young children before the eruption of their first deciduous teeth (4, 5). Older children have a microbial flora closely approximating that of the mature dentulous adult. This includes the presence of Prevotella and spirochetal microorganisms. Actinomyces naeslundii can be recovered from the mouths of most infants but is replaced by Actinomyces viscosus when the teeth erupt (4). The bacteriology of the saliva is somewhat different from that of the gingival crevice. The gingival crevice has approximately 100 times more microorganisms per gram than saliva does and 70% of these microorganisms are anaerobic, whereas most salivary flora are aerobic and facultatively anaerobic. As an illustration of this difference, Streptococcus salivarius accounts for approximately 47% of facultative microorganisms in saliva but less than 1% of those found in the gingival crevice. There appears to be no relationship between the appearance of the tongue (presence of white coating or not) and salivary bacterial load (6). Table 54.1 lists the microorganisms in the saliva and gingival crevice of adults (7, 8, 9). TABLE 54.1. NORMAL FLORA OF THE ORAL CAVITY Saliva Gingival crevice/plaque Streptococci Streptococci S. salivarius S. sanguis S. faecalis S. mutans S. milleri Peptostreptococcus Staphylococcus aureus Veillonellae Veillonellae Neisseria Neisseria Branhamella Branhamella Actinomyces Candida albicans A. naeslundii Herpes simplex A. odontolyticus A. viscosus A. israellii Entamoeba gingivalis Bacterionema matruchotii Trichomonas tenax Bacteroides species Porphyromonas species Prevotella species Capnocytophaga Eikenella corrodens Fusobacterium nucleatum Actinobacillus actinomycetemcomitans Treponema T. macrodentium T. denticola T. orale A study of the short-term effect of full-mouth tooth extractions (in eight patients with severe periodontitis) on periodontal pathogens colonizing the oral mucous membranes revealed that many Prevotella species still colonized the oral mucous membranes of edentulous patients. However, it was unlikely that these patients would continue to be reservoirs for Actinobacillus actinomycetemcomitans or Prevotella gingivalis (10). Although no studies have documented transmission of oral flora from one patient to another in the dental operatory, Genco and Loos (11) reviewed several studies using molecular epidemiologic techniques to demonstrate the transmission of Streptococcus mutans by vertical transmission from mother to infant (but not from father to infant) and intrafamilial transmission of A. actinomycetemcomitans. These studies are the first to document the spread of oral microorganisms and raise the possibility of whether oral bacteria can be transferred in a medical setting from one patient to another. Many aerobic bacteria transiently colonize or infect the pharynx and posterior nasopharynx. Recognition of these agents depends either on characteristic clinical symptoms, such as the chancre of syphilis or the adherent membrane of diphtheria, or on culture to demonstrate the presence of group A ОІ-hemolytic streptococci or Neisseria meningitidis. The viruses often present in saliva are those agents causing latent infection, particularly the herpes group and less commonly hepatitis B virus (HBV), hepatitis C virus (HCV), or HIV (12). Herpes simplex virus (HSV) has been recovered from the saliva of approximately 1% of asymptomatic children and between 0.75% and 5% of asymptomatic adults. Serial sampling of the saliva from normal adults over time has demonstrated that HSV can be recovered from oral secretions in more than 50% of adults in the absence of clinical lesions. More than 50% of seropositive P.914 patients undergoing organ transplantation shed oral HSV asymptomatically (13). HSV is a recognized occupational hazard for dentists, oral surgeons, and dental technicians. Cytomegalovirus (CMV) has been isolated from salivary glands, adenoid tissue, and pharyngeal secretions. The prevalence of antibody to CMV increases with age and is further increased among people from lower socioeconomic groups. In seroprevalence studies performed in the 1970s, between 40% and 80% of adults had serologic evidence of infection with CMV by the age of 40 (14). CMV excretion is increased in the presence of transplantation and immunosuppression. No transmission to dental workers or medical staff has been shown (15). Epstein-Barr virus (EBV) is another herpesvirus that causes acute infectious mononucleosis followed by a chronic infection of lymphocytes. The prevalence of infection as indicated by the presence of antibody is higher at early ages in the tropics and in underdeveloped countries. Prevalence progressively increases with age in developed countries (16). No transmission to dental workers or medical staff has been demonstrated. Herpesvirus 6 (HHV-6), herpesvirus 7 (HHV-7), and herpesvirus 8 (HHV-8) have been identified in up to 29% of gingival biopsies of HIV-seronegative adults with periodontitis suggesting that the periodontium might constitute a reservoir for these viruses (17). The blood-borne viruses, including HBV, HCV, and HIV, may be present in the saliva of persons with chronic infection. Small cuts and abrasions in the oral cavity, especially when made acutely during dental or intraoral surgery, serve as the primary sources for seeding the saliva with virus. These viruses are addressed in the epidemiology section of this chapter. Many other viral agents can be recovered from oropharyngeal secretions during or after acute infection. These agents include polioviruses, coxsackieviruses and echoviruses, influenza viruses A and B, rhinoviruses, and coronaviruses. Despite the occasional isolation from saliva and nasal secretions of these viruses and the childhood respiratory pathogens rubeola (measles), mumps, and rubella, no occupationally related transmission of any of these viral agents to a dental worker has been documented except for one case of coxsackievirus infection (12). Dental procedures should be avoided in patients suspected of having active severe acute respiratory syndrome (SARS), which is caused by a newly discovered coronavirus and is likely transmitted by droplet and contact routes. Yeasts and fungi also are part of the normal flora of the oral cavity. Candida albicans can be isolated from the mouths of approximately 55% of healthy people. Many other species of Candida are found less often. The dorsum of the tongue has the greatest density of yeast. Carriage rates for yeast are increased among hospitalized patients, people with dentures, persons who are blood type O or nonsecretors, and people who are in the HIV-positive population (18, 19, 20). Transient carriage of filamentous soil fungi from the genera Penicillium, Aspergillus, Geotrichum, and others can be shown. Protozoa are also normal inhabitants of the healthy mouth. Entamoeba gingivalis and Trichomonas tenax can be demonstrated in more than 50% of the healthy population. TYPES OF INFECTIONS Infections of the oral cavity and maxillofacial regions can be grouped loosely into the categories of localized infection, infection by direct extension, and distant infection. Localized infections can be classified as dentoalveolar, periodontal, infections of the salivary glands or tonsils, and cellulitis from tissue injury. Dentoalveolar infections are also known as odontogenic infections and include carious teeth with resulting infections of the dental pulp and periapical dental abscess. Infections involving the gingiva, periodontal ligament, and other tissues supporting the teeth are known as periodontal infections. These infections include gingivitis and acute necrotizing ulcerative gingivitis (ANUG). Parotitis and sialoadenitis are infections of glands. Infections resulting from the direct extension of one or more of these localized infections include osteomyelitis of the mandible or maxilla, infection of the deep fascial spaces (e.g., submandibular, canine, retropharyngeal), maxillary sinusitis, noma (necrotizing infection of the cheek), posterior mediastinal infection, and anaerobic pulmonary infection. The anatomy of the deep fascial spaces is beyond the scope of this discussion but is well treated in standard texts (2, 4, 21). Distant infections that may develop secondary to oral infection include brain and liver abscesses and septic arthritis (22, 23, 24, 25, 26). Distant spread of bacteria from the oral cavity to implanted prosthetic devices via the bloodstream is well documented (27, 28). The risk of bacteremia may increase with increasing severity of gingival inflammation. P.915 PATHOGENESIS OF INFECTION Localized Infection Infection of the dental pulp may result from microbial penetration directly through the dentin secondary to dental caries, dental drilling, or tooth fracture or by hematogenous spread. The most common cause of pulpal infection is from dental caries that begins with the formation of dental plaque. Plaque is composed of a large number of bacteria (>108 CFU/mm3), including S. mutans, which firmly adhere to the enamel of the tooth. These bacteria secrete enzymes that progressively dissolve away the tooth enamel and dentin, permitting the bacteria to access the pulp (29). Microbial infection of the pulp (pulpitis) results and manifests clinically with pain and temperature sensitivity in the tooth. If the infection is not recognized and treated, the bacteria may then migrate through the pulpal foramen at the apex of the tooth into the alveolar bone at the root of the tooth, forming a periapical abscess, or extend beyond into the medullary space of the mandible, resulting in osteomyelitis. Needle aspiration and appropriate culture of pus from dentoalveolar abscesses reveal a polymicrobial flora (>2.4 isolates per specimen) with a predominance of facultatively anaerobic streptococci together with obligately anaerobic gram-positive cocci and gram-negative rods (30). More than 60% of these infections include aerobic microorganisms, whereas approximately one third have purely anaerobic isolates. Gingivitis is a periodontal process. Mild inflammation of the gums is present in almost all adolescents and in most American adults (31). Acute and chronic gingivitis begin with the formation of plaque below the gumline. Swelling and hyperemia of the free gum margin occurs, and the gums may bleed easily with brushing. Gingivitis is increased in frequency or severity in certain patient groups such as HIV-positive patients (up to 20% incidence), cancer patients undergoing chemotherapy, and young patients with type I diabetes mellitus (32). The healthy gingival sulcus has few microorganisms. Cessation of dental oral hygiene results in the appearance of dental plaque and gingivitis within 10 to 21 days. The most extreme form of gingivitis is ANUG. ANUG represents tissue invasion and destruction by mixed anaerobes and facultatively anaerobic bacteria. Data suggest an important role for spirochetes and for Fusobacterium species (33). In HIV-seropositive patients, yeasts and herpesviruses may also contribute (34). ANUG manifests as a loss of the papillae between adjacent teeth and results in exposure of the roots of the tooth. It is accompanied by systemic symptoms. The disease is characterized by the sudden onset of pain and tenderness of the gums associated with increased salivation and a peculiar metallic taste. Physical examination demonstrates bleeding of the gingivae with blunting and necrotic punched-out lesions of the interdental papillae. ANUG most often occurs in adolescents and young adults. Risk factors include poor oral hygiene, infrequent dental care, poor nutrition, and possibly diabetes (35). Prevalence studies have demonstrated that 4% of students using dental services at Harvard University, and 6.7% of 9,203 adolescents in Chile have this condition (35, 36). Periodontal infections usually begin with gingivitis. As the infection becomes chronic, it extends deeper into the junction between the tooth and gingiva. This leads to loss of the connective tissue attaching the tooth to the bone (the periodontal ligament) and resorption of the bone. The resulting periodontitis causes a pocket to form between the tooth and the gingiva. This space is ideal for the growth of anaerobes because of the very low reduction oxidation potential. Spirochetes of many morphotypes, some uncultivable, appear to be the most predominant bacteria in advanced lesions (37, 38). The chronic infection that occurs causes loosening and then loss of teeth. Periodontal abscesses result from infection of deep periodontal gingival pockets (39, 40). Acute suppurative parotitis is a nosocomial infection that occurs after surgery or in patients who are predisposed because of malnutrition, immunosuppression, or dehydration or in whom drugs have been used that decrease salivary flow (41). Such drugs include anticholinergic agents, antihistamines, and tranquilizers. The condition is unilateral in 80% to 90% of cases and presents clinically as the acute onset of unilateral facial swelling with pain. Physical examination demonstrates purulent fluid, which can be expressed from the parotid duct. The microbial causes reported in the older literature were Staphylococcus aureus in most patients, occasionally Streptococcus viridans, and rarely Actinomyces (41). Newer studies using proper anaerobic culture methods demonstrate anaerobes in most patients (42). The microorganisms are the same as those recovered from the gingival sulcus. Methicillin-resistant S. aureus has been reported as the cause of one outbreak in a nursing home (43). The pathogenesis of this infection is presumed to be retrograde movement of mouth microorganisms up the parotid duct in patients with diminished rates of salivary flow. Some authors report a correlation between increased numbers of bacteria in saliva and decreased rates of salivary flow (44). Acute tonsillitis is rarely an institutionally related infection unless an outbreak of acute group A ОІ-hemolytic streptococcal infection is spreading through the population. Although group A ОІ-hemolytic Streptococcus is the most commonly recognized cause of tonsillitis, the significance of recovery of other microorganisms such as Mycoplasma and Chlamydia from inflamed tonsils has been debated (45, 46, 47, 48). The pathogenetic role of these bacteria is not known. Microbiologic studies of the core of tonsils removed from 150 children with recurrent tonsillitis resulting from group A ОІ-hemolytic streptococci during three periods beginning in 1977 and ending in 1993 revealed mixed flora (8.1 microorganisms per tonsil) in all tonsils and an increased rate of recovery of ОІ-lactamase–producing bacteria with time (49). Erysipelas, a soft tissue infection of the cheek resulting from direct extension of bacteria from the mouth, often is due to group A or C streptococci. This rare complication follows 2 to 3 days after oral surgery and represents bacterial entry into soft tissues injured by instrumentation. Noma (gangrenous stomatitis) is an acute, fulminant, necrotizing infection of the cheek and facial tissue that destroys the oral and para-oral structures and is found predominantly in malnourished children, particularly in sub-Saharan Africa. Certain groups of adult patients may develop noma-like lesions that are slowly progressive. These adults have chronic lymphocytic leukemia, are receiving cytotoxic chemotherapy, or are neutropenic. The antecedent lesions to noma are believed to be oral herpetic P.916 ulcers, necrotizing gingivitis, or a buccal abrasion resulting from the rubbing of a tooth or from surgery (50, 51). Infection of these precursor lesions with synergistic bacteria, such as Fusobacterium and Prevotella, causes progressive full-thickness necrosis of the cheek, leaving a large open defect through which the mandible and tongue can be seen. Fusospirochetal infection with a mixed flora is the cause (51, 52). Cervicofacial actinomycosis is a rare disease most commonly caused by Actinomyces israelii. Healthy individuals are affected most often, although the incidence of infection is decreasing with time. This decline is possibly related to more frequent antibiotic use, better oral hygiene, and water fluoridation. The portal of entry is through disrupted mucosal barriers after trauma, dental manipulations, or oral and maxillofacial surgery (53, 54). The infection often appears as a chronic, slowly progressive induration or soft tissue mass in the mandibular-preauricular area and is sometimes accompanied by fistulous tracts to the skin that release sulfur-like granules. Systemic signs usually are absent (55). Primary oral tuberculous lesions are seen rarely (56, 57). Primary lesions occur in younger patients, are painless, and are associated with cervical lymphadenopathy. Secondary oral tuberculous lesions are more common and are seen mainly in older persons. Although the lesions are variable in appearance, the ulcerative form is the most usual, occurring on the tongue base or gingiva. These lesions are often painful. Most of these patients have accompanying active pulmonary tuberculosis (58, 59, 60). There are many oral complications from cancer therapy; one of the most prominent is infection. As a result of treatment effects on the mouth and immunosuppression, the oral cavity has the potential to become a reservoir for opportunistic microorganisms. Candida microorganisms are the primary cause of opportunistic fungal disease in patients who are immunocompromised. As many as 60% of cases of fungal septicemia in cancer patients are associated with prior oral infections (61). The most common oral manifestation of a candidal infection is pseudomembranous candidiasis, manifested by removable white curd-like plaques over an inflamed mucosa. Other forms include leukoplakia-like white plaques that are not removable, referred to as chronic hyperplastic candidiasis and chronic erythematous candidiasis that appears as patchy or diffuse mucosal erythema. Oral infections can extend to involve the esophagus. Candidiasis can be diagnosed by microscopic examination of a potassium hydroxide preparation of mucosal scrapings or a Gram-stained smear of mucosal scrapings or by culture. Aspergillosis is the second most frequent fungal infection in cancer patients, particularly in patients with hematologic malignancies (62). The paranasal sinuses are the most common sites of Aspergillus infection in the facial region, but there have been a few reports of primary oral aspergillosis (62, 63, 64, 65, 66, 67). The oral lesions initially manifest on the gingiva and then develop into necrotic ulcers covered by a pseudomembrane. Spread to the alveolar bone and facial muscles may occur rapidly. HSV is the most common viral pathogen in patients receiving cytotoxic agents or bone marrow transplants. The vesicular lesions on an erythematous base may appear anywhere on the mucosa and in addition to the mouth can involve the respiratory and gastrointestinal tracts. In immunocompromised patients, the oral mucositis associated with HSV may be particularly painful, severe, and prolonged. The oral HSV ulcerations may act as portals of entry for bacterial and fungal microorganisms. The most frequent viral infection following solid organ transplants is CMV. This infection can develop in high-risk transplant recipients despite ganciclovir or valganciclovir prophylaxis (68). Oral manifestations, when they occur, are nonspecific and require biopsy to confirm the cause. The infection often consists of a single, large, shallow ulceration (69). Infections by Direct Extension An epidemiologic retrospective study of hospitalized patients with maxillofacial infections noted differences between pediatric and adult patients. Upper face infections predominate in children (81%), whereas in adults, lower face infections, mainly odontogenic or peritonsillar, are more common (66%) (70). Osteomyelitis of the jaw (usually the mandible) most often results from chronic infection of a tooth, either from periapical abscess or from gingivitis. Other risk factors for osteomyelitis of the jaw include compound jaw fractures, diabetes mellitus, treatment with steroids, and surgery (50). Infection is particularly likely to occur when surgery is performed after irradiation of the mandible for tumor removal or after compound fracture of the mandible through the socket of a molar tooth. The causative microorganisms are S. aureus or anaerobes. Peritonsillar abscesses arise by direct extension from infected tonsils and tonsillar remnants and are rarely nosocomial in nature. It is critical to recognize and treat this infection to avoid respiratory compromise and other serious complications (71, 72). Maxillary sinusitis caused by mixed aerobes and anaerobes is recognized as a complication of periapical dental abscesses in the upper teeth (73, 74, 75). Sinusitis sometimes complicates extraction of the premolars and first molar on the upper side, because the root tips of these teeth almost touch the lower border of the maxillary sinuses. Retropharyngeal abscesses arise by direct extension from uncontrolled tonsillar infection or after perforation of the posterior pharyngeal wall by a foreign body. The foreign body may be a bone or another sharp object carried in the mouth. A retropharyngeal abscess presents initially with pharyngeal discomfort, limited neck motion, and nonspecific constitutional symptoms, including fever and chills (76). In its later stages, the abscess can be recognized by forward displacement of the posterior pharyngeal wall (77). A lateral soft tissue film of the neck or computed tomography of the neck is required for diagnosis and will demonstrate air fluid levels or pockets of air in the retropharyngeal space. Clinical differentiation between a retropharyngeal abscess and cellulitis of the retropharyngeal space is difficult and may be accomplished by performing needle aspiration of the area. A return of pus signifies an abscess (78). Prompt recognition and urgent surgical management by incision and drainage is the standard treatment because the retropharyngeal space directly communicates with the posterior mediastinum and life-threatening complications may occur rapidly (79, 80, 81, 82). However, there are recent reports of children being treated successfully without surgical intervention (76). P.917 Anaerobic pulmonary infection (“aspiration pneumonia”) resulting from mixed anaerobes and facultatively anaerobic microorganisms occurs after aspiration of oral contents. Clinical evidence suggests that the presence of severe gingivitis and/or oral surgery on the gums is associated with subsequent development of aspiration pneumonia. Clearly, the very large numbers of microorganisms (>1010 microorganisms per gram of tissue) found in gingival material provide a large inoculum if aspirated into the lungs. The interplay of local host defense and the frequency of dental procedures on patients with gingivitis suggests that local host defense usually overcomes this inoculum. As an extreme example of the aspiration of oral flora, pulmonary actinomycosis has been reported to follow partial full-mouth dental extractions by 4 years (83). Deep fascial space infections in the upper neck and underneath the jaw usually result from direct extension of odontogenic or oropharyngeal infection. Ludwig's angina is a diffuse fasciitis and cellulitis involving the submandibular and submental spaces. It is the result of a polymicrobial infection, often originating in the periodontal region (gingivitis, periapical abscess) or after injury to the gums or soft tissues (84). The fascial space involved secondary to infection of the maxillary or mandibular bicuspid and molar teeth depends on the relation of the root apex to the mylohyoid and buccinator muscles. Progression of infection upward may involve the whole side of the face, including the eyelids and orbit, whereas downward movement of infection can lead to necrotizing cervical fasciitis or mediastinitis (85, 86, 87, 88). The anatomic parameters influencing the spread of infection in these areas is beyond the scope of this chapter but is covered in other works (21, 31). Distant Infection Many distant abscesses have been reported as complications of dental and periodontal infection, including brain abscess, meningitis, paraspinal abscess, liver abscess, suppurative jugular thrombophlebitis, septic cavernous sinus thrombosis, septic arthritis, cellulitis, and necrotizing cavernositis of the penis (22, 23, 24, 25, 26, 89, 90, 91, 92, 93). The route of migration is held to be bacteremia. Hematogenous seeding of oral flora to native heart valves and prosthetic material is discussed in Chapter 67. EPIDEMIOLOGY The accuracy of published rates of infection for common dental procedures is limited generally by small numbers in the denominator. Even with limited data, infection rates for some common procedures appear to be very low. For example, one 1992 review of the complications of oral surgical procedures commented that of approximately 50 million intraoral injections of local anesthetic each year, a literature search turned up only two case reports of injection-associated infection (94). Even if this represents underreporting by 100 times, the rate is still 1 infection per 10,000 injections. Table 54.2 summarizes available data concerning the frequency of procedure-related infections (95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108). TABLE 54.2. NOSOCOMIAL INFECTION RATES FOR ORAL AND DENTAL OPERATIONS Dental caries and periapical abscess usually are not considered nosocomial infections because of their long incubation period and their association with poor oral hygiene and dental plaque formation. However, as noted previously, the periodontal flora begins to change within 10 days of stopping aggressive oral hygiene. Therefore, these complications may occur in head trauma, burns, and other patients with prolonged periods of unconsciousness or intubation. Fourrier et al. (109) studied the relationship between dental status and colonization of dental plaque by aerobic pathogens and the occurrence of nosocomial infections in 57 intensive care unit (ICU) patients in France. The amount of dental plaque increased during the ICU stay. Colonization of dental plaque was present in 40% of patients, either acquired or present on admission. A positive dental plaque culture was associated with the occurrence of nosocomial pneumonia and bacteremia. In 6 of 15 cases of ICU-related nosocomial infection, the pathogen isolated from dental plaque was the first identified source of the nosocomial pneumonia or bacteremia. The results from this study, and others, suggest dental plaque colonization and oral flora may be a source of nosocomial infection (110, 111). Several prophylactic oral measures have been tried to decrease nosocomial pneumonia. One prospective, randomized, double-blind, placebo-controlled trial in a cardiovascular ICU in a tertiary care center studied oropharyngeal decontamination with 0.12% chlorhexidine gluconate oral rinse (112). The nosocomial respiratory infection rate and the use of nonprophylactic systemic antibiotics in patients undergoing heart surgery were reduced. More recent studies have also demonstrated lower rates P.918 of nosocomial pneumonia in ventilated ICU patients treated with chlorhexidine oral decontamination when compared with control groups (113, 114) (see also Chapter 22). Acute suppurative parotitis was often reported in the past among patients undergoing general anesthesia for surgery. With better attention to adequate hydration and oral care and the widespread use of antibiotics in surgery, it is now reported to occur less than 0.5% of the time after use of a general anesthetic (94). Placement of dental implants is usually accompanied by the administration of prophylactic penicillin. When antibiotic prophylaxis was used, Larsen (97) reported five instances of wound dehiscence in 445 implants but without any evidence of inflammation or infection. The relationship between the preoperative administration of antibiotics and the success of 2,641 endosseous implants was investigated as part of the comprehensive Dental Implant Clinical Research Group clinical implant study (115). Overall implant failure rate was very low. When preoperative antibiotics were not used, the risk of implant failure increased two to three times, suggesting that antibiotic prophylaxis may be helpful. Osteotomy for correction of maxillary or mandibular deformities is occasionally followed by surgical site infection. One series recorded eight infections among 600 cases of sagittal split osteotomy, for a rate of 1.8%. Another retrospective analysis of 2,049 patients who underwent maxillofacial orthopedic surgery over a 21-year period reported only eight severe infections requiring incision and drainage, with no results compromised because of infection (116). Maxillary sinusitis after osteotomy of the maxilla is reported to be rare (105). The risk of bacteremia with dental manipulation has been quantified. One such study reported bacteremia in 72% of 183 patients undergoing one or more tooth extractions, and it occurred most often when teeth were extracted for inflammatory conditions (117). Seventy-one percent of isolates were anaerobes. Even minor oral manipulation such as periodontal probing leads to bacteremia in as many as 43% of patients and is more frequent in patients with periodontitis than in patients with chronic gingivitis (118, 119). Some authors suggest using antibacterial mouthwashes preprocedure to reduce gingival bacterial counts and the incidence of procedure-related bacteremias. Flood et al. (120) examined the frequency of bacteremia during routine incision and drainage of dentoalveolar abscesses. They demonstrated bacteremia during incision and drainage in 25% of such abscesses (3 of 12 patients). Bacteremia was found only during the drainage procedure in two of three patients, although, in the third instance, bacteremia was also demonstrated 5 minutes after the procedure ended. When abscesses were aspirated with a needle before incision and drainage, no blood cultures were positive. The authors suggested that the risk of bacteremia may be significantly reduced by needle aspiration of the abscess contents before incision and drainage. Oral infections may be an important cause of septicemia in patients with hematologic malignancies. Dens et al. (121) noted a marked reduction of salivary flow rate in patients after bone marrow transplant that was more pronounced if total body irradiation had been included in the pretransplant therapy. A higher concentration of cariogenic microorganisms and a shift toward a lower buffering capacity in saliva were found. These changes may lead to an increased risk of caries and oral complications posttransplant. Bergmann (122) prospectively followed 46 patients with hematologic malignancies through 78 febrile episodes. He estimated that a probable oral focus for septicemia was demonstrable in 10.5% of these individuals and that an oral origin was possible in an additional 21.1%. Other authors have sought a relationship between the mucositis that often follows chemotherapy for leukemia or bone marrow transplantation and the oral microbial flora as a way to explain the infections seen in these settings. Dreizen et al. (123) prospectively studied patients undergoing treatment for acute leukemia and found that 34.2% developed chemotherapy-related oral infection and 16.3% developed chemotherapy-related oral mucositis. In patients with mucositis, the viridans group of streptococci seems to be the most frequent cause of septicemia. Staphylococcus epidermidis, typically thought of as a skin bacterium, has been suggested to have an oral origin in a bone marrow transplant patient with bacteremia (124). Ferretti et al. (125) demonstrated that antimicrobial mouthwashes such as 0.12% chlorhexidine gluconate protected against these oral complications. Barker et al. (126) showed a possible reduction in incidence of О-hemolytic streptococcal sepsis among children receiving myelosuppressive chemotherapy who received prophylactic oral vancomycin paste. Various oral protocols for preventing oral sequelae in immunocompromised patients have been suggested (127, 128, 129, 130, 131, 132, 131). Not all studies show a correlation between the oral cavity and sepsis in patients with hematologic malignancies. A recent retrospective study of 77 patients after hematopoietic stem cell and bone marrow transplant showed no relationship between advanced periodontal disease and septicemia within the initial 100 days after transplant (132). Fortunately, the frequency of fungal and viral oral infections in cancer and transplant patients has been reduced dramatically by the institution of prophylactic agents. However, one of the unfortunate potential consequences of this practice is the emergence of resistant microorganisms. Oral candidiasis and hairy leukoplakia are conditions that should trigger an assessment of HIV infection risk factors. Many oral diseases in persons with HIV infection appear to be modified presentations of conventional disorders, such as gingivitis, necrotizing periodontal diseases, and exacerbated periodontitis (133, 134). Lower frequencies of oral disease have been seen in those on HIV therapy and in those who receive regular oral healthcare (135, 136). There are insufficient studies on the complication risks for HIV-positive patients undergoing invasive dental procedures (137). Literature on the complications of tongue piercing is beginning to emerge. Adverse outcomes have occasionally been life threatening. The reported complications include pain; tongue swelling; tongue abscess; airway problems; profuse bleeding; mucosal or gingival trauma or recession; chipped or fractured teeth; and interference with speech, mastication, and swallowing (138, 139, 140, 141). In addition, there have been reports of endocarditis resulting from Neisseria mucosa, Haemophilus aphrophilus, and methicillin-resistant S. aureus following tongue piercing (142, 143, 144). Polymicrobial cerebellar brain abscess subsequent P.919 to tongue piercing has also been noted (145). There appears to be fewer problems associated with lip piercing, although gingival recession of previously healthy tissue has been seen (146). Viral Agents HSV is the latent herpesvirus in the oral cavity most commonly expressed during hospitalization. The virus is latent in the trigeminal ganglion and is secreted in saliva from the parotid gland. Reactivation of labial lesions in a patient is triggered by oral surgical procedures, trauma, ultraviolet light, and major injuries such as burns. Because of the frequency of asymptomatic excretion of the virus in saliva (virus can be recovered at intervals from the saliva of more than 50% of adults in the United States), the unprotected hands of dental surgeons, dental hygienists, and oral surgeons, which are bathed in saliva, are exposed to HSV. A common result of such exposure has been herpetic whitlow, which is a painful infection localized to the periungual region of the fingernail. This is recognized as an occupational hazard for dental workers (147). Because acquisition of HSV infection depends on direct contact between saliva or active lesions and an opening in skin, the frequency of herpetic whitlow possibly has been reduced by the use of gloves for blood-borne disease precautions (also see Chapters 43 and 81). Latex and vinyl gloves have been shown to be impermeable to HSV and to protect against HSV infection (148). Despite their recovery from saliva and associated oral tissues, EBV and CMV have not been recognized as nosocomially important pathogens in the dental or oral surgery setting, either for spread to other patients or for spread to staff. However, a seroprevalence study done in the United Kingdom suggests possible occupational risk of infection with EBV in dentists based on a higher seroprevalence to EBV among clinical dental students and qualified dentists than among preclinical dental students (149). Another seroprevalence study done in England showed a greater prevalence of antibodies to influenza A and B viruses and respiratory syncytial virus in dental surgeons compared with control subjects, suggesting occupational risk for respiratory virus infections (150). Based on questionnaire results, reported donning of masks did not reduce seroprevalence with these viruses. Blood-borne Pathogens HBV may be transmitted both from patients to dentists and oral surgeons and from oral surgeons and dentists to patients (151, 152, 153, 154). HBV can cause a chronic latent infection of the liver and is associated with large numbers of virus particles circulating in the blood of chronically infected persons. Because all intraoral surgery and many dental procedures cause breaks in the mucosa of the oral cavity or gums that result in bleeding, the risk of spread of HBV from the patient to the operator is substantial (155). In prevaccine surveys, the annual incidence of HBV was five to ten times higher among physicians and dentists than among blood donors (154, 156). Infections occur when blood from the patient enters the body of the dentist through small breaks in the skin. In recent years, gloves have been used routinely as part of Standard/Universal Precautions. However, exposures of breaks in the skin to blood still occur because of glove perforation. Glove puncture occurs in 2.1% to 16% of oral surgery procedures (157, 158). Aerosol transmission from high-speed drills used in dentistry with resulting aerosolization of saliva and blood has never been documented to result in occupationally related infections. Transmission of HBV via a human bite has been reported (159, 160). Several outbreaks of HBV infection have been reported among patients who underwent surgery by oral surgeons chronically infected with HBV (12, 152, 155). The precise mechanism(s) resulting in transmission of infection has not been determined, but infection was likely transmitted from dental workers to patients rather than from one patient to another. These outbreaks usually have been terminated after the involved surgeon began to wear gloves when performing procedures. The authors are unaware of any reports of transmission of HBV from dentists to patients since 1987. This may be due to increased adherence to universal precautions/standard precautions, higher levels of immunity resulting from use of hepatitis B vaccine, incomplete reporting, or isolated sporadic cases that are difficult to associate with a dental worker (161). HCV was identified in 1989 and is recognized as the main agent of what was previously termed non-A, non-B viral hepatitis. The virus causes chronic latent infection of the liver in most infected persons. In the U.S. general population it is estimated that 3.9 million people have been infected with HCV, of whom 2.7 million have chronic infection. The prevalence of antibody to HCV in the homeless and incarcerated may be as high as 40% (162). In the United States, injecting drug users account for one half of all newly infected persons (163). HCV RNA is variably detected in the saliva of infected persons (164, 165, 166). Oral surgery appears to increase the occurrence of HCV in saliva (167). Transmission of HCV through a human bite has been reported (168, 169). Like HBV, HCV is a known occupational hazard for healthcare personnel by contact with contaminated blood, although HCV seems to be transmitted in the occupational setting less efficiently than HBV. To date, no dental worker is known to have acquired HCV occupationally, but the high frequency of sharps injuries occurring in the dental setting places the dental worker at risk of HCV acquisition (170). Despite this risk, the prevalence of HCV infection among dental workers appears to be similar to that of the general population. Anti-HCV may be more common in dental workers who are older, have more years of practice, and have serologic markers of HBV infection (171). A review of self-reported and observational studies of occupational blood exposures among U.S. dental workers between 1986 and 1995 suggested that percutaneous injuries steadily declined to an average of three injuries per year (172). The data on the frequency of transmission of HCV during dental care are very limited. Gingivectomy performed by a dental surgeon of unknown HCV status was identified as the only risk factor for the seroconversion of one patient (173). A case-control study found an association between dental treatment and HCV positivity (174). HIV infection causes a chronic infection of human lymphocytes and many other cell types and has a latency period of at least several years before onset of symptoms. The epidemiology P.920 of HIV in the medical setting likely is the same as that of HBV (175). As of December 2001, the Centers for Disease Control and Prevention received reports of 57 U.S. healthcare workers with documented HIV conversion temporally associated with an occupational HIV exposure. An additional 138 cases are considered to have possibly been acquired occupationally, but the source of infection cannot be documented with certainty. No dental personnel are among any of the documented cases, but six dental workers are in the group of possible occupational transmissions (176). Occupationally acquired HIV infection recognized among healthcare workers most commonly resulted from blood transmitted by hollow-bore needles (176). Because hollow-bore needles are used less often in dental practice, the risk of occupationally acquired HIV infection for dental workers may be slightly lower than that for some other groups of healthcare workers. A serosurvey combined with a questionnaire administered to 321 oral and maxillofacial surgeons revealed no HIV-seropositive participants despite a mean number of recalled percutaneous injuries within the previous year of 2.4 (most commonly associated with wire) (177). The results imply a low occupational risk for HIV infection. In the United States, the only documented transmission of HIV from an operating surgeon to a patient occurred with one cluster of six cases related to a single dentist in Florida (178). The events that resulted in the infection of these patients remain unknown, although the evidence suggests that HIV was transmitted from dentist to patient rather than from patient to patient (179). PREVENTION AND CONTROL Infections may be transmitted in the dental operatory through direct contact with blood, oral fluids, or other secretions; via indirect contact with contaminated instruments, equipment, or environmental surfaces; or by contact with airborne contaminants present in either droplet splatter or aerosols of oral and respiratory fluids (161). Strategies to prevent dental patient infections have focused on four areas. The first is sterilization of all instruments used in intraoral procedures and disinfection of related equipment. The second is use of good infection control practices in the dental operatory (161). These measures are aimed at preventing the spread of an infectious agent on instruments or dental apparatus from one patient to another. The third is rigid asepsis during intraoral procedures, including the use of preprocedure mouthwashes to reduce the burden of intraoral flora. Local antisepsis with topically applied antiseptic agents must be used particularly for root canal work, endodontic procedures, and gum surgery. The fourth is antibiotic prophylaxis or treatment of infected areas in which work is performed. These measures are directed at preventing the entry of the patient's own resident oral and gingival flora into the operative site in numbers great enough to cause infection. Instruments used to penetrate soft tissue or bone (forceps, scalpels, bone chisels, etc.) are classified as critical and should be sterilized after each use. Instruments that are not intended to penetrate oral soft tissues or bone such as mirrors and amalgam condensers but may come in contact with oral tissues are classified as semicritical and also should be sterilized after each use. If a semicritical item will be damaged by heat sterilization, the instrument should receive, at a minimum, high-level disinfection. Instruments or devices that come into contact only with intact skin such as external components of x-ray heads are classified as noncritical. These items may be reprocessed between patients with intermediate-level or low-level disinfection or detergent and water washing, depending on the type of surface and the degree and nature of the contamination (161). Before sterilization or high-level disinfection, instruments should be cleaned thoroughly to remove debris. They should be placed into a presoak solution immediately after use to prevent the drying of saliva or blood on the instruments and to make cleaning easier. The soak contains an antimicrobial agent to reduce the levels of bacteria and viruses. Cleaning is accomplished by scrubbing in a detergent solution or, preferably, to minimize handling and the exposure of workers to sharps injuries, by placing the instruments into a mechanical device (an ultrasonic cleaner). After cleansing, the instruments should be thoroughly rinsed with water while they are still in the cleaning basket, inspected carefully to make sure all visible debris has been removed, and then allowed to dry. Critical and semicritical instruments then should be sterilized. The most common forms of sterilization used in a dental office include steam (autoclaving) and dry heat. Sterilization processes must be monitored and periodically tested for efficacy using bacterial spores (see Chapter 86). The use of liquid chemical germicides for high-level disinfection of heat-sensitive semicritical instruments may require up to 10 hours of exposure. Indications for wet sterilization are very limited, and manufacturers' directions regarding the correct concentration and exposure time should be followed closely. The process should be followed by aseptic rinsing with sterile water, drying, and, if not used immediately, placing in a sterile container (161) (see Chapter 85). Because of the transmission of both HIV and HBV in the dental setting, much concern has been expressed over the dental handpieces used to transmit rotary energy to dental drills (180, 181). These handpieces are composed of a number of moving parts and typically have many cracks and crevices, which make them difficult to clean. They cannot be adequately disinfected by wet disinfectants (i.e., glutaraldehyde) because the agent cannot penetrate into the crevices. Studies have shown that residual live bacteria are recoverable from handpieces even after cleaning and wet chemical disinfection. Because all currently manufactured high-speed handpieces and most low-speed handpieces are heat tolerant, these items should be cleaned and lubricated, followed by sterilization between successive patients. Handpieces that are not heat tolerant should be modified to make them tolerant to heat. Those that cannot be heat sterilized should not be used (161, 182). Another potential concern is of dental unit water systems supplying dental handpieces and air water syringes becoming contaminated with microorganisms from the incoming water supply and, less often, with oral flora (183, 184). Methods to minimize this risk include use of various flushing techniques (183, 185). P.921 Good infection control practices in the dental operatory are directed at the use of hand washing, appropriate barrier precautions, and attention to reducing the contamination of environmental surfaces by saliva and blood. These measures include the use of impervious paper or plastic covers to protect surfaces that may become contaminated during use and that are difficult to disinfect. Such surfaces include x-ray unit heads and light handles. Between patients, the coverings should be removed and replaced and all flat surfaces potentially soiled with patient material should be wiped off with an Environmental Protection Agency-approved hospital disinfectant, also labeled as “tuberculocidal.” These intermediate-level disinfectants are effective against most bacteria and viruses (161). Other methods to reduce salivary contamination of the operatory include patient use of a mouthrinse before the procedure, use of a rubber dam, and use of a high-speed air evacuator during high-speed drilling. Creutzfeldt-Jakob disease, one of the transmissible spongiform encephalopathies (TSEs), is a rapidly progressive, invariably fatal neurodegenerative disorder believed to be caused by an abnormal isoform of a cellular glycoprotein known as the prion protein. Although epidemiologic investigation has not revealed any evidence that dental procedures lead to increased risk of iatrogenic transmission of TSEs among humans, studies have shown that infected animals develop infectivity in gingival and dental pulp tissues. Transmission to healthy animals can occur by exposing root canals and gingival abrasions to infectious brain homogenate. The World Health Organization guidelines suggest that the usual infection control guidelines are sufficient when treating patients with TSEs during procedures not involving neurovascular tissue (186). Extra precautions should be considered for major dental procedures (186) (see also Chapter 47). Local antiseptics reduce bacterial contamination of the operative site. They are applied to the prepared sites of dental fillings for caries, crowns, and root canals before closing the defect. Topical disinfection of the gingiva with H2O2 or chlorhexidine mouthwashes before elective gingival surgery reduces the rates of soft tissue infection. Marten and van Saene (131) discussed methods to prevent each of the seven major oral infectious complications of cancer therapy. Four of these complications (caries, osteomyelitis, periodontal disease, and mucositis) can be prevented by strict application of local measures in the mouth. These measures include good oral hygiene to prevent caries and periodontal disease and to reduce the likelihood of osteomyelitis and topically applied antimicrobials to prevent osteomyelitis and mucositis. Systemic antibiotics have a more limited role in reducing the rate of infectious complications. Converse and McCarthy (187) list the following indications for the use of prophylactic antibiotics for surgery on the jaw (mandible): (a) use of an intraoral approach, (b) previous irradiation of the operative field, (c) use of a bone graft, (d) use of an alloplastic implant, and (e) surgery in a patient prone to infection (diabetes patient). As demonstrated in Table 54.2, the rate of surgical site infection is increased when a transmucosal or intraoral approach is used or when the socket of a tooth is involved in a fracture. This risk rises because of the large number of bacteria in the gingival sulcuses. In addition to the situations listed by Converse and McCarthy, systemic antibiotic administration is of proven benefit for transoral procedures more than 3 hours in length and for orthognathic or other major maxillofacial surgery. In addition, patients may require antibiotics to protect heart valves or other distant foci from bacteremia originating in the mouth. Routine antibiotic prophylaxis to prevent hematogenous prosthetic joint infections is not recommended, although premedication may be warranted in some patients (188). SAFETY FOR DENTAL HEALTHCARE WORKERS Worker safety is provided by the following measures (161): Every dentist, oral surgeon, or assistant who is exposed to blood or blood products or who handles needles or sharp instruments within the office should be offered and encouraged to accept hepatitis B vaccination. Standard/universal precautions for handling all patient-related and derived substances should be rigorously used. These include the use of gloves for any work done in or around the mouth and for handling any instruments, surfaces, or substances contaminated with saliva or blood. Goggles and a mask or a face shield should be worn during all procedures. Any visible wounds suffered from sharp instruments should be immediately cleansed and assessed. The assessment should include informed consent for testing the patient for HBV, HCV, and HIV according to applicable state and federal laws. In addition, a postexposure prophylaxis protocol should be followed (189, 190). In addition to glove use, hands should be washed after removing gloves. Gowns or dental clinic jackets should be worn in the office and operatory setting. These items should be removed on leaving work and laundered separately from personal clothing. Because many studies have shown widespread salivary contamination of surfaces in the operatory, barrier protection of commonly touched surfaces such as radiographic handles and controls, bucket handles, and light switches should be provided. The bacterial content of saliva can be reduced by rinsing the patient's mouth with water before any dental examination. Additional reductions can be accomplished by use of a mouthwash. A careful general medical history should be completed for all new patients and at periodic intervals to look for symptoms or signs of pulmonary tuberculosis. Patients with signs or symptoms suggestive of tuberculosis should not undergo elective dental or oral procedures until they have been evaluated by a physician. All dental office workers should receive a yearly test for tuberculosis (191). These measures are the minimum for providing reasonable protection to dental staff and patients against communicable diseases. Numerous studies demonstrate improved compliance by dental and oral surgery personnel with recommended infection control practices in recent years (192, 193, 194, 195, 196, 197). However, the results of these studies suggest that further education and encouragement P.922 are needed to attain a more desirable level of understanding and adherence to the recommended practices. REFERENCES 1. Ciesielski C, Marianos D, Chin-Yiu OU, et al. Transmission of human immunodeficiency virus in a dental practice. Ann Intern Med 1992;116:798–805. 2. Krizek TJ, Arigar S. Infection. In: Conley JJ, ed. Complications of head and neck surgery. Philadelphia: WB Saunders, 1979:55. 3. Heimdahl A, Nord CE. Antimicrobial prophylaxis in oral surgery. Scand J Infect Dis Suppl 1990;70:91–101. 4. Schuster GS. Microbiology of oral and maxillofacial infections. In: Topazian RG, Goldberg MH, eds. Oral and maxillofacial infections, 3rd ed. Philadelphia: WB Saunders, 1994. 5. Long SS, Swenson RM. Determinants of the developing oral flora in normal newborns. Appl Environ Microbiol 1976;32:494. 6. Mantilla GS, Danser MM, Sipos PM, et al. Tongue coating and salivary bacterial counts in healthy/gingivitis subjects and periodontitis patients. J Clin Periodontol 2001;28:970–978. 7. Mackiowak PA. The normal microbial flora. N Engl J Med 1983;307:83. 8. Schuster GS, ed. Oral microbiology and infectious disease, 3rd ed. St. Louis: CV Mosby, 1990. 9. Chen CK, Wilson ME. Eikenella corrodens in human oral and non-oral infections: a review. J Periodontol 1992;63:941–953. 10. Danser MM, van Winkelhoff AJ, de Graaff J, et al. Short-term effect of full-mouth extraction on periodontal pathogens colonizing the oral mucous membranes. J Clin Periodontol 1994;21:484–489. 11. Genco RJ, Loos BG. The use of genomic DNA fingerprinting in studies of the epidemiology of bacteria in periodontitis. J Clin Periodontol 1991;18:396–405. 12. Dolan MM, Yankell SL. Transmissible infections in dentistry. In: Slots J, Taubman MA, eds. Contemporary oral microbiology and immunology. St. Louis: Mosby Year Book, 1992. 13. Corey L, Spear PG. Infections with herpes simplex virus. N Engl J Med 1986;314:686–691, 749–757. 14. Gold E, Nankervis GA. Cytomegalovirus. In: Evans AS, ed. Viral infections of humans: epidemiology and control, 3rd ed. New York: Plenum, 1991. 15. Pomeroy C, Englund JA. Cytomegalovirus: epidemiology and infection control. Am J Infect Control 1987;15:107–119. 16. Evans AS, Niederman JC. Epstein-Barr virus. In: Evans AS, ed. Viral infections of humans: epidemiology and control, 3rd ed. New York: Plenum, 1991. 17. Mardirossian A, Contreras A, Navazesh M, et al. Herpesviruses 6, 7, and 8 in HIV- and non-HIV-associated periodontitis. J Periodontal Res 2000;35:278–284. 18. Stenderup A. Oral mycology. Acta Odontol Scand 1990;48:3–10. 19. Hunter KD, Gibson J, Lockhart P, et al. Fluconazole-resistant Candida species in the oral flora of fluconazole-exposed HIV-positive patients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:558–564. 20. Korting HC, Ollert M, Georgii A, et al. In vitro susceptibilities and biotypes of Candida albicans isolates from the oral cavities of patients infected with human immunodeficiency virus. J Clin Microbiol 1989;26:2626–2631. 21. Chow AW. Infections of the oral cavity, head and neck. In: Mandell G, Douglas RG, Bennett JE, eds. Principles and practice of infectious diseases, 4th ed. New York: Churchill Livingstone, 1995:593–606. 22. Saal CJ, Mason JC, Cheuk SL, et al. Brain abscess from chronic odontogenic cause: report of case. J Am Dent Assoc 1988;117:453–455. 23. Andersen WC, Horton HL. Parietal lobe abscess after routine periodontal recall therapy. Report of a case. J Periodontal 1990;61:243–247. 24. Crippin JS, Wang KK. An unrecognized etiology for pyogenic hepatic abscesses in normal hosts: dental disease. Am J Gastroenterol 1992;87:1740–1743. 25. Kawamata T, Takeshita M, Ishizuka N, et al. Patent foramen ovale as a possible risk factor for cryptogenic brain abscess: report of two cases. Neurosurgery 2001;49:204–206. 26. Edson RS, Osmon DR, Berry DJ. Septic arthritis due to Streptococcus sanguis. Mayo Clin Proc 2002;77:709–710. 27. Field EA, Martin MV. Prophylactic antibiotics for patients with artificial joint undergoing oral and dental surgery: necessary or not? Br J Oral Maxillofacial Surg 1991;29:341–346. 28. van Winkelhoff AJ, Overbeek BP, Pavicic MJAMP, et al. Long-standing bacteremia caused by oral Actinobacillus actinomycetemcomitans in a patient with a pacemaker. Clin Infect Dis 1993;16:216–218. 29. Thoden van Velzen SK, Abraham-Inpijn L, Moorer WR. Plaque and systemic disease: a reappraisal of the focal infection concept. J Clin Periodontol 1984;11:209–220. 30. Lewis MA, MacFarlane TW, McGowan DA. A microbiological and clinical review of the acute dentoalveolar abscess. Br J Oral Maxillofac Surg 1990;28:359–366. 31. Trummel CL. Periodontal infections. In: Topazian RG, Goldberg MH, eds. Oral and maxillofacial infection, 3rd ed. Philadelphia: WB Saunders, 1994. 32. Slots J, Rams TE. New views on periodontal microbiota in special patient categories. J Clin Periodontol 1991;18:411–420. 33. Tanner A. Microbial etiology of periodontal diseases. Where are we? Where are we going? Curr Opin Dent 1992;2:12–24. 34. Cobb CM, Ferguson BL, Keselyak NT, et al. A TEM/SEM study of the microbial plaque overlying the necrotic gingival papillae of HIV-seropositive, necrotizing ulcerative periodontitis. J Periodontal Res 2003;38:147–155. 35. Lopez R, Fernandez O, Jara G, et al. Epidemiology of necrotizing ulcerative gingival lesions in adolescents. J Periodontal Res 2002;37:439–444. 36. Burket LW, Greenberg MS. Gingival diseases. In: Lynch MA, ed. Oral medicine diagnosis and treatment, 7th ed. Philadelphia: JB Lippincott, 1977. 37. Moter A, Hoenig C, Choi BK, et al. Molecular epidemiology of oral treponemes associated with periodontal disease. J Clin Microbiol 1998;36:1399–1403. 38. Sela MN. Role of Treponema denticola in periodontal diseases. Crit Rev Oral Biol Med 2001;12:399–413. 39. Hirsch RS, Clarke NG. Infection and periodontal diseases. Rev Infect Dis 1989;11:707–715. 40. Kerkes K, Olsen I. Similarities in the microflora of root canals and deep periodontal pockets. Endodontics Dent Traumatol 1990;6:1–5. 41. Brook I. Diagnosis and management of parotitis. Arch Otolaryngol 1992;118:469–471. 42. Brook I. Acute bacterial suppurative parotitis: microbiology and management. J Craniofac Surg 2003;14:37–40. 43. Raad II, Sabbagh MF, Caranasos GJ. Acute bacterial sialadenitis: a study of 29 cases and review. Rev Infect Dis 1990;12:591–601. 44. Fox PC. Bacterial infection of salivary glands. Curr Opin Dent 1991;1:411–414. 45. Huminer D, Levy R, Pitlik S, et al. Mycoplasma and chlamydia in adenoids and tonsils of children undergoing adenoidectomy or tonsillectomy. Ann Otol Rhinol Laryngol 1994;103:135–138. 46. Charnock DR, Chapman GD, Taylor RE, et al. Recurrent tonsillitis. Arch Otolaryngol Head Neck Surg 1992;118:507–508. 47. Hone SW, Moore J, Fenton J, et al. The role of Chlamydia pneumoniae in severe acute tonsillitis. J Laryngol Otol 1994;108:135–137. 48. Kobayashi S, Tamura N, Ichikawa G, et al. Infection related arthritis induced by tonsillar Chlamydia trachomatis and streptococcal infection. Clin Exp Rheumatol 2002;20:732. 49. Brook I, Yocum P, Foote PA. Changes in the core tonsillar bacteriology of recurrent tonsillitis: 1977–1993. CID 1995;21:171–176. 50. Topazian RG. Uncommon infections of the oral and maxillofacial regions. In: Topazian RG, Goldberg MH, eds. Oral and maxillofacial infections, 3rd ed. Philadelphia: WB Saunders, 1994. 51. Enwonwu CO, Falkler WA, Idigbe EO. Oro-facial gangrene (noma/cancrum oris): pathogenetic mechanisms. Crit Rev Oral Biol Med 2000;11:159–171. P.923 52. Paster BJ, Falkler WA Jr, Enwonwu CO, et al. Prevalent bacterial species and novel phylotypes in advanced noma lesions. J Clin Microbiol 2002;40:2187–2191. 53. Alderson G, Aufdemorte TB, Jones AC, et al. Oral and maxillofacial pathology case of the month. Actinomycosis. Tex Dent J 2000;117:54, 92–93. 54. Maurer P, Otto C, Eckert AW, et al. Actinomycosis as a rare complication of orthognathic surgery. Int J Adult Orthodon Orthognath Surg 2002;17:230–233. 55. Miller M, Haddad AJ. Cervicofacial actinomycosis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:496–508. 56. Carnelio S, Rodrigues G. Primary lingual tuberculosis: a case report with review of literature. J Oral Sci 2002;44:55–57. 57. Rivera H, Correa MF, Castillo-Castillo S, et al. Primary oral tuberculosis: a report of a case diagnosed by polymerase chain reaction. Oral Dis 2003;9:46–48. 58. Eng HL, Lu SY, Yang CH, et al. Oral tuberculosis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:415–420. 59. Hashimoto Y, Tanioko H. Primary tuberculosis of the tongue: report of a case. J Oral Maxillofacial Surg 1989;47:744–746. 60. Smith WHR, Davis D, Mason KD, et al. Intraoral and pulmonary tuberculosis following dental manipulation. Lancet 1982;1:842–843. 61. Greenberg MS, Cohen SG, McKitrick JC, et al. The oral flora as a source of septicemia in patients with acute leukemia. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1982;53:32–35. 62. Dreizen S, Keating MJ, Beran M. Orofacial fungal infections: nine pathogens that may invade during chemotherapy. Postgrad Med 1992;91:349–364. 63. Talbot SH, Huang A, Provencher M. Invasive aspergillus rhinosinusitis in patients with acute leukemia. Rev Infect Dis 1991;13:219–232. 64. Sugata T, Myoken Y, Kyo T, et al. Invasive oral aspergillosis in immunocompromised patients with leukemia. J Oral Maxillofac Surg 1994;52:382–386. 65. Sugata T, Myoken Y, Kyo T, et al. Oral aspergillosis in compromised patients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:632. 66. Myoken Y, Sugata T, Fujita Y, et al. Molecular epidemiology of invasive stomatitis due to Aspergillus flavus in patients with acute leukemia. J Oral Pathol Med 2003;32:215–218. 67. Khoury H, Poh CF, Williams M, et al. Acute myelogenous leukemia complicated by acute necrotizing ulcerative gingivitis due to Aspergillus terreus. Leuk Lymphoma 2003;44:709–713. 68. Akalin E, Sehgal V, Ames S, et al. Cytomegalovirus disease in high-risk transplant recipients despite ganciclovir or valganciclovir prophylaxis. Am J Transplant 2003;3:731–735. 69. Epstein JB, Sherlock CH, Wolber RA. Oral manifestations of cytomegalovirus infection. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1993;75:443–451. 70. Scutari P, Dodson TB. Epidemiologic review of pediatric and adult maxillofacial infections in hospitalized patients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:270–274. 71. Maharaj D, Rajah V, Hemsley S. Management of peritonsillar abscess. J Laryngol Otol 1991;105:743–745. 72. Stevens NE. Vascular complication of neck space infection: case report and literature review. J Otolaryngol 1990;19:206–210. 73. Randall CJ, Jefferis AF. Quinsy following tonsillectomy (five case reports). J Laryngol Otol 1984;98:367–369. 74. Brook I, Frazier EH, Gher ME Jr. Microbiology of periapical abscesses and associated maxillary sinusitis. J Periodontol 1996;67:608–610. 75. Paju S, Bernstein JM, Haase EM, et al. Molecular analysis of bacterial flora associated with chronically inflamed maxillary sinuses. J Med Microbiol 2003;52:591–597. 76. Craig FW, Schunk JE. Retropharyngeal abscess in children: clinical presentation, utility of imaging, and current management. Pediatrics 2003;111:1394–1398. 77. Cottrell DA, Bankoff M, Norris LH. Computed-tomography guided percutaneous drainage of a head and neck infection. J Oral Maxillofac Surg 1992;50:1119. 78. Herzon FS. Needle aspiration of nonperitonsillar head and neck abscesses. Arch Otolaryngol Head Neck Surg 1988;114:1312. 79. Levine TM, Wurster CF, Krespi YP. Mediastinitis occurring as a complication of odontogenic infections. Laryngoscope 1986;96:747–750. 80. Wenig BL, Shikowitz MJ, Abramson AL. Necrotizing fasciitis as a lethal complication of peritonsillar abscess. Laryngoscope 1984;94:1576-1579. 81. Waggie Z, Hatherill M, Millar A, et al. Retropharyngeal abscess complicated by carotid artery rupture. Pediatr Crit Care Med 2002;3:303–304. 82. Hari MS, Nirvala KD. Retropharyngeal abscess presenting with upper airway obstruction. Anaesthesia 2003;58:714–715. 83. Suzuki JB, Delisle AL. Pulmonary actinomycosis of periodontal origin. J Periodontol 1984;55:581–584. 84. Balcerak RJ, Sisto JM, Bosack RC. Cervicofacial necrotizing fasciitis: report of three cases and literature review. J Oral Maxillofac Surg 1988;46:450. 85. Thakar M, Thakar A. Odontogenic orbital cellulitis. Report of a case and considerations on route of spread. Acta Ophthalmol Scand 1995;73:470–471. 86. Chan CH, McGurk M. Cervical necrotising fasciitis—a rare complication of periodontal disease. Br Dent J 1997;183:293–296. 87. Newton CL, deLemos D, Abramo TJ, et al. Cervical necrotizing fasciitis caused by Serratia marcescens in a 2 year old. Pediatr Emerg Care 2002;18:433–435. 88. Haraden BM, Zwemer FL. Descending necrotizing mediastinitis: complication of a simple dental infection. Ann Emerg Med 1997;29:683–686. 89. Weesner CL, Cisek JE. Lemierre syndrome: the forgotten disease. Ann Emerg Med 1993;22:256–258. 90. Pearle MS, Wendel EF. Necrotizing cavernositis secondary to periodontal abscess. J Urol 1993;149:1137–1138. 91. Montejo M, Aguirrebengoe K. Streptococcus oralis meningitis after dental manipulation. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85;126–127. 92. Larkin EB, Scott SD. Metastatic paraspinal abscess and paraplegia secondary to dental extraction. Br Dent J 1994;177:340–342. 93. Manian FA. Cellulitis associated with an oral source of infection in breast cancer patients: report of two cases. Scand J Infect Dis 1997;29:421–422. 94. Malamed SF, Sykes P, Kubota Y, et al. Local anesthetic: a review. Anesth Pain Control Dent 1992;1:11–24. 95. Osbourne TP, Fredericksen G, Small IA, et al. A prospective study of complications related to mandibular third molar surgery. J Oral Maxillofac Surg 1985;43:7671769. 96. Larsen PE. Role of antimicrobials for dentoalveolar surgery. J Oral Maxillofac Surg 1993;51(Suppl 3):155. 97. Larsen PE. Antibiotic prophylaxis for placement of dental implants. J Oral Maxillofac Surg 1993;51(Suppl 3):194. 98. Buser D, Ingimarsson S, Dula K, et al. Long-term stability of osseointegrated implants in augmented bone: a 5-year prospective study in partially edentulous patients. Int J Periodontics Restorative Dent 2002;22:109–117. 99. Romeo E, Chiapasco M, Ghisolfi M, et al. Long-term clinical effectiveness of oral implants in the treatment of partial edentulism. Seven-year life table analysis of a prospective study with ITI dental implants system used for single-tooth restorations. Clin Oral Implants Res 2002;13:133–143. 100. Islamoglu K, Coskunfirat OK, Tetik G, et al. Complications and removal rates of miniplates and screws used for maxillofacial fractures. Ann Plast Surg 2002;48:265–268. 101. Assael L, Hamman K. A comparison of rigid internal fixation with wire osteosynthesis of mandibular fracture. Case reports and outlines of scientific sessions. J Oral Maxillofac Surg 1987;45:M5–M6. 102. Assael LA. Infection in the maxillofacial trauma patient. In: Topazian RG, Goldberg MH, eds. Oral and maxillofacial infections, 3rd ed. Philadelphia: WB Saunders, 1994. 103. Peterson LJ. Antibiotic prophylaxis against wound infections in oral and maxillofacial surgery. J Oral Maxillofacial Surg 1990;48:617–620. 104. Peterson LJ. Principles of antibiotics. In: Topazian RG, Goldberg MH, eds. Oral and maxillofacial infections, 3rd ed. Philadelphia: WB Saunders, 1994. P.924 105. Hinds EC, Kent JN. Surgical treatment of developmental jaw deformities. St. Louis: CV Mosby, 1972. 106. Dingaman RO, Dingaman DL, Lawrence DA. Surgical correction of lesions in the temporomandibular joint. Plastic Reconstruct Surg 1978;55:335. 107. Tarro AW. Arthroscopic treatment of anterior disc displacement: a preliminary report. J Oral Maxillofacial Surg 1989;47:353. 108. Sanders B, Buoncristiani R. Diagnostic and surgical arthroscopy of the temporomandibular joint: clinical experience with 137 procedures over a two year period. J Craniomandib Dis 1987;1:202. 109. Fourrier F, Duvivier B, Boutigny H, et al. Colonization of dental plaque: a source of nosocomial infections in intensive care unit patients. Crit Care Med 1998;26:301–308. 110. Sole ML, Poalillo FE, Byers JF, et al. Bacterial growth in secretions and on suctioning equipment of orally intubated patients: a pilot study. Am J Crit Care 2002;11:141–149. 111. Imsand M, Janssens JP, Auckenthaler R, et al. Bronchopneumonia and oral health in hospitalized older patients. A pilot study. Gerodontology 2002;19:66–72. 112. DeRiso AJ, Ladowski JS, Dillon TA, et al. Chlorhexidine gluconate 0.12% oral rinse reduces the incidence of total nosocomial respiratory infection and nonprophylactic systemic antibiotic use in patients undergoing heart surgery. Chest 1996;109:1556–1561. 113. Fourrier F, Cau-Pottier E, Boutigny H, et al. Effects of dental plaque antiseptic decontamination on bacterial colonization and nosocomial infections in critically ill patients. Intensive Care Med 2000;26:1239–1247. 114. Houston S, Hougland P, Anderson JJ, et al. Effectiveness of 0.12% chlorhexidine gluconate oral rinse in reducing prevalence of nosocomial pneumonia in patients undergoing heart surgery. Am J Crit Care 2002;11:567–570. 115. Dent CD, Olson JW, Farish SE, et al. The influence of preoperative antibiotics on success of endosseous implants up to and including stage II surgery: a study of 2,641 implants. J Oral Maxillofac Surg 1997;55:19–24. 116. Van de Perre JPA, Stoelinga PJW, Blijdorp PA, et al. Perioperative morbidity in maxillofacial orthopaedic surgery: a retrospective study. J Craniomaxillofac Surg 1996;24:263–270. 117. Okabe K, Nakagawa K, Yamamoto E. Factors affecting the occurrence of bacteremia associated with tooth extraction. Int J Oral Maxillofac Surg 1995;24:239–242. 118. Daly C, Mitchell D, Grossberg D, et al. Bacteraemia caused by periodontal probing. Aust Dent J 1997;42:77–80. 119. Daly CG, Mitchell DH, Highfield JE, et al. Bacteremia due to periodontal probing: a clinical and microbiological investigation. J Periodontol 2001;72:210–214. 120. Flood TR, Samaranayake LP, McFarlane TW, et al. Bacteraemia following incision and drainage of dentoalveolar abscesses. Br Dent J 1990;169:51–53. 121. Dens F, Boogaerts M, Boute P, et al. Caries-related salivary microorganisms and salivary flow rate in bone marrow recipients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:38–43. 122. Bergmann OJ. Oral infections and septicemia in immunocompromised patients with hematologic malignancies. J Clin Microbiol 1988;26:2105–2109. 123. Dreizen S, McCredie KB, Bodey GP, et al. Quantitative analysis of the oral complications of antileukemia therapy. J Oral Surg Oral Med Oral Pathol 1986;62:650–653. 124. Kennedy HF, Morrison D, Kaufmann ME, et al. Origins of Staphylococcus epidermidis and Streptococcus oralis causing bacteraemia in a bone marrow transplant patient. J Med Microbiol 2000;49:367–370. 125. Ferretti GA, Ash RC, Brown AT, et al. Chlorhexidine for prophylaxis against oral infections and associated complications in patients receiving bone marrow transplant. J Am Dent Assoc 1987;114:461–467. 126. Barker GJ, Call SK, Gamis AS. Oral care with vancomycin paste for reduction in incidence of О-hemolytic streptococcal sepsis. J Pediatr Hematol Oncol 1995;17:151–155. 127. Jansma J, Vissink A, Spijkervet FK, et al. Protocol for the prevention and treatment of oral sequelae resulting from head and neck radiation therapy. Cancer 1992;70:2171–2180. 128. Solomon CS, Shaikh AB, Arendorf TM. An efficacious oral health care protocol for immunocompromised patients. Special Care Dentistry 1995;15:228–233. 129. Meurman JH, Pyrhonen S, Teerenhovi L, et al. Oral sources of septicaemia in patients with malignancies. Oral Oncol 1997;33:389–397. 130. Anonymous. Periodontal considerations in the management of the cancer patient. Committee on Research, Science and Therapy of the American Academy of Periodontology. J Periodontol 1997;68:791–801. 131. Marten MV, van Saene HKF. The role of oral micro-organisms in cancer therapy. Oral Maxillofac Surg Infect 1992;11:81–84. 132. Akintoye SO, Brennan MT, Graber CJ, et al. A retrospective investigation of advanced periodontal disease as a risk factor for septicemia in hematopoietic stem cell and bone marrow transplant recipients. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002;94:581–588. 133. Robinson PG, Adegboye A, Rowland RW, et al. Periodontal diseases and HIV infection. Oral Dis 2002;8(Suppl 2):144–150. 134. Robinson PG. The significance and management of periodontal lesions in HIV infection. Oral Dis 2002;8(Suppl 2):91–97. 135. Tappuni AR, Fleming GJ. The effect of antiretroviral therapy on the prevalence of oral manifestations in HIV-infected patients: a UK study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;92:623–628. 136. Hastreiter RJ, Jiang P. Do regular dental visits affect the oral health care provided to people with HIV? J Am Dent Assoc 2002;133:1343–1350. 137. Patton LL, Shugars DA, Bonito AJ. A systematic review of complication risks for HIV-positive patients undergoing invasive dental procedures. J Am Dent Assoc 2002;133:195–203. 138. Farah CS, Harmon DM. Tongue piercing: case report and review of current practice. Aust Dent J 1998;43:387–389. 139. Stroud R. Tongue piercing. Br Dent J 2002;193:3. 140. Campbell A, Moore A, Williams E, et al. Tongue piercing: impact of time and barbell stem length on lingual gingival recession and tooth chipping. J Periodontol 2002;73:289–297. 141. Hardee PS, Mallya LR, Hutchison IL. Tongue piercing resulting in hypotensive collapse. Br Dent J 2000;188:657–658. 142. Tronel H, Chaudemanche H, Pechier N, et al. Endocarditis due to Neisseria mucosa after tongue piercing. Clin Microbiol Infect 2001;7:275–276. 143. Akhondi H, Rahimi AR. Haemophilus aphrophilus endocarditis after tongue piercing. Emerg Infect Dis 2002;8:850–851. 144. Harding PR, Yerkey MW, Deye G, et al. Methicillin resistant Staphylococcus aureus (MRSA) endocarditis secondary to tongue piercing. J Miss State Med Assoc 2002;43:109. 145. Martinello RA, Cooney EL. Cerebellar brain abscess associated with tongue piercing. Clin Infect Dis 2003;36:e32–34. 146. OГDwyer JJ, Holmes A. Gingival recession due to trauma caused by a lower lip stud. Br Dent J 2002;192:615–616. 147. Straus SE. Herpes simplex virus infections: biology, treatment and prevention. Ann Intern Med 1985;103:404. 148. Kotilainen HR, Brinker JP, Avato JL, et al. Latex and vinyl examination gloves. Quality control procedures and implications for health care workers. Arch Intern Med 1989;149:2749–2753. 149. Herbert AM, Bagg J, Walker DM, et al. Seroepidemiology of herpes virus infections among dental personnel. J Dentistry 1995;23:339–342. 150. Davies KJ, Herbert AM, Westmoreland D, et al. Seroepidemiological study of respiratory virus infections among dental surgeons. Br Dent J 1994;176:262–265. 151. Siew C, Gruninger SE, Mitchell EW, et al. Survey of hepatitis B exposure and vaccination in volunteer dentists. J Am Dent Assoc 1987;114:457–459. 152. Ahtone J, Goodman RA. Hepatitis B and dental personnel: transmission to patients and prevention issues. J Am Dent Assoc 1983;106:219–222. 153. Mast EE, Alter MJ. Prevention of hepatitis B virus infection among health-care workers. In: Ellis RW, ed. Hepatitis B vaccines in clinical practice. New York: Marcel Dekker, 1993:295–307. P.925 154. West DJ. The risk of hepatitis B infection among health professionals in the United States: a review. Am J Med Sci 1984;287:26–33. 155. Hu DI, Kane MA, Heymann DL. Transmission of HIV, hepatitis B virus and other bloodborne pathogens in healthcare settings: a review of risk factors and guidelines for prevention. Bull WHO 1991;69:623–630. 156. Gibas A, Blewett DR, Schoenfeld DA, et al. Prevalence and incidence of viral hepatitis in health workers in the prehepatitis B vaccination era. Am J Epidemiol 1992;136:603–610. 157. Burke FJT, Baggett FJ, Lomax AM. Assessment of the risk of glove puncture during oral surgery procedures. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;82:18–21. 158. Avery CME, Hjort A, Walsh S, et al. Glove perforation during surgical extraction of wisdom teeth. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;86:23–25. 159. Cancio-Bello TP, de Medina M, Shorey J, et al. An institutional outbreak of hepatitis B related to a human biting carrier. J Infect Dis 1982;146:652–656. 160. Stornello C. Transmission of hepatitis B via human bite. Lancet 1991;338:1024–1025. 161. Centers for Disease Control and Prevention. Recommended infection-control practices for dentistry. MMWR Mortal Morb Wkly Rep 1993;41:1–12. 162. Kim WR. The burden of hepatitis C in the United States. Hepatology 2002;36(5 Suppl 1):S30–34. 163. Marwick C. Hepatitis C is focus of NIH consensus panel. JAMA 1997;277:1268–1269. 164. Hsu HH, Wright TL, Luba D, et al. Failure to detect hepatitis C virus genome in human secretions with the polymerase chain reaction. Hepatology 1991;14:763–767. 165. Couzigous P, Richard L, Dumas F, et al. Detection of HCV-RNA in saliva of patients with chronic hepatitis C. Gut 1993;34:559–560. 166. Wang JT, Wang TH, Lin JT, et al. Hepatitis C virus RNA in saliva of patients with post-transfusion hepatitis C infection. Lancet 1991;337:48. 167. Chen M, Yun ZB, Sailberg M, et al. Detection of hepatitis C virus RNA in the cell fraction of saliva before and after oral surgery. J Med Virol 1995;45:223–226. 168. Dusheiko GM, Smith M, Scheuer PJ. Hepatitis C virus transmitted by human bite. Lancet 1990;336:503–504. 169. Figueiredo JF, Borges AS, Martinez R, et al. Transmission of hepatitis C virus but not human immunodeficiency virus type I by a human bite. Clin Infect Dis 1994;19:546–547. 170. Siew C, Gruninger SE, Miaw CL, et al. Percutaneous injuries in practicing dentists: a prospective study using a 20-day diary. J Am Dent Assoc 1995;126:1227–1234. 171. Thomas DL, Gruninger SE, Siew C, et al. Occupational risk of hepatitis C infections among general dentists and oral surgeons in North America. Am J Med 1996;100:41–45. 172. Cleveland JL, Gooch BF, Lockwood SA. Occupational blood exposures in dentistry: a decade in review. Infect Control Hosp Epidemiol 1997;18:717–721. 173. Prati D, Capelli C, Silvani C, et al. The incidence and risk factors of community-acquired hepatitis C in a cohort of Italian blood donors. Hepatology 1997;25:702–704. 174. Mele A, Sagliocca L, Manzillo G, et al. Risk factors for acute non-A, non-B hepatitis and their relationship to antibodies for hepatitis C virus: a case-control study. Am J Public Health 1994;84:1640–1643. 175. Bell DM. Human immunodeficiency virus transmission in healthcare settings: risk and risk reduction. Am J Med 1991;91(Suppl 3B):294–300. 176. Do AN, Ciesielski CA, Metler RP, et al. Occupationally acquired human immunodeficiency virus (HIV) infection: national case surveillance data during 20 years of the HIV epidemic in the United States. Infect Control Hosp Epidemiol 2003;24:86–96. 177. Gooch BF, Siew C, Cleveland JL, et al. Occupational blood exposure and HIV infection among oral and maxillofacial surgeons. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:128–134. 178. Centers for Disease Control and Prevention. Possible transmission of human immunodeficiency virus to a patient during an invasive dental procedure. MMWR Mortal Morb Wkly Rep 1990;39:489–493. 179. Gooch B, Marianos D, Ciesielski C, et al. Lack of evidence for patient-to-patient transmission of HIV in a dental practice. J Am Dent Assoc 1993;124:38–44. 180. Lewis DL, Boe RK. Cross infection risks associated with current procedures for using high-speed dental handpieces. J Clin Microbiol 1992;30:401–406. 181. Lewis DL, Arens M, Appleton SS, et al. Cross-contamination potential with dental equipment. Lancet 1992;340:1252–1254. 182. Rutala WA. APIC guideline for selection and use of disinfectants. Am J Infect Control 1996;24:313–342. 183. Fayle SA, Pollard MA. Decontamination of dental unit water systems: a review of current recommendations. Br Dent J 1996;181:369–372. 184. Jensen ET, Giwercman B, Ojeniyi B, et al. Epidemiology of Pseudomonas aeruginosa in cystic fibrosis and the possible role of contamination by dental equipment. J Hosp Infect 1997;36:117–122. 185. Matsuyama M, Usami T, Masuda K, et al. Prevention of infection in dental procedures. J Hosp Infect 1997;35:17–25. 186. WHO infection control guidelines for transmissible spongiform encephalopathies. Report of a WHO consultation, Geneva, Switzerland, 23–26 March 1999. Available at http://www.who.int/emc-documents/tse/whocdscsraph2003c.html (accessed 12/29/2003). 187. Converse JM, McCarthy JG. Infection in plastic surgery. Surg Clin North Am 1972;52:1459. 188. Anonymous. Advisory statement. Antibiotic prophylaxis for dental patients with total joint replacements. American Dental Association: American Academy of Orthopaedic Surgeons. J Am Dent Assoc 1997;128:1004–1008. 189. Younai FS. Postexposure protocol. Dental Clin North Am 1996;40:457–486. 190. Cleveland JL, Barker L, Gooch BF, et al. Use of HIV postexposure prophylaxis by dental health care personnel: an overview and updated recommendations. J Am Dent Assoc 2002;133:1619–1626. 191. Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care facilities, 1994. In: Abrutyn E, Goldman DA, Scheckler WE, eds. Saunders infection control reference service. Philadelphia: WB Saunders, 1998;1041–1042. 192. McCarthy GM, MacDonald JK. Improved compliance with recommended infection control practices in the dental office between 1994 and 1995. Am J Infect Control 1998;26:24–28. 193. Gershon RRM, Karkashian C, Vlahov D, et al. Correlates of infection control practices in dentistry. Am J Infect Control 1998;26:29–34. 194. McCarthy GM, Koval JJ, MacDonald JK. Compliance with recommended infection control procedures among Canadian dentists: results of a national survey. Am J Infect Control 1999;27:377–384. 195. McCarthy GM, MacDonald JK. A comparison of infection control practices of different groups of oral specialists and general dental practitioners. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:47–54. 196. DePaola LG. Managing the care of patients infected with bloodborne diseases. J Am Dent Assoc 2003;134:350–358. 197. Williams HN, Singh R, Romberg E. Surface decontamination in the dental operatory: a comparison over two decades. J Am Dent Assoc 2003;134:325–330.
Воскресенье, 06 Ноябрь 2011 | 1934 Просмотры | Печать | PDF
Advances in medical and surgical interventions, world travel, global climate changes, and threats of bioterrorism are rapidly changing healthcare and pose a challenge in controlling occupationally acquired infections (1, 2, 3, 4). Historically, workers in diagnostic laboratories have always been at higher risk for infection from exposure to infectious materials (5). Today, the laboratory worker is faced with increased exposure to infectious material from the recognition of new infectious agents, potential use of bioterrorism agents, increasing antimicrobial resistance, and introduction of new diagnostic techniques and instrumentation. In addition, improper handling of biologic wastes or episodes of laboratory-acquired infection (LAI) could lead to the spread of microorganisms outside the laboratory, although this occurrence has been rare. Thus, implementation and adherence to effective prevention and control measures should be important to all who work in the laboratory environment (6). Interestingly, laboratory workers must be constantly reminded of these hazards, because they often minimize the risk either through constant daily exposure or risk-taking behavior (7). This chapter considers microorganisms likely to cause infections in hospital laboratory workers, usual modes by which these microorganisms are transmitted in the laboratory, and appropriate control measures for preventing such incidents. OCCURRENCE OF LABORATORY-ACQUIRED INFECTION The risk of LAI has been recognized since the end of the nineteenth century but the true incidence is unknown, even today, because of the lack of adequate reporting systems. In the world literature, LAIs are usually reported as individual cases or compiled through laboratory surveys. Often, accurate documentation of the LAI is lacking in these reports. The most extensive surveys in the United States were conducted by Sulkin and Pike from 1949 to 1970 (8, 9, 10). More recent surveys reviewed LAIs in Utah (11) and in public health laboratories (12). Based on these limited data, Wilson and Reller (13) estimated that the annual incidence of LAIs in the United States was between 1 and 5 per 1,000 employees. More systematic surveys of LAIs have occurred in the United Kingdom from 1970 to 1995 (14, 15, 16). The most recent retrospective survey for occupationally acquired infections in 397 laboratories (1994 to 1995) in the United Kingdom found an overall incidence rate of 16.2 per 100,000 person-years compared with 82.7 infections per 100,000 person-years for a similar survey conducted in 1988 to 1989, suggesting that control measures may be reducing the incidence of infections (16). Because of the lack of adequate modern data on LAIs, control measures are proposed and implemented on the basis of old data (11), experience with one infectious agent applied to others, the epidemiology of relevant microorganisms from nonlaboratory settings, and hazard analysis (17). Although laboratory workers will always be at risk for infection, adherence to safety measures will significantly reduce the risk. The previous data document the occupational risk associated with handling patient specimens and microbiologic cultures. By contrast, few reports document the spread of laboratory pathogens from the laboratory to other hospital areas or to the community (18). Thus, the risk of infection to laboratory workers is greatest from specimens originating in the community or hospitalized patients, and the danger of diagnostic laboratory microorganisms affecting the community is small. Mode of Transmission and Etiology In laboratories, the factors that influence occupationally acquired infections are related to host susceptibility and behavior, the virulence and availability of the pathogen, and the work environment (5). The most common types of exposure that cause infections include inhalation of aerosols generated by accidents and work practices; percutaneous inoculation through accidents with needles, blades, and broken glassware; ingestion; and contamination of mucous membranes and skin (5, 19). Often the specific exposure incident is not easily identifiable other than working with infectious material in a diagnostic laboratory environment. In the past, Brucella species, Mycobacterium tuberculosis, Coxiella burnetii, hepatitis B virus (HBV), Francisella tularensis, and Salmonella species caused most of the LAIs (9). During the 1980s, laboratory workers were most frequently infected with M. tuberculosis, Salmonella species, Shigella species, HBV, and hepatitis C virus (HCV) (11, 12, 14, 15). In the 1990s, biosafety measures have emphasized the reduction of infection from blood-borne pathogens in all healthcare workers P.1432 (HCWs). The risk of acquiring a blood-borne infection is influenced by the prevalence of infection in patients, the amount of blood involved, the type of exposure, the concentration of pathogen in the blood or body fluid, and the availability of postexposure prophylaxis (6, 20). In addition to infections from human immunodeficiency virus (HIV), HBV, and HCV, blood-borne transmission of at least 20 different agents has been reported (21). However, non–blood-borne infections from enteric pathogens, Brucella species, and Neisseria meningitidis continue to cause LAIs (17, 22, 23, 24). A list of selected microorganisms that may cause laboratory infections is illustrated in Table 82.1. A more complete compilation can be found in selected publications (6, 17, 25, 26, 27). The following is a categorization of these microorganisms by their likely mode of spread within the laboratory environment. TABLE 82.1. SELECTED MICROORGANISMS INVOLVED IN LABORATORY INFECTION EPISODES REPORTED IN MEDICAL JOURNALS DURING THE PERIOD 1994–2002 BY MICROORGANISM GROUP Blood-borne Viruses The blood-borne viruses (HIV, HBV, HCV) pose the infection risk of greatest concern to hospital workers (6, 20, 28). As of June 30, 2001, the Centers for Disease Control and Prevention (CDC) had received reports of 57 HCWs in the United States who are documented as having seroconverted to HIV and 137 other reports classified as possible occupational transmission (http://www.cdc.gov/hiv/pubs/facts/hcwsurv.htm). These individuals include 19 laboratory workers (16 of who were clinical laboratory workers). Forty-eight of the 57 documented exposures were percutaneous exposures, 5 were mucocutaneous, 2 were both, and 2 were an unknown route of exposure. Forty-nine HCWs were exposed to HIV-infected blood, 3 to concentrated virus, 1 to visibly bloody fluid, and 4 to unspecified fluid. Twenty-six of these individuals developed acquired immunodeficiency syndrome (AIDS). It is estimated that the individual HCW experiences approximately 30 needlestick injuries per 100 beds per year (29) and that laboratory workers experience more mucocutaneous exposures (30). The risk of infection from HIV, HBV, and HCV following occupational exposure to infected blood is related to the concentration of the virus in blood. HBV can be present in concentrations of 108 to 109 infectious particles/mL blood, and the concentrations of HIV and HCV are 100 to 104 and 102 to 103 particles/mL blood, respectively (6, 20). The risk of infection following a percutaneous exposure is approximately 18% (6% to 30%) for HBV, 1.8% (0% to 7%) for HCV, and 0.3% (0% to 0.9%) for HIV. Following the mandatory requirement that employers provide HBV vaccination at no cost to their employees, the incidence of HBV infections in HCWs decreased 95% from 1983 to 1995 (31). The prevalence of HCV infection among HCWs (1% to 2%) appears no greater than the rate observed in the general population (6, 20). Although the blood-borne viruses are found in many different body fluids and tissues, the transmission of HCV, HIV, and HBV is most often associated with blood or visibly bloody body fluids. Airborne Mycobacteria and Bacteria The transmission of M. tuberculosis and Mycobacterium bovis in healthcare facilities and clinical laboratories is a recognized risk (32). Since 1953 the tuberculosis case rate has declined tenfold from 53 cases per 100,000 to 5.6 per 100,000 in 2001 and decreased 40% from 1992 when the case rate most recently peaked in the United States (http://www.cdc.gov/nchstp/tb/surv/surv2001/default.htm). The presence of tubercle bacilli in specimens other than respiratory secretions (e.g., gastric aspirates, cerebrospinal fluid (CSF), urine, exudates, and tissue) may result in nosocomial transmission to HCWs and autopsy personnel (33). In addition, infection may result from direct parenteral inoculation by laboratory workers (34). However, the greatest risk to laboratory personnel is from exposure to aerosols generated during handling of liquid specimens, preparation of frozen sections, and performing autopsies (6). Other bacteria that may be transmitted by airborne droplets or aerosols include Corynebacterium diphtheriae; N. meningitidis; Bordetella pertussis; Streptococcus pyogenes; and the potential agents of bioterrorism, Bacillus P.1433 anthracis, Yersinia pestis, Brucella species, F. tularensis, and Burkholderia pseudomallei (6, 25, 35, 36, 37, 38). Brucellosis is a commonly reported LAI in research and animal laboratories but also occurs in clinical laboratories (22, 24). The agent of whooping cough, B. pertussis, has caused at least 12 LAIs in the past 20 years (5). N. meningitidis is an infrequent cause of LAIs but has been associated with fatal outcomes and should be handled in a manner that minimizes risk for exposure to aerosols or droplets (23). Worldwide, 16 probable cases of LAIs have occurred from 1986 to the present with six probable cases occurring in the United States from 1996 to 2001. The source isolates were recovered from blood or CSF in five of the six cases and probably CSF or middle ear fluid in the sixth case. In 15 of the 16 worldwide cases, common laboratory procedures were not performed in a biologic safety cabinet and may have contributed to the exposure incidents. Airborne Fungi Laboratory-acquired fungal infections have been reported infrequently after 1980. Generally, fungal infections are acquired from the inhalation of the conidia of Coccidioides immitis, Histoplasma capsulatum, or Blastomyces dermatitidis and in one report Penicillium marneffei (5, 17, 19, 39). Occasionally, cutaneous infections occur following accidental inoculation (40, 41). Coccidioidomycosis and histoplasmosis are the most likely fungal infections to be transmitted in the laboratory (17). Arthroconidia from laboratory cultures of C. immitis easily become airborne, whereas spherules from tissue are much less likely to be aerosolized. Laboratory-acquired histoplasmosis also results primarily from handling laboratory cultures. The infective conidia are small and likely to become airborne, resist drying, and can cause infection after small inocula are inhaled. Pulmonary infection resulting from B. dermatitidis has followed inhalation of the conidia by laboratory workers, but this is much less frequent than cases of coccidioidomycosis or histoplasmosis. Airborne Viruses, Chlamydia, and Rickettsiae Most LAIs from C. burnetii arise from aerosols generated in animal research laboratories although there are a few reports of parenteral and mucous membrane transmissions (5, 19, 25). LAIs from Rickettsia typhi, Rickettsia coronii, andOrientia tsutsugamushi have also been reported (5, 42). Before 1960, psittacosis was “among the most commonly reported laboratory-associated infections” (25), but only sporadic cases have been reported in the past 20 years (5). Psittacosis case-fatality rates are high compared with those of infections resulting from other agents. The microorganism Chlamydia psittaci may be present in tissues, feces, nasal secretions, and blood specimens. Few infections occur from exposure to Chlamydia trachomatis and generally result from mucous membrane exposure. Respiratory viral infections acquired in the laboratory are probably underreported, because it is difficult to document occupational acquisition. These viruses can be aerosolized by manipulation of specimens or cultures. Most laboratory-acquired viral infections, other than infections from the blood-borne viruses, occur in animal research laboratories following exposure to aerosols or contamination of skin and mucous membranes (5). Contact-acquired Enteric Bacteria, Viruses, and Parasites Infections from the enteric bacterial pathogens, Salmonella species and Shigella species, are commonly reported LAIs and are probably underreported (5, 11, 16, 25, 43). Infections generally occur from handling laboratory specimens and microbiologic cultures or occasionally from ingestion of intentionally contaminated food (44). Salmonella typhi causes the most serious infection (45). Gastroenteritis resulting from Vibrio species, Campylobacter species, enterotoxigenic Escherichia coli, and hepatitis A (HAV) and hepatitis E (HEV) viruses is infrequently reported (5, 46, 47, 48, 49). The shedding of HAV and probably HEV is diminished by the time a patient is symptomatic, decreasing the risk of transmission in the healthcare facility (4). Parasitic diseases are receiving increasing attention because of world travel and increased susceptibility in immunocompromised individuals (50). Laboratory-acquired malaria, leishmaniasis, trypanosomiasis, and toxoplasmosis infections have all been reported. The two most frequently reported infections from accidental exposure are from Trypanosoma cruzi and Toxoplasma gondii. The rate of occurrence of laboratory accidents during work with T. gondii is reported to be one accident per 9,300 hours of exposure (51) or 1 infection per 24 person-years (50). The infection rate for working with T. cruzi is calculated to be 1 infection per 46 person-years. These infections ranged from asymptomatic to fatal in one case for each microorganism. Herwaldt (50) also reported on Plasmodium species (34 cases) and Leishmania species (12 cases) infections. Most of the infections associated with blood and tissue protozoa occurred from parenteral exposure but other routes included skin and mucous membrane exposure and ingestion. Only 21 cases of LAIs with intestinal protozoans have been reported and involved Cryptosporidium parvum, Isospora belli, and Giardia lamblia. Fewer reports have involved the helminths, including Schistosoma species, Strongyloides species, and Ancylostoma species. The most probable route of infection was ingestion of contaminated material although a few cases were associated with aerosols or skin penetration. Contact-associated Bacteria and Fungi S. pyogenes, Staphylococcus aureus, Neisseria gonorrhoeae, and N. meningitidis all have caused laboratory infection in association with parenteral inoculation or droplet exposure of mucous membranes from laboratory cultures of the microorganism or from clinical specimens. A few LAIs resulting from Cryptococcus neoformans, B. dermatitidis, and Sporothrix schenckii have occurred in which the proposed mode of spread was direct contact, splash, or percutaneous inoculation. Bacteria, Fungi, Spirochetes, Viruses, and Rickettsia with Multiple Modes of Transmission Many microorganisms, including the potential agents of bioterrorism, are transmitted by multiple exposure routes such as P.1434 aerosols, contamination of skin and mucous membranes, ingestion, and percutaneous inoculation. Brucellosis is highly infectious and often causes multiple infections in research or laboratory workers following an accident (22, 24). Infections often occur when laboratory workers do not recognize the pathogen and neglect to take necessary safety precautions (52). The practice of “sniffing” plates for characteristic odors associated with a specific bacterium should be curtailed (5). In addition to aerosol transmission, laboratory-acquired brucellosis has occurred from direct skin contact with cultures or with other infectious material, percutaneous inoculation, and spray onto mucous membranes. These same transmission routes are important for B. anthracis, the agent of anthrax; F. tularensis, the cause of tularemia; C. diphtheriae, the agent of diphtheria, and Y. pestis, the agent of plague. All of these bacteria should be handled with biosafety level (BSL) 2 and 3 safety precautions (37). B. pseudomallei, the microorganism responsible for melioidosis, is cited as a rare cause of LAIs but has been associated with a fatal outcome (53). Direct contact with microbiologic cultures or specimens, ingestion, autoinoculation, and exposure to infectious aerosols and droplets all have been implicated in transmission of B. pseudomallei. Fortunately, the agent of Q fever, C. burnetii, is rare in the United States, so the risk for diagnostic laboratory-acquired Q-fever infection in this country is minor compared with that in many other parts of the world. The microorganism is present in blood, urine, feces, milk, and tissue specimens and resists drying. Airborne spread is the most likely route for laboratory transmission, but parenteral inoculation occurs as well. An extremely small inoculum can produce disease. Leptospira interrogans, the cause of leptospirosis, can be present in urine, blood, and tissues of infected patients. Ingestion, accidental parenteral inoculation, and contact of skin or mucous membranes with cultures or infected specimens all have led to laboratory worker infection. Likewise, syphilis has been an LAI and its agent, Treponema pallidum, can be present not only in blood but also in cutaneous, mucous membrane, and other lesions. Laboratory spread of this microorganism follows from parenteral inoculation, contact of mucous membranes or broken skin with infectious clinical materials, and possibly infectious aerosols. Most laboratory-acquired viral infections occur in animal research facilities and include numerous agents (5). Arenavirus (54), Sabia virus (55), West Nile virus (56), and other viruses causing hemorrhagic disease have caused laboratory infections. Lymphocytic choriomeningitis virus infections in laboratory workers occur in diagnostic facilities when cell cultures become contaminated with the virus, leading to possible aerosolization or skin or mucous membrane contamination. Specimens suspected of harboring the agent of smallpox, variola major, should not be cultured but rather shipped directly to CDC or a state health laboratory (37, 38). Accidental parenteral inoculations are likely sources for laboratory-acquired rickettsial infections, but several infections with typhus have been associated with aerosols or infected airborne particles, and cases of Rocky Mountain spotted fever probably have occurred by this route as well (19, 42). Because most diagnostic clinical laboratories do not perform cultures for rickettsia, these infections are more likely to be a risk in research laboratories. RESERVOIRS AND MODES OF SPREAD As detailed previously, a variety of modes of transmission have been noted in cases of LAIs and include inhalation, ingestion, inoculation, and contamination of skin and mucous membranes (6, 17, 19, 25). Perhaps the most likely mode of transmission is due to accidental inoculation of skin or soft tissue with needles or other sharps such as scalpels and broken glass from specimen containers. Nearly all pathogenic microorganisms can produce infection by this route, which is the most frequent route of transmission for blood-borne pathogens such as HIV or the hepatitis viruses (6). Hopefully, the accidental percutaneous inoculation of infectious material by laboratory personnel will decrease with the increased use of plastic collection tubes, needleless systems, and engineered safety devices (6, 57, 58). Needles should not be used in the laboratory unless there is no other alternative. Although the intact skin is an excellent barrier to penetration by microorganisms, it does contain minor cuts and abrasions that serve as portals of entry. Contamination of mucous membranes by splashes and sprays of infectious material can lead to the laboratory transmission of HIV and other pathogenic agents to laboratory workers (5, 6, 25, 27). In animal research facilities, bites and scratches from infected animals present a risk for transmission of an agent. As on patient care wards, transmission by hand to skin and mucous membrane of the mouth, eye, and nose can cause an infection (59). Ingestion may occur following mouth pipetting, transfer of microorganisms on contaminated fingers or pencils, accidental splashes, or consumption of food and beverages in the laboratory. The laboratory environment is contaminated during the workday from routine specimen processing and the other work practices that produce aerosols or splatters and results in the contamination of hands (25, 27, 60). This stresses the importance of avoiding poor personal hygiene practices, such as applying cosmetics and adjusting contact lenses in the laboratory. Cases associated with contamination of food, drink, or tobacco products also have declined, because attention has been paid to eliminating eating, drinking, and smoking in the laboratory. Indirect contact with microorganisms can occur when environmental objects (e.g., specimen containers, test requisitions, instruments) or surfaces become contaminated with microorganisms. Accidents or spills also can lead to contamination of the workbench or other equipment, which may lead to contamination by hand contact. Airborne spread is one mode of transmission of great concern in the laboratory (6,37, 61). Many laboratory procedures generate aerosols, droplets, or droplet nuclei that can be associated with direct transmission of infection through inhalation by the laboratory worker. Droplet nuclei (<5 Вm in diameter) tend to remain suspended in air and move throughout the room or to build on air currents and reach the alveoli of the lungs when inhaled (62, 63). Relevant procedures that generate aerosols include use of bacteriology loops for transferring cultures and flaming them afterward; pipetting (especially with fixed automatic pipettes); using syringes and needles; opening tubes and bottles; using centrifuges and blenders; performing autopsies; harvesting viral cultures; lyophilizing; and breaking culture plates, bottles, P.1435 and tubes. These work practices also produce droplets that contaminate counters or floor surfaces, permitting transmission from these surfaces to hands. Microorganisms in blood droplets can survive for several days after drying on work surfaces or instruments (64). GUIDELINES FOR PREVENTION Laboratory safety is demanded by the standards of the Occupational Safety and Health Administration (OSHA), which are driven by the premise that the employer must provide a safe workplace (20, 28, 65, 66, 67, 68). Compliance with current OSHA standards is subject to assessment by the agency's inspectors; thus, these regulations are perhaps of greatest importance to clinical diagnostic laboratories. Other groups, such as the CDC (69, 70), the National Institutes of Health (25), the College of American Pathologists (CAP), and the National Coordinating Committee for Laboratory Standards (6) all provide guidelines or regulations for laboratory safety. OSHA, the National Institute for Occupational Safety and Health (NIOSH), the CAP, and the Joint Commission on Accreditation of Healthcare Organizations include safety among their checklists for laboratory inspectors. State and local licensing inspections and federal inspections for participation in Medicare also focus on safety issues. Guidelines for laboratory safety from these groups cover exposures to chemical agents, fire, and other aspects, but the highlight of each is the prevention of laboratory infection. The following discussion is guided by these various regulations and guidelines and centers on the clinical diagnostic laboratory. The prevention of infection in autopsy, surgical pathology, and research and referral laboratories follows the same general plan considered here, but its implementation varies dramatically in each site according to the work done and the microorganisms involved (6). Each laboratory must assess its specific risk from handling infectious material and design an exposure control plan to minimize these potential risks. Safety practices, usually containment measures, are designed to reduce or eliminate the exposure of laboratory workers to infectious material (6, 25). These practices vary with the pathogenicity and infectious dose of the agent, the routes of transmission, the work performed, and the availability of treatment or prophylaxis (25, 71, 72). The CDC/National Institutes of Health guidelines (25) recommend four levels of biosafety (Table 82.2), and each successive level suggests increased occupational risk and more stringent containment practices. These classifications are similar to those adopted by the World Health Organization (WHO) based on increasing level of risk to the individual and community and availability of effective treatment and prevention (73). The clinical diagnostic laboratory typically encounters microorganisms as shown in Table 82.2; thus, further discussion focuses on elements in BSLs 2 and 3. Laboratories that use, receive, or store select agents must address, in addition to BSL 2 to 4 safety practices, security and reporting issues (74). Additional safety practices are necessary for work in research and anatomic laboratories (6, 75, 76, 77). TABLE 82.2. BIOSAFETY LEVEL OF MICROORGANISMS ENCOUNTERED IN THE LABORATORY AND SELECTED MICROORGANISMS REPRESENTATIVE OF THE LEVEL OSHA regulations for prevention of infection emphasize engineering controls, work practices modification, and personal P.1436 protection by immunization and protective equipment (67, 68). Guidelines for the clinical diagnostic laboratory can be placed in these same general categories and compared for BSLs 2 and 3 (Tables 82.3, 82.4 and 82.5). Most guidelines are common to both BSLs 2 and 3 (Table 82.3), whereas some are unique to level 2 (Table 82.4), and others are specific to level 3 protection (Table 82.5). Many of these elements are pertinent to other hospital areas and to laboratories; such policies are discussed in detail in other chapters and are reviewed only briefly here. Several elements of infection prevention are more relevant to the laboratory than to other areas, and these are discussed at greater length in the following sections. TABLE 82.3. CONTROL MEASURES FOR PREVENTION OF LABORATORY-ASSOCIATED NOSOCOMIAL INFECTIONS THAT ARE COMMON TO BIOSAFETY LEVELS 2 AND 3a TABLE 82.4. REQUIREMENTS FOR BIOSAFETY LEVEL 2 THAT DIFFER FROM THOSE FOR BIOSAFETY LEVEL 3a TABLE 82.5. REQUIREMENTS FOR BIOSAFETY LEVEL 3 THAT DIFFER FROM THOSE FOR BIOSAFETY LEVEL 2a P.1437 Engineering Controls Airflow handling is an essential element in several clinical care areas of a hospital where microorganisms likely to be spread by airborne transmission are encountered, especially M. tuberculosis and certain fungi (68). In the laboratory, however, the potential for encountering BSL 3 microorganisms that can be spread by air is so much greater that certain standards and guidelines beyond those for the rest of the institution are mandatory (61, 78). Aerosolization can result from use of blenders, both low- and high-speed centrifuges, and automatic pipettes. Loops used for inoculation of microbiologic cultures can lead to aerosolization if not flamed properly. Other standard and seemingly innocuous laboratory procedures such as pipetting, accidentally dropping infected liquids on a counter, and inoculating a tube with a syringe all can generate aerosols. If one adds to this the presence in clinical specimens of microorganisms prone to spread by the airborne route (e.g., M. tuberculosis, H. capsulatum, C. immitis, and certain viruses), the need for control of aerosols becomes crucial. Thus, building design that ensures inward directional airflow into the laboratory from corridors and hallways and similar engineering for direct exhaust of the air without recirculation are crucial for laboratories handling airborne pathogens. For BSL 3 laboratories, airflow is monitored to ensure that the ventilation system does not fail (61, 69). Air ventilation in the autopsy room is also critical and the room should be under negative pressure, provide 12 air exchanges per hour, and be exhausted directly to the outside (6). Biologic safety cabinets (BSCs) are designed to contain the highly infectious agents that are transmitted by an airborne route through infectious splashes or aerosols generated by microbiologic procedures (6, 61, 79). There are three types of BSCs (Class I, II, and III), but most routine clinical laboratories use Type II BSCs that provide protection to the user and prevent external contamination of the materials inside the cabinet. An effective containment system for handling BSL 2 and 3 agents requires that the BSC is properly maintained, that the BSC be certified annually or whenever the cabinet is moved, and that well-trained employees use good microbiologic technique (6). The characteristics of each type of cabinet and procedures for their correct use have been reviewed and extensively described elsewhere (6, 61, 68). Other engineering controls for decreasing the risk associated with handling infectious material include safety engineered devices and instruments, sharps containers, safety containers for centrifuges, plastic containers and collection devices for specimens, mechanical pipettes and diluters, bench tops impervious to liquids, and personal protection equipment. These controls are discussed in depth in other publications (6, 26). Work Practice Modification Laboratory workers cannot identify specimens that contain infectious agents and, therefore, must practice standard precautions, which is the concept that all patients and all laboratory specimens are potentially infectious and capable of transmitting infection (6, 70, 80). These guidelines represent the first level of protection for the laboratory worker from a wide variety of pathogens. Hand washing is a fundamental procedure to reduce duration of exposure and transmission of an infectious agent within the healthcare facility, including the laboratory. Adequate hand washing should occur before leaving the laboratory, after removing gloves, and after obvious hand contamination by using traditional soap and water or an alcohol-based gel (81). Some work practices promote the transfer of microorganisms from surfaces to hands to mucous membranes and are universally prohibited in the laboratory. These prohibited practices include eating or storing food, drinking, applying cosmetics or contact lens, smoking, chewing gum, and mouth pipetting. Workers with skin lesions or dermatitis on the hands or wrists should not handle potentially infectious materials without adequate protection (82). Personnel who collect and transport specimens should be adequately trained. Whether transported by hand or pneumatic tube, specimens should be placed in a leak-proof primary container. This primary container is placed in a leak-proof secondary container that is usually a sealable plastic bag. Secondary containers and specimen storage areas should be labeled with a biohazard label to alert individuals to the potential infectious hazard. P.1438 Needles should be removed before transporting a syringe to the laboratory. Specimen processing in microbiology requires special steps to prevent infection and should be performed in a BSC. For example, when entering a blood culture bottle with a needle and syringe, the vial should never be held in the worker's hand and the bottle should be placed behind a splashguard or in a BSC. Similarly, unfixed slides should always be handled as if they contained infectious materials (6). Special steps are needed for dealing with the potential hazards associated with the use of diagnostic instruments (6). Prompt decontamination of spills is particularly important in the laboratory. Most laboratory spills involve blood, other body fluids, or microbiologic media that often contain high concentrations of protein. Because many disinfectants are less active in the presence of these proteins, the bulk of the spilled P.1439 liquid must be adsorbed before disinfection (6, 82). For large spills of microbiologic cultures, the spill is flooded with an appropriate disinfectant and left to stand for 20 minutes before clean-up (6). Phenolic disinfectants are not recommended for use on contaminated medical devices that come in contact with laboratory workers but may be used on laboratory instruments, floors, and countertops. Also, instrument parts made in part or wholly of aluminum are corroded by sodium hypochlorite, so other disinfectants are preferred for disinfection of laboratory instruments containing these parts. Surveillance of accidents and exposures is a key feature of infection control in all hospital areas but is especially important in the laboratory. The essential components of postexposure management include incident reporting, wound management, evaluation of the transmission risk, and consideration of postexposure prophylaxis (6, 20, 28). The incident is reported to the supervisor, no matter how trivial the injury or exposure may be and includes the date and time of exposure, the details of the accident, information on the source person, and medical evaluation of the injured employee. The immediate reporting of the incident establishes a time relationship, in the event that an infection develops, and permits preventive measures to be implemented. OSHA regulations require that the facility's exposure control plan includes hepatitis B vaccination at no cost to the employee, postexposure evaluation and follow-up, communication of potential hazards to employees, and appropriate records and reporting (6, 20, 28, 57, 58). (See Chapters 78, 79 and 99.) Follow-up for the individual is vital; equally important is the periodic and regular analysis of the incidents that occur in a given laboratory. Laboratory, occupational health, and infection control personnel should cooperate in the compilation and analysis of incident report data to search for common patterns, to eliminate identified risk factors, and to modify laboratory procedures to minimize occurrence of these incidents (64). Procedures for medical follow-up of exposure to blood-borne pathogens are dealt with elsewhere. Surface cleaning of the laboratory bench or other surfaces must be meticulous, because these surfaces are likely to be contaminated with potential pathogens (60). Many surfaces (countertops, floors, equipment, centrifuges, etc.) become contaminated by microorganisms during routine processing of clinical specimens and cultures. These surfaces should be carefully disinfected at the completion of work and after accidental spills to prevent contamination of laboratory employees and visiting medical personnel who may unknowingly carry the agent to other parts of the facility or the community (60). All unnecessary material should be removed from these surfaces to facilitate proper cleaning and disinfection. Waste disposal and handling of biologic materials at the end of processing are especially important topics for the laboratory, because of the volume of the materials involved and because the processing of the specimens often involves amplification of the potential pathogen (83). The laboratory is a major generator of biohazardous waste and should segregate the material into designated categories such as routine, chemical, and biohazardous waste for proper decontamination and disposal. Fortunately, the same procedures used in other parts of the facility apply to laboratory waste. (See Chapter 100.) The shipment of infectious material is regulated by national and international rules and regulations promulgated by the U.S. Department of Transportation, International Airline Transport Association, and the WHO and are beyond the scope of this chapter (73, 84, 85). Personal Protection Immunization Laboratory workers must be encouraged to participate in the same immunization program that is offered throughout the institution (6, 86). This includes, at a minimum, provision of HBV immunization at no cost to the employee. The laboratory worker may be at greater risk of exposure to body fluids containing one of the hepatitis viruses, so it might be worth the special effort to emphasize immunization to laboratory employees. Immunizing trainees against HBV is particularly important, because the risk of infection often is high during training. Bacillus Calmette-GuГrin (BCG) vaccine is made from an attenuated strain of M. bovis. It is not routinely offered to hospital workers in the United States, because a positive tuberculin skin test when the vaccine is effective is thought to be a hindrance to surveillance for natural tuberculous infection and because adverse effects are associated with immunization (e.g., abscess at the injection site). However, it may be considered for laboratory employees who process large volumes of specimens containing M. tuberculosis. Other possible vaccines for laboratory workers include meningococcal polysaccharide vaccine, rabies vaccine, polio vaccine, and typhoid vaccine. Primary prevention in the laboratory should focus on biosafety practices, but these vaccines are a consideration for personnel who work with these agents on a frequent and regular basis. At this time, smallpox vaccination is recommended only for those individuals in the laboratory who directly handle cultures of the smallpox virus. Personal Protective Equipment Gloves, masks, and gowns are used throughout a hospital to protect workers from contact with blood and other potentially infectious materials. The laboratory is no exception to this practice, because all specimens handled in the laboratory are considered potentially infectious. Laboratory workers must be trained in the appropriate use, limitations, and disposal of personal protective equipment. In general only powder-free latex or other nonlatex gloves should be used in the laboratory as part of the standard precaution guidelines. Puncture-resistant gloves should be available in the autopsy suite or when handling scalpels and other sharps. In addition to protective clothing, laboratory workers should wear face shields or work behind splashguards when removing stoppers or withdrawing samples from specimen tubes (6). When extensive soaking by potentially infectious material is a possibility, waterproof coats, gowns, or aprons should be worn. Respiratory protection in the form of NIOSH-approved masks (e.g., N95 particulate respirator) is recommended when working with M. tuberculosis or other similar BSL 3 microorganisms (6, 25). Shoes should cover the feet to protect the skin from spills or dropped sharps. All personal protective equipment, P.1440 including laboratory coats, gowns, or other protective covers, should not be worn outside the laboratory area. ACKNOWLEDGMENT This chapter contains information presented in Chapter 74 by John E. McGowan Jr. in the second edition of this book. REFERENCES 1. Peterson LR, Brossette SE. Hunting health care-associated infections from the clinical microbiology laboratory: passive, active, and virtual surveillance. J Clin Microbiol 2002;40:1–4. 2. Kiska DL. Global climate change: an infectious disease perspective. Clin Microbiol Newslett 2000;22:81–86. 3. Weinstein RA. Nosocomial infection update. Emerg Infect Dis 1998;416–420. 4. Aitken C, Jeffries DJ. Nosocomial spread of viral disease. Clin Microbiol Rev 2001;14:528–546. 5. Harding AL, Byers KB. Epidemiology of laboratory-associated infections. In: Fleming DO, Hunt DL eds. Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:35–54. 6. NCCLS. Protection of laboratory workers from occupationally acquired infections; approved standard. M29-A2. Wayne, PA: NCCLS, 2001. 7. Straton CW. Occupationally acquired infections: a timely reminder. Infect Control Hosp Epidemiol 2001;22:8–9. 8. Pike RM. Laboratory-associated infections. Summary and analysis of 3921 cases. Health Lab Sci 1976;13:105–114. 9. Pike RM. Laboratory-associated infections: incidence, fatalities, causes and prevention. Annu Rev Microbiol 1979;33:41–66. 10. Sulkin SE, Pike RM. Viral infections contracted in the laboratory. N Engl J Med 1949;241:205–213. 11. Jacobson JT, Orlob RB, Clayton JL. Infections acquired in clinical laboratories in Utah. J Clin Microbiol 1985;21:486–489. 12. Vesley D, Hartmann HM. Laboratory-acquired infections and injuries in clinical laboratories: a 1986 survey. Am J Public Health 1988;78:1213–1215. 13. Wilson ML, Reller LB. Clinical laboratory-acquired infections. In: Bennett JV, Brachman PS, eds. Hospital infections, 4th ed. Philadelphia: Lippincott-Raven, 1998:343–355. 14. Grist NR, Emslie JA. Association of Clinical Pathologists' surveys of infection in British clinical laboratories, 1970–1989. J Clin Pathol 1994;47:391–394. 15. Grist NR, Emslie JA. Infections in British clinical laboratories, 1986–7. J Clin Pathol 1989;42:677–681. 16. Walker D, Campbell D. A survey of infections in United Kingdom laboratories, 1994–1995. J Clin Pathol 1999;52:415–418. 17. Sewell DL. Laboratory-associated infections and biosafety. Clin Microbiol Rev 1995;8:389–405. 18. Reuben B, Band JD, Wong P, et al. Person-to-person transmission of Brucella melitensis. Lancet 1991;337:14–15. 19. Voss A. Prevention and control of laboratory-acquired infections. In: Murray PR, Baron EJ, Pfaller MA, et al., eds. Manual of clinical microbiology, 7th ed. Washington, DC: American Society for Microbiology, 1999:165–173. 20. Beltrami EM, Williams IT, Shapiro CN, et al. Risk and management of blood-borne infections in health care workers. Clin Microbiol Rev 2000;13:385–407. 21. Collins CH, Kennedy DA. Microbiological hazards of occupational needlestick and ‘sharps’ injuries. J Appl Bacteriol 1987;62:385–402. 22. Fiori PL, Mastrandrea S, Rappelli P, et al. Brucella abortus infection acquired in microbiology laboratories. J Clin Microbiol 2000;38:2005–2006. 23. Centers for Disease Control and Prevention. Laboratory-acquired meningococcal disease-United States, 2000. MMWR Morb Mortal Wkly Rep 2000;51:141–144. 24. Memish ZA, Mah MW. Brucellosis in laboratory workers at a Saudi Arabian hospital. Am J Infect Control 2001;29:48–52. 25. Centers for Disease Control and Prevention, National Institutes of Health. Biosafety in microbiological and biomedical laboratories, 4th ed. Washington, DC: U.S. Government Printing Office (HHS Publ. No. CDC 93-8395), 1999. 26. Fleming DO, Hunt DL, eds. Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000. 27. Collins CH. Laboratory-acquired infections: history, incidence, causes, and prevention. Oxford, United Kingdom: Butterworth-Heinemann Ltd, 1993. 28. Centers for Disease Control and Prevention. Updated U.S. public health service guidelines for the management of occupational exposures to HBV, HCV, and HIV and recommendations for postexposure prophylaxis. MMWR Morb Mortal Wkly Rep 2001;50:1–42. 29. National Institute for Occupational Safety and Health. NIOSH alert: preventing needlestick injuries in health care settings. NIOSH publication 2001-108. 1999; November 1-24. NIOSH-Publications Dissemination Cincinnati, OH. 30. Lymer UB, Schutz AA, Isaksson B. A descriptive study of blood exposure incidents among healthcare workers in a university hospital in Sweden. J Hosp Infect 1997;35:223–235. 31. Mahoney FJ, Steward H, Hu H, et al. Progress toward elimination of hepatitis B virus transmission among health care workers in the United States. Arch Intern Med 1997;157:2601–2605. 32. Kao AS, Ashford DA, McNeil MM, et al. Descriptive profile of tuberculin skin testing programs and laboratory-acquired tuberculosis infections in public health laboratories. J Clin Microbiol 1997;35:1847–1851. 33. D'Agata EMC, Wise S, Stewart A, et al. Nosocomial transmission of Mycobacterium tuberculosis from an extrapulmonary site. Infect Control Hosp Epidemiol 2001;22:10–12. 34. Genee D, Siegret HH. Tuberculosis of the thumb following a needlestick injury. Clin Infect Dis 1998;26:210–211. 35. Coggin JH Jr. Bacterial pathogens. In: Fleming DO, Hunt DL eds, Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:65–88. 36. Kakis A, Gibbs L, Eguia J, et al. An outbreak of group A streptococcal infection among health care workers. Clin Infect Dis 2002;35:1353–1359. 37. Gilchrist MJR, McKinney WP, Miller JM, et al. Cumitech 33. In Snyder JW, coordinating ed. Laboratory safety, management, and diagnosis of biological agents associated with bioterrorism. Washington, DC: American Society for Microbiology, 2000. 38. Miller JM. Agents of bioterrorism: preparing for bioterrorism at the community health care level. Infect Dis Clinics North Am 2001;15:1127–1156. 39. Hilmarsdottir I, Coutellier A, Elbaz J, et al. A French case of laboratory-acquired disseminated Penicillium marneffei infection in a patient with AIDS. Clin Infect Dis 1994;19:357–358. 40. Cooper CR, Dixon DM, Salkin IF. Laboratory-acquired sporotrichosis. J Med Vet Mycol 1992;30:169–171. 41. Larson DM, Eckman MR, Alber CL, et al. Primary cutaneous (inoculation) blastomycosis: an occupational hazard to pathologists. Am J Clin Pathol 1983;79:253–255. 42. Oh M, Kim N, Huh M, et al. Scrub typhus pneumonitis acquired through the respiratory tract in a laboratory worker. Infection 2001;29:54–56. 43. Mermel LA, Josephson SL, Dempsey J, et al. Outbreak of Shigella sonnei in a clinical microbiology laboratory. J Clin Microbiol 1997;35:3163–3165. 44. Kolavic SA, Kimura A, Simons SL, et al. An outbreak of Shigella dysenteriae type 2 among laboratory workers due to intentional food contamination. JAMA 1997;278:396–398. 45. Ashdown LR, Cassidy J. Successive Salmonella give and Salmonella typhi infections, laboratory-acquired. Pathology 1991;23:233–234. 46. Rao GG, Saunders BP, Masterton RG. Laboratory acquired verotoxin producing Escherichia coli (VTEC) infection. J Hosp Infect 1996;33:228–230. P.1441 47. Burnens AP, Zbinden R, Kaempf L, et al. A case of laboratory acquired with Escherichia coli O157:H7. Zbl Bakt 1993;279:512–517. 48. Penner JL, Hennessy JN, Mills SD, et al. Application of serotyping and chromosomal restriction endonuclease digest analysis in investigating a laboratory-acquired case of Campylobacter jejuni. J Clin Microbiol 1983;18:1427–1428. 49. Coia JE. Nosocomial and laboratory-acquired infection with Escherichia coli O157. J Hosp Infect 1998;40:107–113. 50. Herwaldt BL. Laboratory-acquired parasitic infections from accidental exposures. Clin Microbiol Rev 2001;14:659–688. 51. Parker SL, Holliman RE. Toxoplasmosis and laboratory workers: a case-control assessment of risk.Med Lab Sci 1992;49:103–106. 52. Batchelor BI, Brindle RJ, Gilks GF, et al. Biochemical mis-identification of Brucella melitensis and subsequent laboratory-acquired infections. J Hosp Infect 1992;22:159–162. 53. Ashdown LR. Melioidosis and safety in the clinical laboratory. J Hosp Infect 1992;21:301–306. 54. Vasconcelos PF, Travassos da Rosa RA, Rodrigues SG, et al. Laboratory-acquired human infection with SP H 114202 virus (Arenavirus: Arenaviridae family): clinical and laboratory aspects. Rev Inst Med Trop Sao Paulo 1993;35:521–525. 55. Barry M, Russi M, Armstrong L, et al. Brief report: treatment of a laboratory-acquired Sabia virus infection. N Engl J Med 1995;333:294–296. 56. Centers for Disease Control and Prevention. Laboratory-acquired West Nile virus infections-United States, 2002. MMWR Morb Mortal Wkly Rep 2002;51:1133–1135. 57. Occupational Safety and Health Administration. Occupational exposure to bloodborne pathogens: needlestick and sharps injuries; final rule. Federal Register 2001;66:5317–5325. 58. NCCLS. Implementing a needlestick and sharps injury prevention program in the clinical laboratory; a report. Standard X3-R. Wayne, PA: NCCLS, 2002. 59. Bolyard EA, Tablan OC, Williams WW, et al. Guideline for infection control in healthcare personnel, 1998. Infect Control Hosp Epidemiol 1998;19:407–463. 60. Collins SM, Hacek DM, Degan LA, et al. Contamination of the clinical microbiology laboratory with vancomycin-resistant enterococci and multidrug-resistant Enterobacteriaceae: implications for hospital and laboratory workers. J Clin Microbiol 2001;39:3772–3774. 61. Richmond JY, Knudsen RC, Good RC. Biosafety in the clinical mycobacteriology laboratory. Clin Lab Med 1996;16:527–550. 62. Cole EC, Cook CE. Characterization of infectious aerosols in health care facilities: an aid to effective engineering controls and preventive strategies. Am J Infect Control 1998;26:453–464. 63. Gilchrist MJR, Fleming DO. Biosafety precautions for Mycobacterium tuberculosis and other airborne pathogens. In: Fleming DO, Hunt DL eds. Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:209–219. 64. Groschel DHM, Strain BA. Laboratory safety in clinical microbiology. In: Balows Hausler WJ Jr, Herrmann KL, Isenberg HD, et al., eds. Manual of clinical microbiology, 5th ed. Washington, DC: American Society for Microbiology, 1991:49–58. 65. Occupational Safety and Health Administration. Enforcement procedures for the occupational exposure to bloodborne pathogens. CPL 2-2.69 2001. 66. Occupational Safety and Health Administration. Occupational exposure to bloodborne pathogens: request for information. Federal Register 1998;63:48250–48252. OSHA, Washington, DC. 67. Occupational Safety and Health Administration. Occupational exposure to bloodborne pathogens: final rule. Federal Register 1991;56:64003–64182. 68. Occupational Safety and Health Administration. Occupational exposure to tuberculosis: proposed rule. Federal Register 1997;62:52149–54309. 69. Centers for Disease Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care facilities. MMWR Morb Mortal Wkly Rep 1994;43:1–128. 70. Garner JS, Hospital Infection Control Practices Advisory Committee. Guideline for isolation precautions in hospitals. Infect Control Hosp Epidemiol 1996;17:53–80. 71. Fleming DO. Risk assessment of biological hazards. In: Fleming DO, Hunt DL eds, Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:57–64. 72. Knudsen RC. Risk assessment for biological agents in the laboratory. In: Richmond JY, ed. Rational basis for biocontainment: proceedings of the Fifth National Symposium on Biosafety. Mundelein, IL: American Biological Safety Association, 1998. 73. World Health Organization. Laboratory biosafety manual, 2nd ed. Geneva: WHO, 1993. 74. Centers for Disease Control and Prevention. Laboratory security and emergency response guidance for laboratories working with select agents. MMWR Morb Mortal Wkly Rep 2002;51:1–6. 75. Cipriano M. Cumitech 36, Biosafety considerations for large-scale production of microorganisms. Washington, DC: American Society for Microbiology, 2002. 76. Johnson MD, Schaffner W, Atkinson J, et al. Autopsy risk and acquisition of human immunodeficiency virus infection: a case report and reappraisal. Arch Pathol Lab Med 1997;121:64–66. 77. McCaskie AW, Calder SJ, Roberts M, et al. Sectioning fresh human bone: the reduction of aerosol and physical hazards. Br J Biomed Sci 1997;54:88–90. 78. Crane JT, Richmond JY. Design of biomedical laboratory facilities. In: Fleming DO, Hunt DL eds, Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:283–311. 79. Stuart DG. Primary barriers: biological safety cabinets, fume hoods, and glove boxes. In: Fleming DO, Hunt DL eds, Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:313–330. 80. Henderson DK. Raising the bar: the need for standardizing the use of “standard precautions” as a primary intervention to prevent occupational exposures to bloodborne pathogens. Infect Control Hosp Epidemiol 2001;22:70–72. 81. Centers for Disease Control and Prevention. Guideline for hand hygiene in health-care settings. MMWR Morb Mortal Wkly Rep 2002;51:1–47. 82. Hunt DL. Standard (universal) precautions for human specimens. In: Fleming DO, Hunt DL eds. Biological safety: principles and practices, 3rd ed. Washington, DC: American Society for Microbiology, 2000:355–367. 83. Zaki AN, Campbell JR. Infectious waste management and laboratory design criteria. Am Indust Hyg Assoc J 1997;58:800–808. 84. U.S. Department of Transportation, Research and Special Programs Administration. Hazardous materials: revision to standards for infectious substances and genetically modified organisms; final rule. Federal Register 2002;67:53117–53144. 85. International Air Transport Association. Dangerous goods regulations, 41st ed. 2000, 57–88. IATA, Montreal, Quebec Canada. 86. Sepkowitz KA. Occupationally acquired infections in health care workers, part I. Ann Intern Med 1996;125:826–834. 87. Knobloch J, Demar M. Accidental Leishmania mexicana infection in an immunocompromised laboratory technician. Trop Med Int Health 1997;2:1152–1155.
Воскресенье, 06 Ноябрь 2011 | 900 Просмотры | Печать | PDF
Infections are an important cause of neonatal morbidity and mortality worldwide. Although most neonatal infections are of maternal or community origin, an increasing proportion are acquired in the nursery. Advances in newborn intensive care have permitted the survival of low-birth-weight and sick infants and have simultaneously created risks for nosocomial infection, which are themselves a significant cause of mortality in these infants (1, 2, 3). Prevention of infection in the premature infant who starts life in an intensive care unit and whose immature defenses are further depleted by illness and invasive procedures is a major challenge. DESCRIPTIVE EPIDEMIOLOGY Surveillance The newborn may acquire infection from the mother in utero or during delivery or postpartum from maternal, hospital, or community sources. Many infections transmitted from mother to infant during delivery, such as group B ОІ-hemolytic streptococci (GBS), Listeria, hepatitis B virus (HBV), or herpes simplex virus (HSV), have not traditionally been considered nosocomial. On the other hand, infections classified as nosocomial are often caused by microorganisms acquired from the mother that become part of the flora of the newborn and subsequently invade because of immature or impaired defenses. Difficulty in distinguishing between maternal and hospital sources makes identification of newborn nosocomial infections imprecise. Because of this difficulty, the Centers for Disease Control and Prevention (CDC) has defined all neonatal infections, whether acquired during delivery or during hospitalization, as nosocomial unless evidence indicates transplacental acquisition (4). Most published reports of nosocomial infections have included only those infections with onset within a specified period after admission to the nursery, whereas some have attempted to separately define infections from maternal and hospital sources. The need to distinguish between maternal and hospital sources is more than semantic. Infection control measures designed to prevent acquisition of microorganisms within the nursery will not affect pathogens acquired perinatally, for which control measures involve prevention, diagnosis, and treatment of infection in the pregnant woman; intrapartum antibiotic prophylaxis; postpartum antibiotic or immune prophylaxis for the infant; and prevention of obstetric complications known to be associated with increased intrapartum transmission. Nosocomial infections may become manifest only after discharge from the hospital, especially in normal newborns whose short hospital stay is within the incubation period for many infections (5). Such infections may be brought to the attention of the infant's primary physician rather than the nursery. Surveillance for nosocomial infection in normal newborns requires close communication between community physicians and the nursery; otherwise, recognition of hospital-based outbreaks may be delayed. Programs for active postdischarge surveillance have been developed (6) but may not be cost effective in the absence of an outbreak, because most sporadic infections in normal newborns are benign. CDC definitions of nosocomial infections in the newborn are based on those for older children and adults with certain modifications for children younger than 12 months (4). Methods of case finding vary and may include prospective daily clinical case review and/or review of laboratory results or retrospective chart review. Laboratory-based surveillance, although very sensitive for bloodstream, urinary tract, and central nervous system (CNS) infections, is less sensitive for infections at other sites and depends on the availability of laboratory facilities and the intensity of testing. Not all centers have ready access to virology cultures. Surveillance may include all infections or infections at specific sites such as bacteremia. The high-risk nursery component of the CDC's National Nosocomial Infections Surveillance (NNIS) system collects data on all infections at all sites (7). An overall newborn nosocomial infection rate is of limited use, because it is influenced by the type of hospital or nursery, the patient mix, referral patterns, whether or not newborn surgery is performed, the type of surveillance, and the denominator used (5, 8, 9). Also, total infections must be distinguished from the numbers of infected patients, because many patients have more than one infection. The denominators most often reported are infections per 100 admissions or discharges or, for maternity hospitals, infections per 1,000 deliveries or live births. For the P.852 neonatal intensive care unit (NICU), where infection risk is related to duration of stay, patient-days is a more appropriate denominator (9) (see Chapter 94). Infection rates expressed per admission or per patient-day may be useful for following infection rates in a specific NICU over time or for interhospital comparison, provided that the rates are adjusted for severity of illness or are expressed by risk group. Severity of illness scores (10, 11, 12, 13) and a measure of the intensity of care required (14) have been used to control for differences in severity of illness in recent reports. Birth weight has been used as a marker for severity of underlying illness in the NICU. Goldmann et al. (15) found a strong correlation between infection and low birth weight, with a mean birth weight of 1,581 g for infants with major nosocomial infections versus 2,607 g for those without. The NNIS system stratifies data by birth-weight groups (16). Use of invasive devices may be a more relevant marker for average severity of illness and for the type of NICU. NICU infection rates vary with intensity of device use. In a study of 35 hospitals using NNIS protocols, assessment of device use (central or umbilical lines and ventilators) by total device-days and calculation of device-associated infection rates by device-days controlled for this variation. Stratification by birth weight did not eliminate the need to control for device use (9). With infants staying in the NICU for longer periods, the need for another classification, the “late, late onset” infection or infection onset after 30 days of age, has arisen. Risk factors and the predominant microorganisms involved differ from those of the short-stay NICU patient (17). Where a system of continuing surveillance has not been established or is not feasible, cross-sectional prevalence studies provide information on the spectrum of nosocomial infections and may help in allocation of resources. In addition, prevalence studies may permit collection of more detailed patient data. National prevalence studies from Spain (18), Norway (19), and the United States (20) have reported NICU infection rates of 16.7%, 14%, and 11%, respectively. Reported Infection Rates There are few data on infection rates in normal nurseries, where the infant is healthy and the hospital stay short. Reported rates are low, from 0.3 to 1.7 per 100 newborns (21, 22, 23). NNIS rates for all newborn nurseries reporting in 1984 were 0.9 and 1.7 per 100 discharges for nonteaching and large teaching hospitals, respectively (24). Similar rates for a small number of hospitals surveyed in Canada in 1984 were 1.4 and 3.1 per 100 discharges (25). Reported infection rates in the NICU in the last 20 years vary from 3.2 to 30 per 100 admissions or discharges, illustrating the wide variability among centers (Table 52.1) (26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37). The low rate from Nice was attributed to minimal use of invasive procedures and antibiotics, rapid enteral feeding, and the limited number of personnel to which each infant was exposed (29). NICUs that admit surgery patients may have higher rates. A rate of 58 per 100 admissions was reported in one small series from a newborn surgery unit (38). Rates controlled for length of stay are 4.6 to 23.8 infections per 1,000 patient-days (Table 52.1). TABLE 52.1. NOSOCOMIAL INFECTION RATES IN THE NICU Infections per 100 admissions or discharges Infections per 1,000 patient-days Year Location Reference 14.0 — 1984–1987 Toronto 26 30.0 — 1987–1989 Curitiba, Brazil 27 14.5 — 1988–1989 Brooklyn 28 3.2 — 1989 Nice 29 28.4 — 1992–1995 Trinidad 30 27.1 21.9 1991 Freiburga 31 6.2 4.8 1994–1995 New South Wales, Australiaa 32 19.1 11.6 1996 Italy (mulicenter) 33 7.0 — 1996 European (multicenter)b 34 19.0 9.9 1987–1997 Montreal 35 18.9 23.8 1993–1997 Sao Paulo, Brazila 36 — 4.6–18.1 (median 8.9) 1997 USA (multicenter) 37 NICU, neonatal intensive care unit. a Stated that maternally acquired are included. b Major infections only. Infection rates in infants of birth weights less than 1,500 g, 1,501 to 2,500 g, and more than 2,500 g were 63%, 8.2%, and 6%, respectively, in a study from Brooklyn (28) and 74%, 28%, and 13%, respectively, in a study from Brazil (39). The association of birth weight with infection rate may be lost if length of stay is considered (Table 52.2). TABLE 52.2. NOSOCOMIAL INFECTION RATES IN THE NICU: RELATION TO BIRTH WEIGHT Birth weight (g) Freiburg 1991 (31) Infections per Montreal 1992–1997 (35) Infections per 100 admissions 1,000 patient-days 100 admissions 1,000 patient-days <1,000 70.4 22.8 25.7 9.9 1,001–1,500 31.9 18.5 39.4 9.6 1,501–2,500 21.4 23.8 19.4 9.0 >2,500 13.6 23.5 9.6 7.1 NICU, neonatal intensive care unit. ACQUISITION OF INFECTION IN THE NEWBORN NURSERY AND NEONATAL INTENSIVE CARE UNIT Bacterial Colonization of the Newborn The fetus is relatively protected from acquisition of microorganisms, although infections may occur transplacentally or by direct extension from the maternal genital tract. Initial exposure to maternal microbial flora usually occurs during passage P.853 through the birth canal. Colonization proceeds more slowly after delivery by cesarean section (40). Postnatally, colonization continues with new microbes acquired from mother, other family members, hospital personnel, and, occasionally, inanimate objects. Healthy newborns establish normal flora within a few days of birth. Gram-positive microorganisms predominate in the pharynx (41, 42), and coagulase-negative staphylococci (CONS) predominate at the umbilicus (41). Gastrointestinal colonization is more complex. Anaerobic bifidobacteria are found in very high concentrations, with smaller numbers of Bacteroides, other anaerobes, and Escherichia coli in stools of breast-fed infants. With formula feeding, Enterobacteriaceae predominate, and there are more Bacteroides microorganisms and other anaerobes and fewer bifidobacteria (40, 41, 42, 43, 44). The lower pH in the gastrointestinal tract of breast-fed infants may permit preferential growth of bifidobacteria (43). The normal newborn with a short hospital stay has little opportunity to acquire nursery flora with the exception of Staphylococcus aureus and some viruses. However, in a report from nurseries in Japan wherein mothers and infants were separated for up to 72 hours after birth, 85% of healthy newborns acquired E. coli strains of nonmaternal, presumably nursery, origin (45). Colonization of the infant in the NICU follows a different pattern as a result of limited maternal contact, delayed feeding, antibiotic treatment, and exposure to NICU flora. Each NICU has its unique endemic flora with colonized infants serving as reservoirs for transmission to newly admitted infants. The pattern of antibiotic use influences NICU flora (5). Colonization occurs later, especially when antibiotics are used or feeding is delayed (46), and gram-negative aerobic bacilli are prominent at multiple sites (41, 42, 46, 47). Graham et al. (41) found that low-birth-weight and ill infants acquired fewer gram-positive and more gram-negative microorganisms in the mouth and umbilicus and more gram-negative aerobic rods and fewer anaerobes in the rectum than did healthy newborns. Klebsiella, Enterobacter, Serratia, and Pseudomonas were more frequent in those receiving antibiotics. Klebsiella, Enterobacter, and Citrobacter were often isolated from stool, nose, throat, or umbilicus of NICU infants studied by Goldmann et al. (47). Risk of acquisition of these microorganisms was increased in infants receiving antibiotics for more than 3 days and with increased duration of stay in the NICU. Antibiotics suppress stool anaerobic flora and favor growth of aerobic gram-negative rods (46, 48). NICU strains are often antibiotic resistant (5, 49, 50, 51, 52). The NICU offers ideal conditions for emergence of resistance: intense antibiotic pressure in a population with high bacterial loads. The high risk of severe infection and the difficulty in making a definitive diagnosis of infection or of identifying the causative microorganisms result in widespread empiric use of broad-spectrum antibiotics. Reports indicate that 50% to 81% of infants admitted to NICU receive antibiotics (47, 53, 54, 55). Resistance appears to occur more rapidly with routine use of broad-spectrum cephalosporins than with penicillins and aminoglycosides (52, 56, 57, 58). CONS were the predominant flora of stool, skin, and mucous membranes in a series of premature infants delivered by cesarean section and not treated with antibiotics (59). Gram-negative bacilli were rare even in the stool. Antibiotic resistance patterns suggested that these CONS were acquired from the nursery. Although strains of CONS are normal flora, infants in the NICU may acquire unique nursery strains characterized by antibiotic resistance and enhanced slime production (60, 61). Acquisition of abnormal flora does not necessarily lead to infection, although the risk of infection may be increased. Sprunt (62) reported nosocomial infections in 0.5% of infants with normal pharyngeal flora and in 15% of those with abnormal flora. She considered normal flora to be protective against colonization with potential pathogens and showed that artificial implantation of О-hemolytic streptococci interfered with the growth of gram-negative rods in the pharynx. A more recent study showed that neonates colonized with viridans group streptococci were less likely to acquire methicillin-resistant S. aureus (MRSA) (63). Experimental evidence in animals suggests that the normal intestinal flora also has a protective effect (64). Sources of Infectious Agents and Modes of Transmission The usual mode of transmission of microbes in the nursery is by contact, either direct physical contact with an infected or colonized person or, more often, transfer from one infant to another on the hands of personnel. The classic experiments of Rammelkamp and colleagues (65, 66) during the 1960s demonstrated hand transfer as the predominant mode of transmission of S. aureus. Hands have been implicated in several nursery outbreaks with various gram-negative bacilli, S. aureus, Enterococcus, and viruses (5, 67, 68, 69, 70, 71, 72, 73, 74). Goldmann et al. (47) documented the P.854 presence of gram-negative bacilli on the hands of 75% of NICU personnel. Usually, hands are transiently contaminated, and hand washing removes the microorganisms and interrupts transmission (75). Less often, personnel who are persistent carriers have been implicated in nursery outbreaks (76, 77, 78). Artificial fingernails have been associated with transmission of Pseudomonas aeruginosa. (68, 69) Hand washing may itself perpetuate outbreaks if hand-washing agents become contaminated (79, 80). Transmission by indirect contact may also occur through patient care equipment that is not adequately decontaminated between patients or through equipment, solutions, and topical preparations contaminated from hands. Infections have been related to contamination of such items as resuscitation equipment (81, 82, 83, 84, 85), suction devices (86, 87, 88), ventilator circuits (89), rectal thermometers (90), and feeding bottles (91, 92) and to application of contaminated eyewash (93), umbilical cord wash (94), mineral oil (95), ultrasound gel (96, 97), glycerine (98), and antiseptic soap (99). Water sources are additional potential hazards in the NICU (100, 101). Microorganisms such as Pseudomonas, Serratia, Stenotrophomonas, and Flavobacterium are prevalent in the inanimate environment and grow readily at room temperature in water and moist environments such as incubators, humidifier reservoirs, and respirator nebulizers and tubing. Pseudomonas infection has been associated with a water bath used to thaw plasma (102) and Stenotrophomonas with tap water used to bathe premature babies (103). Intravenous fluids, especially those used for parenteral nutrition (PN), were frequent sources of infection in the past (104, 105). With current standards of preparation, intravenous fluids are rarely intrinsically contaminated (106) but may become extrinsically contaminated by handling during use (107, 108, 109, 110, 111, 112). Blood transfusions may be a source of viruses such as cytomegalovirus (CMV) (113) hepatitis A virus (HAV) (114), HBV (115), hepatitis C virus (HCV) (116), and human immunodeficiency virus (HIV) (117). Neonatal transfusion-acquired malaria has also been reported (118, 119) (see Chapters 44 and 69). Breast milk is another source of blood-borne viruses (120). Approximately one third of CMV-seropositive women excrete CMV in their breast milk, and two thirds of these women were found to transmit CMV to their newborns by breast-feeding (121). HIV is transmitted from mother to newborn by breast milk, and breast-feeding may be the major means of acquisition of human T-cell leukemia virus type I (HTLV-I) (120). Breast milk may contain bacteria if the mother is bacteremic or has mastitis but is more likely to be contaminated with bacteria during collection or handling. One study reported growth of gram-negative bacilli in 36% of samples of unpasteurized human milk (122). Outbreaks of Klebsiella (123) and Serratia (122, 124) have been associated with contaminated breast milk pumps. Contamination of banked breast milk with Salmonella (125), enteropathogenic E. coli (126), and MRSA (127) has been reported. Formula feeds may also become contaminated during preparation or handling (128). Contaminated blenders have been identified as a source of infection (129, 130). Bacteremia and meningitis have been associated with powdered infant formula intrinsically contaminated with Enterobacter sakazakii (131). Infected personnel and visitors may introduce pathogens into the nursery, especially during community outbreaks of viral infections. Most respiratory viruses are spread by large droplets expelled from the respiratory tract that travel distances of less than a meter and settle on surfaces close to the infected person. Infection is acquired by close contact with the infected person or with contaminated objects (132). This is probably not an important means of transmission between newborns, because they do not cough vigorously and are inefficient generators of droplets. However, it can be an important means of transmission to newborns from infected personnel or visitors (133). True airborne transmission, in which microorganisms remain suspended in aerosols for significant periods and are carried by air currents over considerable distances, occurs infrequently in the nursery. Diseases transmitted in this manner include varicella, measles, and tuberculosis, all of which are rare in newborns. Airborne fungal spores are rare causes of infection in NICUs, but such infections may be devastating (5). Risk Factors for Infection The Newborn The newborn is at risk for infection because of immaturity of the normal structural barriers and the immune system. The defenses of the premature newborn are particularly inadequate. Before 32 weeks' gestation, the stratum corneum is poorly developed and the skin is fragile, easily traumatized, and very permeable. The skin matures rapidly postnatally and, by 2 weeks of life, is well developed in most newborns regardless of gestational age (134). Immune function in the newborn has been extensively reviewed by Lewis and Wilson (135) Although the fetus is capable of early antibody production, there is little antibody synthesis in utero, and the newborn initially depends on passively transferred maternal antibody. The repertoire of antibodies received depends on maternal exposure. The newborn of less than 34 weeks' gestation may not receive protective levels of antibody. Premature newborns respond adequately to most protein antigens, but response to polysaccharide antigens is poor in the first 2 years of life. Opsonization activity of the alternate complement system and serum fibronectin levels are deficient in the term infant. These deficiencies put the newborn at risk for overwhelming bacterial and fungal infections. T-cell function is important for control of intracellular pathogens such as Listeria, Toxoplasma, and Salmonella. The newborn has a high total T-lymphocyte count, but phenotypic surface markers differ from those in the older child. Cytotoxic T-cell activity is decreased as is T-cell helper function. T-cell dependent antigen specific response is delayed, and there is limited production of several cytokines. Natural killer cell activity, important in control of herpes group viral infections, is also decreased. The newborn has a decreased granulocyte storage pool and defective neutrophil and monocyte chemotaxis. Under optimal conditions, phagocytosis and microbial killing by neonatal granulocytes are normal, but these functions are impaired if opsonization is deficient. P.855 Environment Infection rates in the NICU increase with overcrowding and understaffing. Haley and Bregman (136) reported a 16-fold increase in outbreaks of S. aureus infection when the infant-to-nurse ratio exceeded 7 and a sevenfold increase when the nursery was crowded. More recently, they reported increasing rates of endemic MRSA linked to overcrowding and understaffing, with eradication of MRSA when these conditions improved (14). An outbreak of Enterobacter cloacae infection was associated with understaffing and overcrowding in another report (137). Goldmann et al. (15) observed a decrease in percentage of nosocomial infections from 5.8 to 1.8 after a move to a new NICU with more nurses and space per infant, more accessible sinks, and improved ventilation. Invasive Procedures Any procedure that disrupts the normal barriers to infection is likely to present a higher risk of infection in the newborn than later in life. The normal newborn escapes most invasive procedures but may be subjected to scalp electrodes or percutaneous punctures for blood sampling. Scalp electrodes provide a portal of entry for maternal genital microorganisms. Infectious complications occur in less than 1% of infants and most are benign abscesses, but severe cellulitis, bacteremia, osteomyelitis, and disseminated HSV infection have been reported (138, 139). Osteomyelitis has resulted from infected toe and heel punctures (140). Premature and ill newborns often require feeding by nasogastric tubes, which provide a portal of entry and potentiate overgrowth of microbes in the upper gastrointestinal tract (74, 141, 142, 143). Breast milk (122) and formula feeds (128) administered by continuous infusion remain at room temperature for several hours, allowing microbes to proliferate in the reservoir or tubing during infusion. Botsford et al. (122) found that colony counts of enteric gram-negative rods in milk samples taken from connecting tubings changed after 24 hours of use were often greater than 106/mL. Counts of 103/mL or higher were associated with feeding intolerance. Counts of 106/mL or higher were associated with symptoms of sepsis and necrotizing enterocolitis (NEC) in some infants. In another study, formula contamination with 106/mL of gram-negative bacteria was associated with NEC (143). Bacterial colonization of intravascular catheters, including peripheral lines, occurs more often in neonates than in older children (144). Infections are rare with radial artery and scalp vein needles (145), although the latter did serve as significant portals of entry in an outbreak of Serratia infection (146). Catheter-associated bloodstream infection (BSI) rates compiled from several studies were 3.7% for umbilical vein and 1.2% for umbilical artery catheters and 3% to 8% for central venous lines (CVLs) used for PN (145). BSI was associated with 5% of umbilical artery and 3% of umbilical vein catheters in one report (147) and with 17.6% of umbilical artery and 13.2% of umbilical vein catheters in another. Umbilical catheterization for more than 5 days was an independent risk factor for sepsis (33). Rates of BSI associated with Broviac cuffed silicone catheters are 2.9 to 5.4 per 1,000 catheter-days (148). These are higher than in older children and are particularly elevated in low-birth-weight infants (149). Peripherally inserted, noncuffed, silicone CVLs are being increasingly used in newborns because of ease of insertion and decreased trauma to small central vessels. Infection rates of 1 to 5.7 per 1,000 catheter-days are reported (148, 150, 151). In a recent multicenter study, rates of BSI for umbilical, percutaneous central, and tunneled catheters were 7.2, 13.1, and 12.1 per 1,000 catheter-days, respectively (152). Risk factors for catheter-associated BSI include catheter disconnection, blood sampling, and colonization of the catheter hub (153). Contamination of intravascular pressure-monitoring devices has resulted in neonatal infections (154, 155). Extracorporeal membrane oxygenation was associated with a BSI rate of 3.4%; duration of bypass was the major risk factor (156). Other invasive devices associated with infection in the newborn include endotracheal tubes, urinary catheters, and ventriculoperitoneal shunts (see later discussion). Infants have higher rates of catheter-associated urinary tract infections (UTIs) (157) and ventriculoperitoneal shunt infections (158) than older children. However, NNIS data suggest a lower rate of ventilator-associated pneumonia in NICUs than in other intensive care units (16). INFECTIONS AT SPECIFIC SITES In the normal nursery, most infections are superficial infections of the skin, mouth, and eyes. Outbreaks of infections with common viral pathogens also occur (5, 22). These outbreaks may represent admission of large numbers of maternally infected infants or the introduction of a microbe by one or a few maternally infected infants and subsequent transmission in the nursery. In the NICU, skin and mucous membrane infections predominate in most early reports (8). More recently, BSIs have been more common (Table 52.3). In general, BSIs are more common and surgical site infections and UTIs are less common than in older children and adults (7, 8, 24). NNIS data of infection site by birth-weight group are summarized in Table 52.4 (7). Approximately 15% of BSIs and pneumonia and less than 8% of the other infections were maternally acquired. TABLE 52.3. NOSOCOMIAL INFECTIONS IN THE NICU: PERCENTAGE OF INFECTION BY SITE IN SELECTED STUDIES Site of infection Toronto 1984–1987 (26) Freiburg 1991 (31)a Montreal 1987–1997 (35) SГЈo Paulo 1993–1995 (36)a USA multicenter 1999 (20)b Skin, mucous membranes 15 11 19 17.5 8.6 Pneumonia 7 32 6 42.1 Not specified Upper respiratory tract 4 2 9 0.3 Not specified Respiratory tract unspecified — — — — 12.9 Blood 45 27 27 29.1 52.6 Urinary tract 5 3 14 2.4 8.6 Surgical site 2 11 5 1.4 Not specified Gastrointestinal tract 16 2 19 6.5 Not specified Other 7 2 6 — 17.2 NICU, neonatal intensive care unit. a Includes maternally acquired. b Prevalence study. TABLE 52.4. NOSOCOMIAL INFECTIONS BY SITE OF INFECTION AND BIRTH WEIGHT <1,000 (n = 3,987) % Birth weight (g) 1,001–1,500 (n = 1,881) % 1,501–2,500 (n = 3,547) % >2,500 (n = 3,764) % Blood 43 38 26 28 Pneumonia 16 12 13 18 Gastrointestinal tract 7 10 11 5 Conjunctivitis 4 10 16 9 Ear, nose, throat 4 4 5 4 Skin, soft tissue 6 7 10 9 Surgical site 1 1 3 7 Clinical sepsis 6 7 6 8 Other 13 11 10 12 Data from the National Nosocomial Infections Surveillance, 1986–1994. From Gaynes RP, Edwards JR, Jarvis WR, et al. Nosocomial infections among neonates in high-risk nurseries in the United States. Pediatrics 1996;98:357–361, with permission. Skin, Subcutaneous Tissues, Mouth, and Eyes Pustules, cellulitis, subcutaneous abscesses, lymphadenitis, and infections at sites of percutaneous punctures are most often due to S. aureus, although streptococci may also be involved. Infections resulting from gram-negative bacilli and other microorganisms may occur in the NICU (159). Microbes causing infections at scalp monitor sites are more diverse and include maternal genital microorganisms such as HSV (138, 139), Mycoplasma hominis (160), and Gardnerella (161). Candida infections of the skin and mouth are frequent in infants in NICUs (162). Omphalitis is uncommon; it occurred in 0.5% of term and 2% of preterm infants in one report (163). The presentation varies from mild erythema or serous drainage to purulent discharge, cellulitis, and acute necrotizing fasciitis of the abdominal P.856 wall. S. aureus is most often isolated, but group A Streptococcus, CONS, enterococci, gram-negative rods, and anaerobes may also be involved (164, 165). A mortality rate of 7% was reported in a series of hospitalized patients. All deaths were in infants presenting with rapidly progressing cellulitis or necrotizing fasciitis (165). Tetanus secondary to umbilical infection is common when conditions of delivery and cord care are not hygienic and mothers are nonimmune (166) (see Chapter 47). Circumcision is the most common surgical procedure performed in the newborn. For CDC surveillance purposes, circumcision infections are classified with skin and soft tissue infections and not with surgical site infections. Reported infection rates are low, at 0.06% to 0.4% (167, 168). Most are simple site infections, but bullous impetigo, staphylococcal scalded skin syndrome, necrotizing fasciitis, and bacteremia have been reported (159). Infectious conjunctivitis in the normal newborn is most often due to Chlamydia and presents after discharge from hospital (159). Gonococcal infections, which are rare, may present earlier. S. aureus infection is next in frequency to Chlamydia and may be of nursery origin with outbreaks occurring in normal nurseries and NICUs (71). In the NICU, conjunctivitis is a frequent finding in outbreaks of Serratia (80, 87). P. aeruginosa conjunctivitis has been associated with contaminated resuscitation equipment (86), and infection in intubated patients has been related to endotracheal tube colonization and eye contamination during suctioning (169). P. aeruginosa eye infection is particularly severe in premature infants, in whom it may progress rapidly to destruction of the cornea by proteolytic enzymes, invasion of the eye, and secondary bacteremia (170). Endophthalmitis secondary to BSI occurs with disseminated Candida infection (171). Ophthalmologic examination was a common factor in an outbreak of adenoviral conjunctivitis (172). Bloodstream Infections Neonatal BSI is reported in one to eight newborns per 1,000 live births (173). Early-onset BSI, occurring shortly after birth as a result of infection in the birth canal, is characterized by fulminant multisystem disease with a high mortality rate. Risk P.857 factors are prematurity, low birth weight, prolonged rupture of membranes, maternal chorioamnionitis, and maternal fever. Predominant causes are GBS and E. coli. Pathogens generally seen in older infants, such as Haemophilus influenzae and Streptococcus pneumoniae, are now encountered in increasing numbers in newborns (17, 54, 173, 174). Late-onset BSI, usually occurring after the first week of life, is also often due to maternal microorganisms but shows less relation to obstetric complications, occurs more often in term infants, and is often associated with focal infection, especially meningitis. Late-onset BSI acquired in the nursery is more often due to CONS, S. aureus, enterococci, or gram-negative rods (17, 175). In BSI occurring after more than 30 days in the NICU, CONS and Candida are prominent (17). Microorganisms isolated in neonatal sepsis have been monitored at Yale-New Haven Hospital since 1928 (17). Changes over time are presented in Table 52.5. Beta-hemolytic streptococci, S. aureus, and E. coli were the main microorganisms encountered during the first 20 years. With increasing use of antibiotics, E. coli, P. aeruginosa, and Klebsiella-Enterobacter became more important, reaching a peak in the period 1958 to 1965. GBS became prominent in the 1970s. In the 1980s, GBS and E. coli remained the most frequent isolates, but CONS became more prominent. There were increasing rates of late, late onset infections related to increasing admission and survival of infants of birth weight less than 1,000 g. In 1979 to 1988, the overall rate of BSI was 2.7 per 1,000 live births for newborns less than 30 days of age and 3.8 per 1,000 if those remaining in the NICU for more than 30 days were included. Of these infections, 46% were of early onset, 27% had onset at 5 to 30 days, and 28% had onset after 30 days of life. TABLE 52.5. MICROBIOLOGY OF NEONATAL SEPSIS IN INFANTS BORN AT YALE-NEW HAVEN HOSPITAL, 1928 TO 1988 A 1993 review from Spain reported that sepsis and/or meningitis occurred in 4.9 per 1,000 live births, again with increasing GBS and CONS infections and decreasing gram-negative bacillary infections over time (176). A similar pattern has been reported in many centers in North America and Europe (173). In developing countries, GBS is less common and gram-negative rods and S. aureus are more prominent (177) In data reported to NNIS from 1986 to 1994, GBS accounted for 46% of maternally acquired BSI, followed by CONS (12%) and E. coli (10%). CONS were associated with 58% of BSI acquired in the NICU, followed by Candida (9%), S. aureus (8%), and enterococci (7%) (7). Anaerobes are an uncommon cause of neonatal bacteremia, accounting for only 2.2% of infections over an 18-year period in a series from New York. Those occurring in the first 48 hours of life were usually penicillin-sensitive gram-positive microorganisms associated with chorioamnionitis. Gram-negative anaerobes were more common after 48 hours and were often associated with NEC and gastrointestinal perforation (178). Intrapartum prophylaxis for early-onset GBS sepsis has resulted in a shift in predominant pathogens. At Yale-New Haven Hospital, an increasing incidence of gram negative rod bacteremia was noted in 1995 to 1997; maternal intrapartum antibiotic prophylaxis was an independent risk factor (179). However, another study in Connecticut showed a decrease in incidence of early-onset GBS sepsis from 0.61 to 0.23 per 1,000 live births but no increase in early-onset non-GBS sepsis (180). A multicenter study in the United States showed no increase in rate of gram-negative sepsis but an increasing proportion of E. coli infections that were resistant to ampicillin (181). In Australia, decreased P.858 rates of early-onset GBS sepsis and also non-GBS sepsis were reported (182). Low birth weight is a major risk factor for BSI, with very-low-birth-weight (VLBW) infants at especially high risk. A multicenter study of sepsis in VLBW infants was carried out by the U.S. National Institute of Child Health and Human Development (NICHD) from 1991 to 1993 (54, 175). Early-onset BSI (within 72 hours of birth) was uncommon, occurring in 1.9% of infants. The predominant microorganisms were GBS (31%), E. coli (16%), and H. influenzae (12%). In contrast, late-onset BSI occurred in 25% of infants. The microorganisms most often isolated were CONS (55%), S. aureus (9%), enterococcus (5%), and Candida (7%), with enteric gram-negative rods accounting for 12%. Early- or late-onset infection increased length of hospitalization, but only late-onset disease increased risk of death. Infants infected with gram-negative bacteria or Candida were more likely to die. BSI rates by birth weight reported from the Yale-New Haven Hospital (17), the Vermont-Oxford Trials Network Database (183), and the NICHD study (54, 175) are presented in Table 52.6. TABLE 52.6. BACTEREMIA RATES IN RELATION TO BIRTH WEIGHT Yale-New Haven Hospital (17) Vermont-Oxford Trials Network (183) NICHD (54,184) Birth weight (g) Bacteremiaa Birth weight (g) Bacteremiab Bacteremiab Early onset Late onset <1,000 17.2 в‰750 26 2.4 50 1,000–1,499 6.1 751–1,000 22 2.3 33 1,500–2,499 1.5 1,001–1,250 15 1.6 21 ≥2,500 0.11 1,251–1,500 8 1.7 10 a Bacteremias per 100 births. b Bacteremic infants per 100 births. In a NICHD report from 1998 to 2000, the rate of early-onset sepsis in VLBW infants was 1.5%, similar to the previous rate. However, the rate of GBS sepsis decreased from 5.9 to 1.7, whereas the rate of E. coli sepsis increased from 3.2 to 6.8 per 1,000 live births (184). Rate of late-onset sepsis and the microorganisms isolated were similar to those reported earlier (185). The effect of birth weight on risk of BSI remains when CVL use is taken into account. NNIS data for 1995 to 2002 showed median rates of CVL-associated BSI of 10.3, 6.6, 3.9, and 2.7 per 1,000 CVL-days for infants of birth weights below 1,000 g, 1,001 to 1,500 g, 1,501 to 2,500 g, and more than 2,500 g, respectively (16). Although low birth weight and use of CVL and PN are interrelated, CVL and PN are independent risk factors for BSI. Donowitz et al. (186) reported Broviac catheters, PN, and surgery as risk factors; catheters were the major risk in infants weighing less than 1,500 g, and surgery was the major risk in those weighing more than 3,000 g. Administration of lipid is an independent risk factor for CONS bacteremia (12, 187). Low birth weight; admission for respiratory illness; and treatment with H2-blockers (11), mechanical ventilation (33, 36), and dexamethasone (188) were independently associated with BSI in other studies. Neonatal bowel resection was a risk factor for late, late onset nosocomial gram-negative bacteremia; a report showed that 39% of infants developed infection at a mean of 17 weeks after surgery (189). Enteral feeding was associated with an increased risk of gram-negative bacteremia in infants with short bowel syndrome, possibly resulting from an effect of enteral feeding on bacterial translocation across the gut wall (190). Central Nervous System Infections The microorganisms isolated and risk factors identified for BSI also apply to neonatal meningitis, which occurs in approximately 25% of bacteremic newborns (173). In a review of neonatal meningitis from Parkland Memorial Hospital, Dallas from the period 1969 to 1989, microorganisms isolated were GBS (53%), gram-negative enteric bacilli (31%), Listeria (7%), and other gram-positive cocci (6%). E. coli, Klebsiella, Enterobacter, and Citrobacter were the principal enteric bacilli; Klebsiella and Enterobacter were more frequent in premature infants (191). In contrast, gram-negative bacteria, predominantly Klebsiella and E. coli, accounted for 64% of cases of neonatal meningitis in a report from Panama. CONS and S. aureus were the most common gram-positive isolates (192). Outbreaks of nosocomial CNS infections in NICUs are usually due to gram-negative bacilli (91, 93, 96, 109, 193, 194, 195, 196, 197), although Listeria (82, 95), group A streptococci (198, 199), and S. aureus (200) have also been involved. Citrobacter diversus, a cause of sporadic meningitis of maternal origin in normal newborns, has also caused clusters of hospital-acquired CNS infections in normal nurseries and in NICUs (201, 202). Meningitis is complicated by focal brain lesions in 77% of cases (203). Outbreaks of meningitis from Campylobacter fetus (204) and Campylobacter jejuni (205) have been reported. Viruses causing maternally or nursery-acquired CNS infection are discussed later. Neural tube defects (173, 191) and ventricular drains and shunts (158) are risk factors for CNS infection. In a study of shunt placement for hydrocephalus in newborns of less than 2,000 g, the shunt infection rate was 25% after primary placement and 36% after revision (206). Pople et al. (158) found that contamination during surgery correlated with a high preoperative skin bacterial density, which was found in younger infants. CONS was the most frequent etiology, and CONS strains with high adherence were more common in the newborn. The role of perioperative antibiotic prophylaxis in the newborn has P.859 not been established, but preoperative antiseptic bathing and use of intraoperative topical antiseptics may be beneficial (158). (For additional information on CNS infections in newborns, see Chapters 27 and 49.) Respiratory Tract Infection Early-onset pneumonia is usually related to intrapartum infection and is most often due to GBS (7, 207). Maternally transmitted Chlamydia (208) and Ureaplasma infection (209) may cause pneumonia, usually mild, with onset at 1 to 3 months of age. Ureaplasma may cause earlier more severe infection in the very premature infant. Respiratory viruses may be acquired postnatally from the mother or from nursery sources (see later discussion). S. aureus and gram-negative enteric bacilli are important causes of nosocomial pneumonia in NICUs (7, 15). CONS is also implicated (207, 210, 211). In the 1986 to 1994 NNIS data, the predominant etiologic agents reported were S. aureus (16.7%), CONS (16.5%), P. aeruginosa (11.7%), and Enterobacter (8.2%) (7). Gram-negative microorganisms have often been reported in outbreaks, which have occurred after exposure to contaminated resuscitation and respiratory therapy equipment (83, 86, 87). Endotracheal intubation is a major risk factor; pneumonia developed in 10% of intubated neonates in one report (207) and 16% in another (211). The pathogenesis of ventilator-associated pneumonia in the newborn has been less well studied than in the adult, but mechanisms are expected to be similar (212). Colonization of the gastrointestinal tract precedes colonization of the respiratory tract (213). The mean rates of NICU ventilator-associated pneumonia in NNIS hospitals reported for 2002 were 2.4, 1.7, 2.0, and 0.7 infections per 1,000 ventilator-days in birth-weight groups of less than 1,000 g, 1,001 to 1,500 g, 1,501 to 2,500 g, and greater than 2,500 g, respectively. These rates are lower than those reported for any other intensive care units (16). However, diagnosis of pneumonia in the intubated newborn is difficult, and many infants are treated empirically for presumed lung infection. Underlying lung disease complicates interpretation of radiographic changes, and procedures used to diagnose pneumonia in the older patient, such as bronchoscopy and lung biopsy, are rarely performed. Specific microbiologic diagnosis is rarely obtained unless there is secondary bacteremia. Endotracheal cultures are useful in predicting the etiology of perinatal pneumonia (214). However, the respiratory tract of the intubated newborn rapidly becomes colonized, and, subsequently, such cultures are not helpful in determining the cause of invasive infection or diagnosing pneumonia (207, 215, 216). Intubated newborns are also at risk for otitis media (217) and tracheitis (218). Surfactant is now used in many premature infants to prevent or treat respiratory distress syndrome. There has been no report of an increased infection rate with surfactant use; in fact, decreasing the need for ventilation would be expected to reduce the risk of infection (5). However, two infants receiving surfactant developed rapidly progressive necrotizing pneumonia resulting from Bacillus cereus, an unusual pathogen in the newborn. Surfactant was not found to be contaminated, but it was postulated that it might have served as a growth factor for this lecithinase-producing microorganism, which is commonly found in the environment (219). Gastrointestinal Infections The newborn is at increased risk for infections with gastrointestinal pathogens, because local gastrointestinal immunity and normal flora have not yet developed and the high gastric pH and short gastric emptying time allow ingested microorganisms to survive and to be passed to the intestine. Normal newborns and those in the NICU are susceptible. When outbreaks have occurred, the index case has often been infected by vertical transmission, and subsequent transmission has occurred by way of the contaminated hands of personnel or contaminated equipment. Infected personnel have rarely been involved (5, 220). Epidemic nosocomial diarrhea is especially a problem in developing countries, where outbreaks of Salmonella and enteropathogenic or enterotoxigenic E. coli occur often (88, 166, 220, 221). Nursery outbreaks of enteropathogenic E. coli infection were common in developed countries in the past but are now unusual. Outbreaks may be explosive, with rapid progression and symptomatic infection of most newborns at risk, or more indolent, with occasional clinical infections and many carriers. Salmonella outbreaks are often protracted and recognized late because of delayed onset of symptoms and prolonged carriage in the newborn (220, 222). The newborn is at high risk for Salmonella bacteremia, and focal infections are also common. Although Shigella has a low infective dose and spreads very rapidly in older children, symptomatic infection in the newborn is surprisingly rare and transmission to other newborns is unusual; transmission to nursery personnel has been reported (223). C. jejuni is an uncommon newborn pathogen and is usually of maternal origin, but nosocomial outbreaks have been described (220, 224, 225). Rotavirus is the gastrointestinal pathogen most often identified in nurseries in developed countries. Infections may be epidemic or endemic and are often asymptomatic (220, 226) (see later discussion). There is little information on the role of other viruses in neonatal gastrointestinal disease. (For additional information on nosocomial gastrointestinal infections, see Chapters 24 and 50.) Necrotizing Enterocolitis NEC is a disease of multifactorial origin involving an immature gastrointestinal tract, ischemia, overgrowth of gastrointestinal bacteria, oral feeding that provides substrate for bacterial growth or production of toxins, and local production of inflammatory mediators (227, 228). Although not strictly an infectious process, microbes are involved in the pathogenesis. It is included in the CDC definitions of nosocomial infections; therefore, data on NEC are found in most surveillance reports. Cases often occur in clusters, suggesting a transmissible agent. It is likely that any microbe capable of causing damage to the gastrointestinal tract can contribute to this disease; thus, NEC may accompany any outbreak of gastrointestinal infection in a population at risk. Microorganisms isolated from stool often reflect those predominant in the NICU at the time, and isolates from blood P.860 and peritoneal fluid are those that have invaded through the damaged gut wall and are not necessarily causative. NEC is mainly a disease of the convalescing premature infant, occurring in up to 10% to 15% of VLBW infants. Term infants are also affected, making up 5% to 10% of cases (159). Microbes temporally associated with outbreaks of NEC include Klebsiella; Clostridia; E. coli; Serratia; Pseudomonas; Staphylococcus epidermidis; and enteric pathogens such as Salmonella, toxigenic E. coli, and S. aureus (227, 228, 229). Delta toxin produced by S. epidermidis and S. aureus is enteropathogenic, and many of the other microorganisms encountered are toxin producers. Free cytotoxins have been detected in stools in some outbreaks (229). The role of Clostridium difficile toxin, which is often found in the stools of asymptomatic infants, is not clear. Rotavirus has also been associated with outbreaks of NEC (230). When compared with other infants with NEC, those with rotavirus were of older gestational age at birth, had been fed at an earlier age, were older at onset of symptoms, and had less severe disease (231). Prophylactic oral vancomycin protected VLBW infants against NEC and may be indicated in specific situations, but routine use may increase the risk of colonization with resistant microorganisms. (232). Breast milk is protective (233). In one study, oral immunoglobulin containing IgA and IgG was protective when fed to low-birth-weight infants for whom breast milk was not available (234). Feeding neonates with Lactobacillus and Bifidobacterium reduced the incidence of NEC when compared with historical controls (235) (see also Chapter 24). Urinary Tract Infection Neonatal UTI has been reported in approximately 0.7% of term infants and 1.9% to 2.9% of high-risk and premature infants (236, 237). In a study of nosocomial UTIs in a children's hospital, the highest rate of infection was observed on the neonatal surgery unit at 4.8 UTIs per 100 admissions. The NICU rate was 1.9 per 100 admissions. Seventeen percent of the NICU and 44% of the neonatal surgery infections were catheter-related. One third of the infections were due to CONS, followed in frequency by Candida, enterococci, and Klebsiella (238). Another hospital reported an infection rate of 0.79 UTIs per 100 NICU admissions. Two thirds of these were catheter-associated, and one third was due to CONS (239). In another pediatric study, the rate of catheter-associated UTI was highest in neonates, at 10.2 per 1,000 patient-days (240). Catheters are not often used in newborns, but when they are the risk of infection is high, possibly because of the use of feeding tubes, which are not well stabilized, rather than the balloon-tipped catheters used in older children (157). Bacteria acquired in the nursery may also cause pyelonephritis later in infancy. Tullus (241) found an association between pyelonephritis resulting from a nephritogenic strain of E. coli in children younger than 2 years and previous admission to a particular neonatal ward. Similar strains were found in stools of personnel and newborns on the ward. UTI rates over several years correlated with the bed occupancy rate on the ward at the time of admission there. Surgical Site Infections Neonates are also at elevated risk for surgical site infections (242, 243). Reported infection rates are presented in Table 52.7. Small-for-date infants undergoing major procedures had an increased risk of infection compared with premature and term infants. Use of prophylactic antibiotics was associated with a lower infection rate after potentially contaminated surgery but not after clean surgery (244). In the largest study reported to date involving 1,433 operations over 12 years, infection rates increased over time. Risk factors for infection were increased incision length, increased duration of surgery, and contamination of the operative site, but there was no relation to gestational or chronologic age or birth weight (245). Data from our institution (246) show neonates to have higher infection rates than older children, but rates were lower than those reported in earlier publications (Table 52.7). TABLE 52.7. SURGICAL SITE INFECTION RATES AFTER NEONATAL SURGERY: PERCENTAGE OF SURGICAL SITES INFECTED Location and year(s) of study (reference) Surgical site classification United Kingdom 1973 (242) United Kingdom 1975–1987 (245) United Kingdom 1989a (244) India 1986a (243) Montreal 1992–1997 (246) Clean 18.5 11.1 3.9 5.3 4.6 Potentially Contaminated 45 20.9 11.2 21.2 5.0 Contaminated/dirty 55.5 20.5 16.7 42.9 1.9 Total 38 16.6 — 13.7 4.5 a Year of report. In contrast to these reports is one of a lower infection rate in newborns than older children, with only 1 of 137 neonatal surgical sites becoming infected (247) and another showing no difference in infection rates by age (248). The reason for these discrepancies is not apparent, but differences in patient population and types of surgical procedures must be considered. Infection rates classified by surgical site infection risk, severity of illness, and duration of operation are now used for surgical site infection surveillance (16), and a classification system is needed for newborn surgery. S. aureus, CONS, and E. coli are the microorganisms most often reported (7, 242, 244, 245). Despite the high risk of surgery-related P.861 infection, there are few data on the efficacy of antibiotic prophylaxis for surgical procedures in the newborn (245), and there are no precise guidelines. The American Academy of Pediatrics (AAP) includes “body cavity exploration in neonates” in a statement concerning circumstances in which prophylaxis in clean surgical site procedures may be justified (249). (For additional information on surgical site infections, see Chapter 21.) Bone and Joint Infections Neonatal osteomyelitis and septic arthritis are usually secondary to bacteremia or fungemia and, as such, may be of maternal or nursery origin. Osteomyelitis may also occur from local infection at sites of invasive procedures (140). INFECTIONS CAUSED BY SPECIFIC MICROORGANISMS Microorganisms causing infections in the normal nursery are usually true pathogens, acquired either from the mother (e.g., GBS, Listeria, HSV) or in the nursery (S. aureus, group A Streptococcus, bacterial enteric pathogens, and respiratory and enteric viruses). In the NICU, additional causes of infection are microorganisms that are not usually pathogenic in the healthy newborn, including gram-negative enteric bacilli, other gram-negative bacilli from water sources, and commensal microorganisms such as CONS and Candida. With the development of neonatal intensive care and widespread use of antibiotics and respirator therapy, Pseudomonas replaced S. aureus as the microorganism of main concern in the nursery. With better control of environmental sources of infection, gram-negative enteric bacilli such as Klebsiella, Enterobacter, and Serratia became more prominent. Increased survival of VLBW infants and increasing use of invasive procedures have resulted in the increasing prevalence of CONS and Candida. Gram-positive cocci predominate in infections reported to the NNIS system in recent years (7). Viruses, not included in the earlier reports, are recognized as important causes of infection in NICU. Staphylococcus aureus S. aureus was recognized as a cause of newborn infection in the nineteenth century. Investigations of nursery outbreaks during the pandemic in the 1950s and 1960s contributed significantly to the understanding of nosocomial infections and the development of infection control policies. In the 1970s, less virulent strains became prominent, but S. aureus remains an important cause of newborn infections (250, 251). Newborns become colonized with S. aureus within the first few days of life at rates of 40% to 90%. The microorganism may be acquired from the mother but is more often of nursery origin (250). Mortimer and co-workers (65, 66) demonstrated that transfer of infection between infants was by the hands of personnel rather than airborne. Infants rarely became infected with nursery strains unless cared for by the same personnel, even when bassinets were in close proximity. Transmission may be reduced by hand washing (65). When nurses who were staphylococcal carriers handled infants through the portholes of incubators for 10 minutes, 20 of 37 infants (54%) became colonized, yet a carrier nurse sitting between two bassinets for 8 hours a day did not transmit the microorganism (66). In most instances, personnel are transiently colonized, but outbreaks have occasionally been linked to true staphylococcal carriers (77, 252). Rates of endemic S. aureus disease are usually in the range of three to six infants with mild skin infection per 1,000 live births. Overall rates of colonization do not correlate with outbreaks of disease (253), which occur when more virulent strains are introduced (250). During epidemic periods, attempts were made to reduce colonization and infection by bathing infants with hexachlorophene. Concern about neurotoxicity (254) led to a discontinuation of this practice and a subsequent increase in rates of infection (255, 256). Because colonization usually begins at the umbilicus, application of various antiseptic and antimicrobial agents to the umbilical stump has been used as a control measure with varying degrees of success (250). Artificial colonization with an avirulent strain of S. aureus has been used successfully to prevent colonization with virulent strains (257). Infections occur in normal newborns, usually after discharge, and in infants in the NICU. Risk factors for infection include skin abrasions and invasive procedures. Most infections are benign skin pustules, but bullous impetigo, cellulitis, scalded skin syndrome, staphylococcal toxic shock, omphalitis, breast abscess, conjunctivitis, and infections of surgical sites and puncture sites occur, as do occasional cases of bacteremia, pneumonia, osteomyelitis, and enterocolitis. S. aureus is an important cause of infections of ventricular drains and shunts (250) (see also Chapters 27 and 49). MRSA infections have become a problem in newborn nurseries. Most outbreaks have been described in NICUs, but outbreaks occur in normal newborn nurseries as well, where large numbers of mothers and infants may become infected because of delayed recognition (258). Infection may become endemic with persistent newborn colonization rates of more than 30% (259) and may involve multiple strains (200). The major reservoir in the nursery is the infected newborn, although infected staff are occasionally involved (78, 260). MRSA strains are not necessarily more virulent but are more difficult to treat when infection does occur, which has led to renewed attempts to eliminate colonization. Cohorting and hand washing have been successful in some outbreaks (200, 261). Contact Precautions were effective in another (262). Other measures used have been replacement of chlorhexidine hand wash with hexachlorophene (72, 200) or triclosan (263) and application of mupirocin to the cord and nares of newborns (264). Mild skin infection is the most common manifestation, but outbreaks of severe disease have occurred (200, 265) (see also Chapters 28 and 29). Coagulase-negative Staphylococci CONS are normal inhabitants of the newborn skin and nose and rarely cause disease in the healthy infant (210). They can colonize prosthetic materials, especially intravascular catheters P.862 and endotracheal tubes, and they produce large amounts of extracellular matrix or slime and other adherence factors (104). The increasing prominence of CONS infections in NICUs is due to increasing numbers of infants at risk, namely, small premature infants with invasive devices. Low birth weight and prolonged NICU stay are major determinants of CONS bacteremia (266). Intravenous lipids and intravascular catheters are significant independent risk factors (12, 187). It is suggested that lipids enhance the rate of bacterial growth in colonized catheters (187). Indolent bacteremia without focal findings is the most frequent manifestation of infection. Endocarditis, abscesses, omphalitis, surgical site infections, and meningitis occur occasionally, and a mild form of NEC has been described. Pulmonary infiltrates on chest radiograph are frequent with CONS bacteremia. CONS are also the major cause of ventricular shunt and drain infections (210). Freeman et al. (267) reported that CONS infection in the newborn was associated with extra hospital stay but not with excess mortality. Control measures involve limiting the use of invasive devices and aseptic technique for insertion and handling of intravascular and other prosthetic devices. Addition of vancomycin to intravenous fluids has been shown to reduce the incidence of CONS bacteremia in neonates but is not recommended for routine use because of potential for inducing vancomycin resistance (250). Transmission of CONS with heteroresistance to vancomycin in NICU has been reported (268). In view of increasing indications for invasive devices and the ubiquitous nature of CONS, infection rates probably will not be reduced until materials used for prosthetic devices are improved (104) (see also Chapter 30). Streptococci Group B Streptococci GBS is a leading cause of neonatal sepsis and meningitis in most centers in North America and Europe. The mother is the most important source of infection with this microorganism. Approximately 20% to 35% of pregnant women are colonized with GBS, and 50% to 75% of these women will transmit the microorganism to their newborns. Only 1% to 2% of colonized newborns develop disease. Early-onset disease usually presents on the first day of life with septic shock, pneumonia, and severe multiorgan failure. Risk factors include premature labor, premature rupture of membranes, multiple births, maternal bacteremia or bacteriuria, and low maternal levels of type-specific antibody to GBS capsular polysaccharide. Several serotypes are involved. Late-onset disease, occurring after 7 days of age, usually presents as meningitis, although other focal infections and bacteremia without an identified focus may occur. Lack of maternal antibody is a risk factor, but obstetric complications are not. Most late-onset infections and meningeal infections are due to serotype III. Late-onset disease may be due to microorganisms acquired from the mother at delivery or acquired postnatally (269). Transmission in the nursery is uncommon but has been reported, especially in crowded nurseries with a high rate of maternal colonization (270). The risk of early-onset GBS disease is reduced by maternal intrapartum antibiotic prophylaxis. The AAP in 1992 recommended screening of all pregnant women for GBS carriage and intrapartum prophylaxis for selected culture-positive women at high risk of delivering infected infants. Revisions in 1996 included alternative strategies based on screening and intrapartum treatment of all women with positive cultures or no screening and intrapartum treatment of all women with risk factors (271). Prophylaxis led to a 70% decline in the incidence of early-onset disease. However, screening was found to be more effective than the risk-based strategy (272). Current recommendations are to screen all women at 35 to 37 weeks' gestation and offer intrapartum prophylaxis to all with positive cultures except those undergoing planned cesarean section before rupture of membranes. All women with GBS bacteriuria during pregnancy or previous delivery of an infant with invasive GBS disease should be receive intrapartum treatment. If GBS status is not known at delivery, prophylaxis is indicated if gestation is less than 37 weeks, duration of amniotic membrane rupture is 18 hours or longer, or there is intrapartum fever (271). Active immunization of pregnant women is a promising approach if effective vaccines can be developed (269, 271) (see also Chapter 31). Enterococci Newborn enterococcal infections are occasionally acquired from the mother, but there is an increasing incidence of nursery-acquired infection in low-birth-weight infants (273, 274). NICU outbreaks of S. faecium (73) and S. faecalis (275) have been reported. Infants who have undergone bowel resection or CVL placement and those with prolonged hospitalization are at risk (274, 275). Bacteremia is often associated with focal infection such as NEC, soft tissue abscesses, pneumonia, and meningitis (273) and is often polymicrobial (274). Vancomycin-resistant enterococcus may cause widespread colonization and occasional invasive infection in the NICU and poses a therapeutic challenge (274, 276) (see also Chapter 32). Group A Streptococci Infection with group A Streptococcus, a prominent cause of newborn sepsis in the past, is now rare. Most infections were of nursery origin, introduced by the infant of an infected mother or by infected personnel and then transmitted from infant to infant. In normal newborns, infection may become manifest only after discharge (198, 277). Disease is usually mild, with most infants presenting with omphalitis, but sepsis and meningitis may occur (198, 199, 277, 278, 279). Administration of prophylactic penicillin to all exposed infants has been successful in terminating some outbreaks but not others (198, 199, 277). Cohorting (199) and application of bacitracin (277), triple dye (199), or chlorhexidine (278) to the umbilical stump have been used successfully in some outbreaks (see also Chapter 31). Other Gram-positive Bacteria Listeria is usually maternally acquired (280). Maternal infection is foodborne, and clusters of infection in newborns usually P.863 indicate community outbreaks. Early-onset disease, often associated with maternal symptoms, presents with pneumonia and rash and multisystem disease. Meningitis is the major form of late-onset disease. Control measures include advising pregnant women to avoid unpasteurized milk products and foods epidemiologically associated with an outbreak and diagnosing and treating infection in pregnancy. Nursery transmission is reported but rare. Contaminated resuscitation equipment (82) and mineral oil used to bathe infants (95) have been implicated (see Chapter 24). S. pneumoniae is an unusual cause of neonatal sepsis. Early-onset infection may be associated with maternal sepsis and has a poor prognosis (174). Nosocomial transmission has been reported (281). The role of C. difficile as a pathogen in the newborn is questionable, although it has been associated with outbreaks of NEC. Cultures are usually negative at birth. A high incidence of asymptomatic postnatal colonization has been detected in the normal nursery and in NICUs (159, 220, 229) (see Chapter 36). Enterobacteriaceae Members of Enterobacteriaceae include normal stool flora transmitted to the newborn at birth or acquired in the nursery. These microorganisms are common causes of nosocomial bacteremia, meningitis, pneumonia, and UTI. E. coli, Enterobacter, and Klebsiella are encountered most often (7). Transmission is usually person to person via hands, although contaminated patient care items have also been involved. The Enterobacteriaceae that are normally enteric pathogens are discussed in the section on gastrointestinal infections. E. coli, the second most frequent cause of sepsis and meningitis in the newborn, is usually of maternal origin. Newborns are particularly susceptible to strains bearing the K1 capsular antigen (173). Risk factors for infection are similar to those for GBS, although infants with E. coli infection tend to be of lower birth weight (17). Newborns may also acquire nursery strains of E. coli (51). Klebsiella may also be transmitted from the mother but is an unusual cause of sepsis in healthy newborns. It is an important cause of epidemic and endemic infections in the NICU and was the major cause of infection in many NICUs in the 1970s and 1980s. Newborns in NICU readily become colonized; the gastrointestinal tract is the major reservoir (49). Klebsiella survives well on skin and is more resistant to desiccation than other Enterobacteriaceae (282). Outbreaks have been associated with enteral feeding (141) and infusion therapy practices (283) and may result in ward closure (284).Control was achieved in one outbreak after introduction of alcohol hand rinses (285). Enterobacter species are frequent components of NICU flora and may now be more common than Klebsiella as causes of neonatal infections (7, 67, 286). Outbreaks have been associated with contaminated infant formula (130, 131) and intravenous fluids (105, 106, 111). Sepsis may be accompanied by meningitis and focal brain lesions (194). C. diversus infections in newborns have become increasingly prominent in recent years. This microorganism is part of the normal gastrointestinal flora and may be transmitted vertically to the newborn, occasionally causing serious infection (287, 288). Citrobacter strains may also be endemic in NICUs (5). Outbreaks affect normal newborns and those in NICUs (76, 201, 202, 287) and are characterized by large numbers of colonized infants with small numbers of symptomatic infants over extended periods (201, 202). Single strains differing from one hospital to another may be implicated (201), or several strains may be present in one outbreak, suggesting multiple introductions (287). Infection may be manifested as meningitis with focal brain lesions (203). Serratia, once considered a benign commensal, is now known to cause serious endemic and epidemic infection in NICUs. Outbreaks are characterized by widespread newborn gastrointestinal colonization with the infants serving as reservoirs (87, 213, 289). Contaminated intravenous fluids (109), delivery room equipment (87), breast pumps (124), and soap (80) have been implicated. Sepsis, meningitis, pneumonia, and UTIs are often associated with invasive procedures (87, 109, 146, 289), and low-birth-weight infants are especially at risk (80, 213). Conjunctivitis is often described (80, 87). Resistance to antibiotics is common and may arise during treatment (196) (see also Chapter 33). Other Gram-negative Bacilli P. aeruginosa was a major NICU pathogen in the past but has become less prominent. It is ubiquitous in the environment and proliferates in water (101). Outbreaks have been associated with contaminated equipment (81, 86). It readily colonizes skin and gastrointestinal and respiratory tracts, especially when antibiotics are used. P. aeruginosa causes sepsis and pneumonia (68, 81). Low-birth-weight infants are particularly at risk, and the case fatality rate is high (290). It is also an important cause of nosocomial conjunctivitis (86, 169) and the leading cause of neonatal endophthalmitis (170). Burkholderia cepacia (291), Ralstonia pickettii (292), and Stenotrophomonas maltophilia (103) may also be acquired from environmental water sources. Outbreaks of neonatal meningitis resulting from B. cepacia (195) and sepsis and meningitis with S. maltophilia (96, 293) have been described. Flavobacterium meningosepticum is another microorganism of environmental origin that occasionally causes nosocomial infections, usually meningitis, in newborns. Outbreaks of meningitis have been related to contaminated water sources (91, 93, 193). Acinetobacter calcoaceticus is often isolated from the hospital environment and skin and is usually considered a commensal microorganism. Outbreaks of bacteremia (70, 112, 294), meningitis, (197), and pneumonia (83) have been described. Low birth weight was a risk factor for infection. C. fetus is an uncommon cause of stillbirths, prematurity, and severe early-onset sepsis and is usually of maternal origin (295). An outbreak of nosocomial meningitis in an NICU has been described (204). C. jejuni has been responsible for an outbreak of nosocomial meningitis in a normal nursery (205). H. influenzae is found with increasing frequency in newborns with early-onset sepsis associated with prematurity and maternal complications. Infections are usually due to nonencapsulated nontypable strains (296). P.864 Pertussis is unusual in the newborn but when it occurs is often severe. Newborns may acquire the disease from visitors (297) or personnel (298) with atypical unrecognized infection (see also Chapters 34 and 81). Legionella infections have rarely been described in the newborn but sporadic cases occur. Infection has been linked to water used in an oxygen nebulizer and for heating feeding bottles and to postoperative contamination of a sternal incision with tap water (299). Pneumonia has been reported after water birth (300) (see also Chapters 35 and 55). Other Bacteria Colonization from vertical transmission of Ureaplasma urealyticum occurs in 22% to 58% of infants of colonized mothers, but neonatal disease is rare. U. urealyticum has been found in the lungs of premature infants with severe pneumonia (208, 209, 301). M. hominis is also transmitted vertically (209). Both microorganisms have been isolated from cerebrospinal fluid and from blood of ill newborns. Nosocomial transmission has not been reported. The infant of a mother with Chlamydia trachomatis infection has a 50% risk of becoming colonized. Conjunctivitis occurs in 25% to 50% of colonized infants, and 5% to 20% develop pneumonia. Infections in normal newborns become evident only after discharge from hospital. Transmission in the nursery has not been reported. Prevention involves diagnosis and treatment of the mother before delivery (208). Congenital or perinatal tuberculosis may be acquired from an infected mother. It is rare, and diagnosis may be delayed. Infants are unlikely to transmit infection by coughing but suctioning may generate infectious aerosols. Tuberculin skin test conversions have occurred in healthcare workers exposed to infected neonates (302, 303). Candida Newborns often acquire Candida at birth; 10% are colonized in the first 5 days of life. Many normal newborns develop oral thrush or diaper dermatitis, usually after discharge from hospital. Candida may also be acquired postnatally in the NICU (171). Common-source outbreaks have been associated with contaminated pressure transducers (154, 155), syringes (110), and glycerine suppositories (98). Systemic Candida infection, a disease of sick or VLBW infants, is increasing in frequency because of increased survival of infants at risk (171, 304). Antibiotic therapy suppresses normal flora and allows overgrowth of Candida, invasive procedures provide portals of entry, PN provides growth medium, and the immature defense system of the newborn permits invasion. In a multicenter study, risk factors for colonization included use of third-generation cephalosporins, central venous catheters, intravenous lipid, and H2-blockers (305). In one series, 27% of infants of birth weight less than 1,500 g became colonized; of those, 30% developed mucocutaneous disease and 8% developed systemic infection. Rates of systemic candidiasis in VLBW infants are 2% to 4%. Sick term infants requiring invasive procedures, especially after abdominal surgery, are also at risk. (171, 304). Duration of antibiotic therapy was the most strongly associated independent variable identified in a case-control study of risk factors for invasive candidiasis (306). Duration of hyperalimentation, lipid infusion, and endotracheal intubation were also significantly associated with infection. In another series, prolonged antibiotic therapy and endotracheal intubation were significant risks (162). The presence of CVLs and duration of use were less important than the infusate used (306), and in one series, most infected infants received PN through peripheral lines (307). Administration of intravenous hydrocortisone was associated with dissemination in another study (308). Endotracheal colonization and colonization in the first week of life identified VLBW infants at high risk to develop systemic disease (309). Petrolatum ointment skin care increased risk of invasive disease in neonates with birth weight of less than 1,000 g (310). Systemic disease is more frequent in newborns hospitalized for more than 4 weeks (17, 171). Infection may be limited to fungemia in association with intravenous lines or to the urinary tract in association with urinary catheters. Disseminated infection may involve the lungs, kidneys, CNS, eyes, gastrointestinal tract, or skin. Symptoms are often nonspecific and suggestive of bacterial sepsis with respiratory deterioration, hypotension, and gastric distension. Diagnosis is difficult and requires a high index of suspicion, because cultures may be negative despite extensive disease (171). Candida albicans and Candida parapsilosis are now the species most often involved in North American series, followed by Candida glabrata and Candida tropicalis (304, 305). C. parapsilosis is more often associated with vascular catheters and appears to be less invasive than C. albicans (304, 311). C. glabrata may be relatively resistant to fluconazole (312). Newborn infections with Candida lusitaniae, which may be resistant to amphotericin B, have been reported (313). Prevention of Candida infection in the NICU is a challenge. Antibiotic therapy is the major risk factor for invasion, yet it is often impossible to withhold empiric antibiotic therapy in sick premature infants when bacterial infection cannot be ruled out. Attempts to prevent invasive disease with prophylactic nystatin have been unsuccessful (162, 307). Administration of fluconazole prophylaxis was successful in terminating a prolonged NICU outbreak of C. parapsilosis (314) and reduced risk of invasive disease in VLBW babies (315). Fluconazole reduces but does not prevent colonization, and, although it may be beneficial in selected risk groups, widespread use may select for resistant Candida species (316) (see also Chapter 39). Other Fungi Yeasts Malassezia furfur, a dimorphic lipophilic yeast causing tinea versicolor in older children, is a cause of fungemia in the NICU (317). Cultures of infants in a normal nursery were negative, whereas skin (318) and rectal (319) colonization were frequent in the NICU. The increased humidity in the incubator and the moist macerated skin of the premature infant may enhance fungus P.865 replication. Skin colonization was correlated with younger gestational age, lower birth weight, and longer NICU stay (318). Rectal colonization was increased in infants receiving antibiotics (319). Fungemia occurred exclusively in infants receiving lipids intravenously; lipids serve as a growth factor (317). Symptoms were apnea, bradycardia, interstitial pneumonitis, and thrombocytopenia (317). Although most infants recovered with removal of intravenous lines, three infants with severe pneumonia were described, two of whom died (320). Isolation of the fungus is enhanced by the use of lipid-supplemented media (317). Malassezia pachydermatis, an animal pathogen, can also colonize infants in NICU and cause fungemia in association with intravenous administration of lipids. Disease is generally mild (321), but meningitis has been reported (322). One NICU outbreak was associated with healthcare workers' pet dogs (322). Clusters of invasive infections with Trichosporon asahii (beigelii), a yeast found in the soil and water, have been described in NICUs. Infection is often fatal (323). Outbreaks of Pichia anomala fungemia have recently been reported in NICUs in Brazil and India (324, 325). Filamentous Fungi Invasive infections with filamentous fungi are rare in the newborn, but sporadic cases of aspergillosis and zygomycosis have been reported (5, 326, 327, 328). They occur as a result of environmental contamination with dust containing fungal spores such as may occur during hospital renovation (327) or with faulty cleaning practices. Contamination of adhesive tape has been implicated (5, 326). An outbreak of Rhizopus infections was associated with contamination of wooden tongue depressors used to support intravascular cannulation sites (328). Spores may be inhaled or may infect puncture sites or wounds. Extreme prematurity, acidosis, renal failure, and treatment with steroids are risk factors. Infection usually progresses rapidly to death, and diagnosis is often made only at autopsy. Neonatal infection with dermatophytes is also rare. Nurses with unrecognized infections were the sources of two nursery outbreaks of Microsporum canis skin infection (329). In another outbreak, nurses were infected by contact with an infected newborn (330). Respiratory Viruses Nosocomial respiratory virus infections reflect virus activity in the community. Because of their incubation periods, these infections are rarely recognized in normal nurseries, but respiratory viruses, especially respiratory syncytial virus (RSV), are important causes of infection in the NICU. The newborn may acquire infection from the mother, other family members or hospital staff, or other infants in the nursery. Most respiratory viruses are spread by direct, indirect, and respiratory droplet contact (132). Newborns shed viruses for prolonged periods after symptoms cease, and their surroundings become contaminated. RSV, parainfluenza, and influenza viruses survive on hands or contaminated surfaces or equipment long enough to permit transfer between patients (132, 331). Hospital personnel often become infected and play an important role in nosocomial transmission (133). Newborns are likely to present with atypical features such as apnea, lethargy, and feeding difficulties, and pulmonary infiltrates on chest radiograph are common. RSV outbreaks are most common and may include large numbers of infants (133, 332, 333, 334, 335). In one outbreak, 35% of newborns in the NICU for more than 6 days and 34% of the staff were infected (133), whereas, in another outbreak, infections occurred in 84% of newborns in the NICU for more than 3 weeks (335). Clinical manifestations range from nonspecific symptoms or mild upper respiratory tract disease to severe respiratory compromise and death. Maternal antibody may decrease the risk or modify the severity of disease (133). Disease is more severe in premature infants (335). Concurrent outbreaks of RSV and rhinovirus (332) and parainfluenza 3 virus (334) have been described. Symptoms with rhinovirus and parainfluenza 3 virus infections were similar to those with RSV infection. Risk factors for transmission included contiguous bed space, nasogastric tubes, and tracheal intubation (332, 334). Outbreaks of parainfluenza 3 infection have occurred at times of crowding and understaffing (336, 337), with attack rates of 63% in infants and 25% in personnel (337). Nosocomial influenza infections in neonates have been described less often. Symptoms may be mild or may resemble bacterial sepsis. In two recent outbreaks, attack rates in neonates were 35% and 32% (338, 339). Most nursery personnel had not received influenza vaccine, and 16% were symptomatic in one outbreak (338). Nosocomial adenovirus infections may present as mild respiratory tract infections and conjunctivitis or as severe pneumonia, sepsis syndrome, and death (340, 341). Severe neonatal disease may result from symptomatic maternal infection, presumably resulting from lack of passive immunity in the newborn (340). An outbreak of conjunctivitis and pulmonary disease was associated with ophthalmologic examination (172). Nosocomial respiratory adenoviral infections may contribute to the development of bronchopulmonary dysplasia in premature infants (341) (see also Chapter 48). Enteroviruses Neonatal enteroviral infections are commonly associated with community outbreaks. Perinatal echovirus and Coxsackie B virus infections are most frequent (342). Infants are usually infected at birth. During a community outbreak, 3.4% of mothers were found to have enterovirus in the stool at delivery. Transmission rates from mother to infant of 29% and 57% have been reported. Horizontal transmission occurs in the nursery by fecal-oral contamination. Infants may be symptomatic within the first day of life, and outbreaks occur in normal newborn nurseries and NICUs with attack rates of 22% to 54%. Personnel may also become infected. Risk factors for infection include mouth care, gavage feeding, proximity to an infected child, and care by the same nurse (74, 343). Mild febrile illness and aseptic meningitis are the most frequent presentations, but disease may resemble bacterial sepsis. Severe hepatic necrosis may occur with echovirus infection. Coxsackie B viruses may cause myocarditis and, less often, hepatitis. Infection acquired from the mother tends to be most severe, probably because of lack of maternal antibody. Coxsackie A P.866 infections are rare in the newborn, although outbreaks of herpangina (344) and aseptic meningitis (345) have been reported. Rotavirus Rotavirus is endemic in many normal newborn nurseries and NICUs, and epidemics have often been described (220, 226). Unlike respiratory viruses and enteroviruses, nursery rotavirus infections tend to be unrelated to community outbreaks and are usually acquired from nursery rather than maternal sources. Infection rates vary considerably from 3.5% to 15% in endemic periods to 50% during nursery outbreaks. Transmission is fecal-oral via contaminated hands or equipment. The virus survives on environmental surfaces for prolonged periods (142). Many neonatal rotavirus infections are asymptomatic. Clinical disease occurs in up to 28% of those with positive stools and is often mild, although occasional severe disease with dehydration, bloody diarrhea, or NEC has been reported (220). Abdominal distension and bloody mucoid stools may be more prominent than watery diarrhea in premature infants (346). Enzymatic cleavage of the rotavirus outer capsid protein enhances infectivity in vitro. Immaturity of proteolytic enzymes in the newborn gut, the presence of antibody or trypsin inhibitors in breast milk, or immaturity of receptor sites on enterocytes have been suggested as explanations for the paucity of symptoms in newborns (220, 226). An alternative explanation may be attenuation of endemic nursery strains (226) (see also Chapter 50). Hepatitis A HAV may be introduced into the nursery by an infant infected by blood transfusion (114) (see Chapter 69) or by vertical transmission from a mother with acute infection (347). The virus is readily transmitted to other infants, staff members, and parents by fecal-oral contact. In one outbreak, 20% of exposed infants and 24% of exposed nurses became infected. Newborns excreted the virus for prolonged periods of up to 4 to 5 months (114). Because infection is usually asymptomatic in the newborn, outbreaks may be recognized late, only after staff members or parents become ill. Infection has spread to other nurseries when infants were transferred (348). Administration of immune globulin to all exposed contacts may be necessary to bring the outbreak to a halt (114) (see also Chapter 45). Herpes Simplex Virus Neonatal HSV infections are usually acquired from the mother. The infant is especially at risk if the mother has primary HSV in pregnancy. Approximately 33% to 50% of mothers with primary genital lesions transmit infection to the infant perinatally. These infants are exposed to a high virus load and lack protective maternal antibody. Risk of transmission from mothers with reactivation at delivery is low, at less than 2% to 5% (208, 349). Scalp monitors may increase the risk of infection by providing a portal of entry for the virus (138, 139). Infected infants may present with mucocutaneous lesions, encephalitis with or without mucocutaneous lesions, or disseminated multisystem disease. Despite antiviral therapy, morbidity and mortality remain significant (350). Management of HSV infection in pregnancy has been a controversial issue. Cesarian section reduces risk of transmission (351). Current recommendations are to assess the parturient for lesions at the onset of labor and consider cesarean section if lesions are present (349). The risks and benefits of cesarean section should be considered if membranes have been ruptured for more than 6 hours (208). Scalp monitors should not be used. An infant exposed during vaginal delivery should have specimens obtained for culture at 24 to 48 hours of life and be carefully observed for signs compatible with HSV infection. Prophylactic acyclovir may be indicated for exposed high-risk infants (208, 349). Acyclovir suppression in late pregnancy reduces the rate of recurrent infection at delivery and may have a role in reducing the need for cesarean section (349, 350, 352). Transmission of HSV in the nursery is rare, but small outbreaks have been described (353, 354). Infection has been transmitted to an infant suctioned by a healthcare worker with orolabial lesions (355) (see also Chapter 43). Varicella-Zoster Virus Varicella is readily transmitted by the airborne route, and transmission may occur before onset of the rash. Nevertheless, varicella is rare in the newborn nursery, because most adults are immune and most infants are protected by maternal antibody. The newborn is at risk for severe perinatal disease acquired from a mother who has onset of varicella lesions from 5 days before to 2 days after delivery, presumably because virus but no antibody is transmitted to the fetus (208, 349). Prophylactic varicella-zoster immune globulin (VZIG) is recommended for these newborns. Varicella may be introduced into the nursery by mothers, employees, or visitors with unrecognized infection (356, 357) or by an infant with perinatal varicella (358). Hospitalized premature infants of seronegative mothers and those born before 28 weeks of gestation are at risk for more severe disease and should receive prophylactic VZIG after postnatal exposure (208, 349). Severe varicella may occur despite VZIG prophylaxis; acyclovir prophylaxis may be more effective and warrants further evaluation (359) (see also Chapter 42). Cytomegalovirus Intrauterine and perinatal CMV infection is common and usually benign. Approximately 1% of all newborns are infected in utero. In most cases, the mother has reactivation of a past CMV infection, and the infant is protected from severe illness by passive transfer of antibody. Severe disease occurs when the mother has primary CMV infection in pregnancy. Perinatal CMV acquisition during delivery or postnatal acquisition from breast milk is usually asymptomatic but may present as mild self-limited pneumonia or hepatitis at 1 to 4 months of age (360). Premature infants of birth weight less than 1,500 g who receive little maternal antibody may have severe disease (361, 362). Transfusion-acquired CMV infection has been a problem in NICUs, especially for low-birth-weight infants of seronegative P.867 mothers (113) (see Chapter 69). A sepsis-like syndrome with hepatomegaly, respiratory deterioration, and atypical lymphocytosis was described in low-birth-weight infants. Some infants died with severe multisystem involvement (113). Transmission of CMV between infants in the NICU has been reported (363) but is extremely rare (364). The evidence suggests that nursery personnel are not at increased risk of acquisition of CMV despite widespread concern (113). Transmission of CMV requires direct inoculation of mucous membranes with fresh secretions and can be prevented by the normal nursery routine of hand washing after handling respiratory secretions or diapers (365). For the protection of both neonates and personnel, personnel should not kiss the newborns in their care (see also Chapter 44). Human Immunodeficiency Virus The usual source of neonatal HIV infection is the mother. Transfusion-acquired infection occurred in newborns before routine screening of blood donors for HIV (117). Vertical transmission may occur in utero, at delivery, or postnatally through breast milk. Zidovudine treatment during pregnancy and delivery followed by postnatal treatment of the newborn reduced the rate of HIV transmission from 25% to 8% (349). More aggressive measures such as combination antiretroviral therapy, monitoring of viral suppression, and cesarean section for women with elevated viral load at delivery reduce the risk of transmission to less than 2 % (366). In countries with resources for antiretroviral treatment, screening for HIV should be part of routine prenatal care, and all pregnant women with HIV infection should be offered antiretroviral therapy. It is recommended that the HIV-infected mother should not breast-feed if a safer alternative source of feeding is available (208, 349, 366). Hepatitis B and C The newborn may acquire HBV from a mother with chronic or acute infection in pregnancy. Infection is almost always asymptomatic in the infant, but neonatal acquisition carries a high risk of lifetime chronic infection and consequent serious liver disease in adulthood. Transmission usually occurs at delivery, and administration of hepatitis B hyperimmune globulin and vaccine at birth prevents infection in the newborn (208, 349). Transfusion-acquired HBV infection in the newborn has rarely been reported (115), even in the era before universal screening of blood donors, probably because infections were asymptomatic and not recognized (see also Chapter 45). Approximately 5% of infants of seropositive mothers acquire HCV vertically. Whether transmission occurs in utero or intrapartum is not known. Breast-feeding does not appear to increase risk of transmission. Transmission risk correlates with maternal serum virus concentration and is higher in women co-infected with HIV. Infection in the newborn is usually asymptomatic but chronic. The AAP recommends screening of infants born to HCV-infected mothers and mothers at high risk. There is no prophylaxis available at present (208, 349). Neonates have also acquired HCV from transfusions before universal screening of blood (116) PREVENTION AND CONTROL OF INFECTIONS Policies for prevention of infection in nurseries have evolved over the years from a combination of custom, common sense, and epidemiologic study. Many procedures that have been considered important are little more than rituals (5, 367) carried out because of tradition and have little bearing on transmission of infection. They were based on the assumption that the sources of infection in the nursery are extrinsic and that preventing entry of microorganisms from outside will keep the nursery infection-free. The available evidence indicates that the initial source for microbial colonization of the newborn is the mother and that, subsequently, within the nursery the infants themselves are the major reservoirs. Transmission of microbes between infants is not affected by rituals performed at entry to the nursery. Thus, it makes little sense to perform surgical scrubs and don gowns or other protective apparel at entry to the nursery if similar precautions are not taken between infants. The emphasis is shifting from consideration of the nursery as a unit to consideration of each infant as a potential source and recipient of microorganisms. The AAP states that “Neonates should be approached as though they harbored colonies of unique flora that should not be transmitted to any other neonate” (365). Infection control strategies can have an impact on NICU endemic infection rates. A multidisciplinary quality improvement model was developed by the Vermont Oxford Perinatal Network and implemented in selected NICUs. Rates of CONS bacteremia decreased in the project NICUs in comparison with control NICUs (368). In an NICU in Argentina, implementation of locally developed guidelines resulted in a decrease in bacteremia rate from 20.0 to 7.7 per 1,000 patient-days (369). Nursery Design and Personnel Design Nursery design should provide adequate space for appropriate care of the infant and for the necessary equipment and sufficient numbers of strategically placed sinks. Specific recommendations have been published by the AAP in collaboration with the American College of Obstetricians and Gynecologists (ACOG) (370) and by others (371, 372). A space of 30 net square feet per neonate with at least 3 feet between bassinets is recommended in the normal newborn nursery. For continuing care of low-birth-weight infants who are not ill but require more nursing hours than term infants, 50 square feet per infant with 4 feet between bassinets is recommended. Intermediate-care nurseries should have 100 to 120 square feet per patient station if subspecialty care is required, with at least 4 feet between incubators or bassinets and 5-foot-wide aisles. NICUs should have 150 square feet per infant with at least 6 feet between incubators and 8-foot aisles (370). There should be one sink for every six to eight patients in the normal newborn nursery and one sink for every three or four patients in intermediate- and intensive-care nurseries (370). Central ventilation should be equipped with filters with efficiency of at least 90% (370). The AAP-ACOG guidelines recommend a minimum of six air exchanges per hour (370). A higher P.868 rate of 10 to 15 air exchanges per hour has been suggested (5). Canadian guidelines recommend positive pressure airflow with air passing from a ceiling entry to a floor exhaust, pulling dust downward and out (371). Each nursery should have access to at least one negative pressure isolation room, with exhaust air vented to the outside, to accommodate newborns with airborne infections (365, 371, 372). During construction or renovation, including dust-generating activities such as breaking into or drilling holes in walls or ceilings or removal of ceiling tiles, newborns and patient care equipment should be protected from exposure to dust and debris that may contain fungal spores (327, 371). Newborns should be moved to a separate hospital area unless impermeable barriers can be set up to prevent influx of air into the nursery from the construction zone. With the trend toward encouraging early contacts between newborns and their families, rooming-in programs for healthy newborns and early discharge from the hospital have become prevalent in North America (370, 371). Both practices reduce the risk of exposure of the newborn to flora of other infants. Studies suggest that infection rates are lower for infants rooming with their mothers than for those in communal nurseries (373, 374). Staffing Staffing should be sufficient to allow for adequate care of infants with sufficient time for hand washing between patient contacts. For normal newborn nurseries, recommendations are one nurse for every six to eight infants or for every three to four mother-infant pairs. A ratio of one nurse for every two to three patients is recommended in intermediate-care units and of one nurse for every one to two patients in NICUs (370, 371). Employee Health Personnel should be immune to rubella, measles, mumps, varicella, HBV (375), and polio (208) and should receive influenza vaccine annually (365, 375). Tuberculin reactivity should be determined on employment and periodically (365, 375). Acellular pertussis vaccine should be considered if an adult formulation is available (376). It is important that employees understand the risks of transmission of contagious diseases to newborns and report acute infections for assessment. It is rarely feasible to remove all persons with communicable infection from the nursery, and decisions should be made on an individual basis, taking into consideration the mode of transmission of the particular infection and the ability of the employee to comply with preventive measures. Employees with exudative or herpetic hand lesions should not have direct patient contact or handle patient care equipment. Personnel with herpes labialis are unlikely to transmit infection but should avoid touching the lesions during patient care and cover any external lesions (365). Nonimmune persons exposed to varicella, measles, or rubella should not work during the latter part of the incubation period, because these diseases may be transmitted for a few days before eruption of the rash (208) (see Chapter 99). Personnel should take precautions to minimize the risk of potential infection with blood-borne viruses and should be familiar with hospital protocols for postexposure prophylaxis after occupational exposures to blood (365) (see also Chapters 78 and 79). Routine Procedures Hand Washing Hand washing has been recognized as an important means of prevention of transmission of infection since the experiments of Semmelweiss (75, 377); however, it is difficult to monitor or enforce, and studies show poor compliance with this procedure in the NICU (367, 378, 379, 380) and in other hospital areas (367). It is recommended that, before handling neonates for the first time on a work shift, personnel wash hands and arms to above the elbows, with care to clean all parts of the hands and beneath the nails. Watches, rings, and bracelets should be removed. Nails should be trimmed short, and no false fingernails should be worn (365, 381). The optimal duration of hand washing has not been established, but sufficient time should be taken to thoroughly wash and rinse all parts of the hands. Two- (371) or 3-minute (365) scrubs have been suggested. However, performance of a prolonged scrub on entry to the nursery is likely of less benefit than careful hand washing between patients (5). A hand wash of 10 to 15 seconds before and after patient contacts should be sufficient to remove transient flora unless hands are heavily contaminated. Whether this should be with soap or antiseptic is controversial. Some recommend an antiseptic hand wash for certain specified activities only (365), and others recommend antiseptic hand wash for all hand washing in the nursery (5, 371). Antiseptic agents commonly used for hand washing are chlorhexidine and iodophors. Chlorhexidine has the advantage of being tightly bound to skin and leaving a residual antibacterial effect. It is less irritating than iodophors. In outbreaks of S. aureus infection, the more potent antistaphylococcal agent hexachlorophene may be indicated; however, it is less effective against gram-negative microorganisms than the former agents and, therefore, should not be used routinely. (365). Triclosan and chloroxylenol are more recently introduced hand-washing agents and less is known about their effectiveness; further evaluation is needed (381). Alcohol-based antimicrobial hand rinses are as effective as soap or water-based antiseptics, are well tolerated and convenient to use, and are especially useful when water is not available or access to sinks is limited. Hand rinses may be less effective on heavily soiled skin (381) (see also Chapter 96). Special Attire The requirement for special attire for persons entering the nursery has caused much controversy and needless expenditure. Unlike hand washing, this ritual is easily monitored and enforced, and compliance is good (367). Unfortunately, the donning of special attire often replaces hand washing as the entrance ritual to the nursery, and, once garbed, personnel may consider themselves to be incapable of transmitting infection. Over the P.869 years, routine use of caps, masks, and beard bags has been dropped. Most nurseries provide scrub suits or dresses that are laundered by the hospital for personnel spending most of the day in the nursery. This practice may prevent soiling of personal clothing and may be reassuring to parents in providing easy recognition of personnel but should not be considered an infection control measure. The routine use of gowns by persons entering the nursery is of no proven value in infection control. A gown protects the infant from contact with the wearer's forearms and clothing. In practice, most contact occurs via the hands of personnel and is not prevented by use of gowns. Several studies have shown that gowns have no effect on colonization or infection rates in the newborn nursery or the NICU (379, 382, 383, 384, 385). In one report, there was a decrease in NEC during the period of gown use but no effect on other infections (386). Because NEC often occurs in clusters, this may have been coincidental. Although it is sometimes argued that gowns serve as reminders for hand washing, this has not been shown to be so. Studies in NICU showed no difference in hand washing frequency or in traffic through the NICU when gowns were or were not used (379, 385). Current recommendations are for a long-sleeved gown to be worn by personnel holding newborns outside of the bassinet or incubator. A separate gown should be used for each infant and discarded after use or maintained exclusively for the care of that neonate and changed regularly (365), such as at the end of a shift or if wet or soiled (387). Visitors The advantages to the family of allowing siblings to visit newborns has been stressed (388). Limited data suggest that neonatal colonization and infection are not increased with such visits (389, 390). However, introduction of highly contagious diseases such as varicella, pertussis, or RSV into a nursery has the potential for serious results. In the normal newborn nursery, visiting in the mother's room or a special visiting room reduces the exposure of other newborns. Visiting in the NICU requires clearly defined policies. Visitors should be screened for infection and individually assessed as to potential for transmission of infection and ability to comply with instructions. The visitor should not have been recently exposed to varicella or measles (unless already immune); should not have fever or acute respiratory, gastrointestinal, or skin infection; and should be prepared in advance for the visit. The visiting child's hands should be washed before contact with the newborn, and parents should ensure adult supervision of the child during the visit. The visiting child should not have contact with newborns other than the sibling and should not handle patient care equipment (371, 388, 391). Limiting the number of visitors per visiting period and the duration of visits is advisable. It may be prudent to restrict visiting during community epidemics of respiratory tract infection (333). Decontamination and Cleaning Examples of outbreaks of infection related to contaminated equipment were given earlier. It is important that all equipment in direct contact with skin or mucous membranes of newborns be decontaminated with a high-level disinfectant or sterilized between patients. Examining equipment such as stethoscopes should be reserved for use with one patient or decontaminated with alcohol or iodophor between uses (365, 371). Respiratory support equipment such as resuscitation bags, masks, and laryngoscope blades should be in sufficient supply to permit decontamination between patients. The nursery should be kept clean and dust free by daily cleaning using cleaning methods that minimize dust dispersal. Quaternary ammonium, chlorine, and phenolic compounds are satisfactory low-level disinfectants for nursery cleaning (365). These do not sterilize but reduce the concentration of microbes to an acceptable level. Phenolic compounds should be used with caution, because inappropriate use has been associated with absorption by the newborn resulting in hyperbilirubinemia (392) (see also Chapter 85). It is important that NICU equipment be kept dust free, because fungal spores from dust may result in serious infections. Responsibility for the cleaning of delicate equipment, especially monitoring equipment, radiant heaters, or infant care units in constant use, must be clearly assigned, because these items are often not handled by the regular cleaning personnel (5). Incubators, open care units, and bassinets should be cleaned between infants and changed and cleaned periodically for those infants with prolonged stay (365). Humidifier reservoirs in incubators are potential sources of Pseudomonas, Legionella, and other water-borne microorganisms and should not be used in nurseries in which central humidification provides sufficient humidity. If used, the reservoir should be drained, cleaned, and refilled with sterile water every 24 hours. Nebulizers and attached tubing and water traps should be replaced regularly with equipment that is sterile or has undergone high-level disinfection. Sterile water should be used in nebulizers and humidifiers (365, 371). Ventilator tubing, bags, and masks should be replaced and decontaminated according to hospital protocol periodically and between patients (see Chapter 68). Linen for newborns does not need to be autoclaved. There is no evidence of infection related to linen, and cultures of NICU linen yielded small numbers of normal skin flora only (393). Clean linen should be stored in closed cabinets to prevent dust contamination. Used linen should be handled as little as possible to avoid hand contamination and aerosolization of microorganisms (365). Skin and Cord Care Once the newborn's temperature has stabilized, blood and meconium should be removed with sterile cotton sponges and warm water. If soap is used it should be supplied in a single-use container or reserved for use with one infant. Because of potential exposure to blood-borne viruses, personnel should wear gloves when handling the neonate until this has been done (365). Localized skin care using warm water and a mild soap for the diaper area and other soiled areas may be sufficient throughout the nursery stay (388, 371, 394). Whole-body bathing and antiseptic agents are not necessary for routine newborn care but may be indicated in outbreaks. P.870 Antiseptic agents should be used only if the benefit outweighs the risk of toxicity. When nursery S. aureus infections were rampant, it was common practice to bathe newborns daily with hexachlorophene. This practice was discontinued when reports of neurotoxicity in premature infants appeared (254, 395). Bathing with hexachlorophene reduced S. aureus colonization and infection, and many nurseries reported increasing infection rates when it was discontinued (255, 256, 396). Hexachlorophene is useful in control of S. aureus outbreaks but should be used only for term infants, for no more than two baths, and at a concentration of no more than 3%; hexachlorophene should be carefully rinsed off (250). A safer alternative, although possibly less effective against S. aureus, is chlorhexidine. It is less toxic, and unlike hexachlorophene, cutaneous absorption is negligible (381, 394). No absorption was detected when 4% chlorhexidine solution was used to bathe term infants (397) or for cord care (398). Minimal blood levels were detected in some premature infants bathed with 4% solution (399) and when 1% chlorhexidine in ethanol was used for cord care (400). No significant toxicity has been reported. Iodophors may not be safe for bathing newborns because of absorption of iodine, and isopropyl alcohol has caused skin necrosis in premature infants (394). Triclosan has been used to bathe neonates (263) but should be used with caution because there is little information about safety and efficacy (381). Care should be taken to avoid damage to the newborn skin from excessive drying, manipulation, exposure to irritating chemicals, or other trauma (388, 394, 401). The skin of the premature infant is especially fragile, and minor trauma such as removal of adhesive tapes or oxygen probes may remove the outer layer of the epidermis (134, 394, 401). Application of topical ointment to the skin of VLBW infants decreased skin damage, intensity of skin colonization, and nosocomial bacteremia in one study (402). To prevent contamination of the product, unit dose containers are advised (394). The cord should be cut and tied under aseptic conditions. Subsequent procedures for cord care vary, and none is clearly superior. Local applications of triple dye or bacitracin delay or reduce cord colonization in comparison to dry cord care. Alcohol hastens drying but may not affect colonization (388, 371, 403). Triple dye delayed, but did not prevent, eventual MRSA colonization in NICU patients in one report (259). Chlorhexidine cord care has been effective in reducing colonization and infection (404) and may be the most effective product with the least toxicity (394). Any agent used should be provided in single-dose containers or reserved for use with a single patient. Eye Care The eyes of the neonate should be cleaned with sterile cotton to remove secretions and debris. Topical prophylaxis against gonococcal eye infection has been routine for many years; the agents used are 1% silver nitrate drops, 0.5% erythromycin or 1% tetracycline ointment, or 2.5% povidone-iodine solution (249, 388). Single-dose containers should be provided. Topical prophylaxis appears to be ineffective against neonatal Chlamydia conjunctivitis. Diagnosis and treatment of Chlamydia infection in the mother before delivery is a more effective strategy (249). Nosocomial conjunctivitis is a frequent occurrence in the NICU. Eyes may become infected with water-borne microorganisms or from infected respiratory tract secretions. Care should be taken to avoid contaminating the eyes with secretions dripped from catheters used to suction the nasopharynx or endotracheal tube (169). Infant Feeding Natural breast-feeding is the optimal method of infant feeding. Human milk provides immunologic and nutritional benefits and has been reported to reduce the risk of sepsis in premature infants (405, 406). When the sick newborn cannot suck, the mother may express and store breast milk. This should be done aseptically to minimize bacterial contamination. Hands should be washed with an antiseptic, and milk should be expressed into sterile containers. If a breast pump is used, all pump components in contact with milk should be washed with hot soapy water after each use (388) and sterilized or disinfected daily. Expressed milk stored in the refrigerator showed no significant growth in 48 hours (407). It is recommended that milk be stored in the refrigerator for a maximum of 48 hours or frozen for up to 6 months. Frozen milk should be thawed in the refrigerator or quickly under running or fresh warm water, because standing water may become contaminated. Milk should not be subjected to excessive heat from hot water or a microwave oven. After thawing, milk should be used promptly or stored in the refrigerator for up to 24 hours (388, 408). When breast milk is stored in hospital, protocols should be established to ensure proper identification to prevent infants inadvertently being fed milk from mothers other than their own (409). Routine microbiologic monitoring of expressed milk is not recommended (388), but screening may be advisable if there is concern about collection technique or if gastrointestinal intolerance or sepsis is suspected. The presence of gram-negative bacilli suggests contamination during collection. (122, 371). Human milk banking is an established practice in many countries (120, 405), but concerns over transmission of infection have led to a decline in this practice (388). Milk donors require careful screening (371, 388, 391, 405). They should be (a) able to carry out aseptic technique, (b) healthy and without acute or chronic infections, and (c) screened for use of drugs and medications and other factors that might impair the quality of their milk. Donors should be serologically negative for hepatitis B surface antigen (HBsAg), HIV-1, HIV-2, HCV, HTLV-I, HTLV-II, and syphilis (391) and should not have active untreated tuberculosis (388, 405). Screening for CMV has also been recommended (371) but may not be indicated if milk is heated to 62.5ВC (391). To ensure microbiologic safety, especially against blood-borne viruses, it is recommended that all donor milk be pasteurized at 56ВC or 62.5ВC for 30 minutes, even though this process results in loss of some immunologic factors (120, 371, 391, 388, 405). The higher temperature results in more destruction of protective components (405) but may be more effective in inactivation of CMV (391). Contamination may occur after pasteurization if appropriate care is not taken during handling (126, 127). The Human Milk Banking Association of North America recommends that donor milk should be used only if it contains no pathogenic bacteria and less than 104 nonpathogenic P.871 bacteria per milliliter (391, 405). The Canadian Paediatric Society does not recommend the use of donor milk, considering that the potential risks and costs outweigh the benefits (408). Most hospitals in North America use sterile commercial formula prepared ready to feed. This should be used within 4 hours of uncapping (388). Commercial liquid concentrates are sterile. Powdered formulas are not sterile but undergo microbiologic testing to ensure that the level of contamination meets safety requirements. These should be used only if there is no alternative (131). Formula made from liquid concentrates or powders must be prepared using aseptic technique. Detailed guidelines for aseptic preparation of infant formula have been published by the American Dietetic Association (410). Water should be sterile or boiled for 5 minutes. Utensils and containers should be sterilized or undergo decontamination to remove vegetative forms of microorganisms. Boiling for 5 minutes and cooling before use may suffice. Blenders should be cleaned after each use and sterilized daily (129). Formula should be bottled in quantities for individual feeds or for 4 hours continuous feeding, refrigerated for a maximum of 24 hours, and used within 4 hours of opening. Routine microbiologic testing is not recommended but may be indicated if problems occur. Nasogastric feeding administration sets should be changed every 24 hours (371). Continuous infusion tube feeding should be set up with the same aseptic precautions as used for intravenous fluids. Syringes and tubing used for continuous feeding should be changed at 4 to 6 hours (388) because bacteria may multiply to high levels in small volumes held at room temperature (122). Special Procedures Invasive Devices Technologic advancements in neonatal care have given rise to new and sometimes unexpected infection risks. As a general principle, whenever a new invasive procedure or device is introduced, the potential risk for nosocomial infection should be considered, protocols established to minimize this risk, and surveillance set up to monitor for infection. The need for any invasive device should be assessed daily, and use should be discontinued promptly when no longer essential. Infection control recommendations for the insertion and maintenance of intravascular catheters, endotracheal tubes, and urinary catheters in the newborn are not different from those in other patients and are discussed elsewhere in this book and not repeated here (see Chapters 17, 18, 19, and 20, 22, and 49). Umbilical vessel catheters warrant special mention. Although now replaced by central lines for infants requiring long-term vascular access, they are still often used in initial management of the sick newborn. These lines should be inserted and maintained with aseptic technique, but the nonsterile insertion site and devitalized cord tissue increase the potential for colonization. It is recommended that umbilical arterial catheters be removed before 5 days and venous catheters before 14 days of use (411). A chlorhexidine-impregnated dressing reduced colonization of intravascular catheters in neonates. Local skin reactions occurred in some low-birth-weight infants (412), precluding use of the dressing for infants younger than 7 days, especially if gestational age is less than 26 weeks (411). Central vascular catheters coated with antiseptics or antibiotics have a lower risk of infection but have not been evaluated in neonates. Blood Indications for transfusions of newborns have become more stringent in recent years; nevertheless, many premature or ill newborns receive blood products (413, 414). In many countries, blood donors are routinely screened for HBV, HIV-1, HIV-2, HTLV-I, HTLV-II, HCV, and syphilis (391, 414). All cellular blood products given to low-birth-weight infants should be from CMV-seronegative donors or treated to remove CMV (113, 391, 414). Some centers use these products for all newborns. Fatal disseminated CMV infection was reported in a term infant of a CMV-seronegative mother after exposure to large volumes of unscreened blood and blood products during extracorporeal membrane oxygenation (415). If CMV-negative blood is not used for all newborns, it should be considered for seronegative infants receiving large volumes of blood and others at elevated risk of CMV disease (see also Chapter 44). Surveillance Surveillance for nosocomial infections in nurseries permits early detection of infection trends and clusters and identification of new risks, provides information on which to base empiric antibiotic therapy, and is a measure of quality of care. The intensity of surveillance varies with the type of nursery and the facilities available. Development of a surveillance program involves selection of definitions and determination of the types of infections to be monitored, the methods of case finding, and the denominator data to be collected (births, admissions, patient-days, birth weights, device-days). For the NICU, infections at all sites should be monitored if this is feasible (7). If resources are limited, targeted surveillance with collection of appropriate denominators such as birth weight and device use will be more useful than total surveillance without relevant denominators. Surveillance in the normal newborn nursery should concentrate on infections likely to be associated with nursery outbreaks such as staphylococcal or streptococcal skin infections, gastroenteritis, and viral infections. Because many of these infections manifest only after discharge, coordination with community healthcare providers is essential. Routine surveillance cultures are generally not recommended, because colonization is not a good positive predictor of infection and correlation of isolates from surveillance cultures and invasive infections has been poor (215, 416, 417, 418). In outbreaks, surveillance cultures may be indicated to identify colonized infants for purposes of cohorting or isolation or for assessment of risk factors for acquisition of the microorganism (5, 365, 371). (For more details on the surveillance of nosocomial infections, see Chapter 94.) Isolation Procedures Revised guidelines for isolation precautions for hospitalized patients were published by CDC in 1996 (419) and Health P.872 Canada in 1999 (387). CDC Standard Precautions refer to barrier precautions to be taken with all patients to reduce transmission from recognized and unrecognized sources. This principle embodies the concept that the flora of neonatal patients should not be shared. Standard Precautions include (a) hand washing between patient contacts and after removal of gloves; (b) use of gloves for touching the patient's mucous membranes or nonintact skin and for all contact with blood, body fluids, secretions, excretions, and contaminated items; (c) removal of gloves promptly after use and before going to another patient; (d) use of masks, protective eyewear, and gowns to protect the healthcare worker's mucous membranes and uncovered skin during procedures that are likely to generate splashes or sprays of body substances; (e) care in handling patient care equipment and linen to avoid contamination of skin and mucous membranes; (f) provision of resuscitation bags, mouthpieces, and mechanical suctioning equipment to eliminate the need for emergency procedures involving oral suctioning; and (g) taking precautions to reduce the risk of injury from needles and other sharp instruments. Standard Precautions should be sufficient to prevent transmission of most infections encountered in nurseries. Extensive use of gloves may not be indicated if appropriate hand washing is performed. The AAP suggests that gloving is not mandatory for diapering of infants (391). Additional transmission-based precautions (Airborne, Droplet, or Contact) are recommended for certain clinical conditions and infectious agents based on modes of transmission of the microbes known or suspected to be involved (387, 419). Prevention of airborne transmission, such as occurs with varicella, measles, or tuberculosis, requires a single room with negative pressure ventilation. The infant of a mother with perinatal varicella or measles requires similar isolation. Fortunately, these infections are rare in the nursery. Forced-air incubators cannot be substituted for isolation rooms, because they discharge unfiltered air into the nursery (391). High-efficiency filtration masks should be worn by nonimmune healthcare workers who must enter the room. Droplet Precautions require that a mask is worn when within 3 feet of the patient. For Contact Precautions, gloves are recommended for entry into the room or the patient's designated bed space in a shared room and gowns for substantial contact with the patient, environmental surfaces, or items in the room. Although single rooms are recommended with Droplet and Contact Precautions, they are not mandatory and may be inadequate for the care of the critically ill child needing close observation. Newborns are nonmobile, so transmission by direct contact is not a problem, and they do not grossly contaminate their environments. Separate isolation rooms are not considered to be necessary for newborns if the following conditions are met: (a) the infection is not transmitted by the airborne route, (b) there is sufficient space for a 4- to 6-foot aisle between infant stations, (c) there are an adequate number of nursing and medical personnel and they have sufficient time for hand washing, (d) an adequate number of sinks are available for hand washing, and (e) continuing instruction is given to personnel about the ways that infections are spread (365, 391). Isolation requirements are determined by the number of infected or colonized newborns in the nursery and the care required by the newborn. In the normal newborn nursery, the most feasible policy may be to isolate the occasional newborn with gastroenteritis, respiratory tract infection, or infectious skin lesions in a single room or to have the newborn room with the mother. When multiple cases of infection occur, as is common during community outbreaks of viral infection, cohorting in a communal nursery is more feasible than use of isolation rooms. An isolation area can be defined in the nursery or NICU by curtains, partitions, or other markers. A closed incubator may be helpful in maintaining barrier precautions, but because incubator surfaces and entry ports readily become contaminated with the microorganisms carried by the infant, the outside of the incubator should be considered contaminated and the boundaries of the isolation area should extend beyond the incubator itself. Where droplet precautions are required, infected infants should be separated from other patients by a distance sufficient to prevent transmission of large droplets (at least 3 feet). Newborns are unlikely to generate large-droplet aerosols, but aerosolization may be a problem with infected infants on respirators. Respiratory viral infections are a significant problem in NICU. Most respiratory viral infections are spread by large droplets and by contact with respiratory secretions. Healthcare workers are at high risk for nosocomial respiratory viral infections (133, 336, 337, 338) and should take precautions to prevent inoculation of their eyes or mucous membranes with infectious respiratory secretions. Gloves reduce the risk of accidental inoculation (420). Masks alone are of little value, but goggles and face shields may give added protection against inoculation (421). The Infected Mother Infection control precautions for the mother with peripartum communicable infection are similar to those for other patients (387, 419). Transmission of maternal infection to the newborn usually occurs during delivery, and postpartum separation of mother and newborn is rarely necessary. Most maternal postpartum infections are urinary or gynecologic from endogenous flora. The mother with a communicable infection should wash her hands before handling the infant and take measures to prevent contact of the infant with potentially contaminated clothing, bedclothes, tissues, and other fomites (365). Untreated active pulmonary tuberculosis in the mother is an indication for separation of the newborn until the mother is considered noninfectious (see Chapter 37). The newborn who has received VZIG may remain with the mother (349). Separation should be considered if a mother has extensive S. aureus infection with drainage not contained by dressings (371) or if a mother has a group A Streptococcus infection until she has received antibiotic therapy and the infection is no longer communicable (365). Mothers with HIV infection should not breast-feed. Those with active untreated tuberculosis should not breast-feed until they have received adequate therapy. Otherwise, breast-feeding by the infected mother is rarely dangerous for her infant. HSV lesions around the nipples are a contraindication. AAP suggests that mothers with antibody to HTLV-I or HTLV-II should not breast-feed, pending further knowledge about transmission of P.873 these agents. For mothers with primary CMV infection who are seronegative at delivery and CMV-seropositive mothers who deliver VLBW infants, potential risks and benefits of breast-feeding should be weighed; pasteurization of milk may be considered (362, 388, 391). Antibiotic therapy per se is not a contraindication but depends on the choice of antibiotic, because many are harmless to the newborn or are excreted in minimal amounts in breast milk (388, 391). Breast-feeding is not contraindicated for mothers with simple mastitis on antibiotic therapy but is contraindicated for those with breast abscesses because of the risk of transmission of large doses of bacteria to the newborn (391). Outbreak Control A significant change over background rate in infections at a certain site or with a particular microbe should be considered an outbreak, and measures should be taken to identify the microorganisms involved, the reservoir, and the risk factors for transmission or acquisition of infection (422). Suspicion of an outbreak should lead to review of general infection control procedures, with emphasis on compliance with hand washing between infants and review of practices for sterilization and decontamination of equipment, preparation of infant formula, and aseptic techniques for invasive procedures. In many instances, such review alone has ended an outbreak before or without identification of a specific point source or problem in procedure. Increased infection rates involving a number of different microbes or strains of the same microbe are likely to be related to (a) breakdown in infection control procedures such as occurs with crowding, understaffing, or other major disruption of the routine functioning of the unit; (b) defective sterilization or disinfection technique; or (c) a change in the use of invasive procedures. If an epidemic microbe is suspected, attempts should be made to identify and isolate or cohort infected or colonized patients using rapid microbiologic testing. If this is not possible, infants who are symptomatic, infants who are asymptomatic but exposed, and infants who are not exposed (including new admissions) should be cohorted. Cohorts should be kept in separate rooms or in well-demarcated areas of a large room (365). Cohorting of personnel has also been recommended, but the efficacy of this has been questioned (423). It may be counterproductive if it results in understaffing and disruption of nursery routine. Enhanced attention to hand washing and barrier precautions may be more productive. Surveillance should be instituted for infants recently discharged. Potential environmental sources or personnel should be cultured only if preliminary epidemiologic investigation suggests an association with infection (371). If an outbreak is not brought under control by these measures, it may be necessary to close the unit to new admissions until all exposed infants have been discharged. If the microorganism implicated is endemic to nursery populations, such as S. epidermidis, E. coli, or Candida, further typing by molecular techniques may be necessary to determine if a true common-source outbreak exists or if several strains are involved (61, 287, 424). Outbreak investigation is discussed in detail in Chapter 7. Enhancement of Neonatal Defenses Bacterial Interference The principle of using one strain of bacteria to prevent colonization with another was used during the S. aureus pandemic, when Shinefield implanted the avirulent S. aureus strain 502A into the nose and umbilicus of newborns and found protection against colonization with virulent strains. This approach was used successfully to control outbreaks of S. aureus infections in several nurseries (250, 257). Artificial colonization of the pharynx with О-hemolytic Streptococcus strain 215 protected infants in NICUs from pharyngeal colonization with gram-negative microorganisms (425). Pharyngeal implantation of this strain was used, along with other infection control measures, to control an NICU outbreak of infections caused by antibiotic-resistant Enterobacteriaceae (426). More recently, attempts to control fecal colonization with gram-negative aerobes by feeding premature infants Lactobacillus were unsuccessful (427, 428). In other reports, antibiotic-sensitive E. coli strains were used successfully to suppress gastrointestinal colonization with resistant enteric microorganisms (429), and infants artificially colonized with an avirulent E. coli strain had fewer nosocomial infections than control infants (430). Postexposure Prophylaxis Postexposure antibiotic prophylaxis is recommended for the newborn of a mother with untreated gonorrhea, syphilis, infectious tuberculosis, or pertussis (208, 371). Antiretroviral prophylaxis is recommended for infants born to HIV-infected mothers (366). There may be an indication for antibiotic or acyclovir prophylaxis in selected high-risk newborns with intrapartum exposure to GBS or HSV (271, 208, 349). Newborns exposed postnatally to pertussis or to invasive H. influenzae or meningococcal infection should receive prophylaxis (208). Postexposure immunoprophylaxis for HBV and varicella were discussed earlier. Immune globulin is recommended for nonimmune newborns exposed to measles (208). Immune globulin has been administered to newborns and nursery staff to control outbreaks of HAV (114) and has been used in nursery outbreaks of enteroviral infections with variable results (342). Immunizations Premature infants respond well to protein antigens, including diphtheria and tetanus toxoids (431). Newborns remaining in the NICU should receive diphtheria, tetanus, acellular pertussis, H. influenzae B conjugate, and inactivated polio vaccines at full dose at the usual chronologic age (375). Failure to vaccinate newborns who remain in the hospital leaves them at risk for pertussis from in-hospital exposure (432). Vaccination with live polio vaccine should be deferred until discharge because of the risk of transmission of the vaccine virus to immunocompromised patients. VLBW infants respond to HBV vaccine given at birth, but response is suboptimal (433). Those born to HBsAg-positive mothers should be vaccinated at birth. Otherwise, HBV vaccination of premature infants of less than 2-kg birth weight should be deferred until weight is 2 kg or until age 2 months (375). P.874 Conjugated pneumococcal vaccine is immunogenic in preterm infants and should also be given at the usual chronologic age (434). Neonates with chronic pulmonary disease, congenital heart disease, or other specified high-risk conditions should receive influenza vaccine at age 6 months (208, 375). Influenza vaccine is recommended for pregnant women who will be in the second or third trimester during influenza season (375). Immunization during pregnancy to protect the newborn against other pathogens is an approach that is being explored (435). Immunotherapeutic Agents Attempts to prevent infections in premature newborns by intravenous administration of Оі-globulin have had conflicting results, but recent analyses suggest there is no benefit (436, 437). Effective prophylaxis probably awaits the development of immunoglobulin preparations with sufficient concentrations of antibodies against common neonatal pathogens. Immunoglobulin with high antibody titer to RSV and monoclonal anti-RSV antibody (palivizumab) are protective against RSV disease and recommended for selected high-risk infants (208). Neutrophil transfusions may be useful as adjunctive therapy in sepsis but are not a practical prophylactic measure. Granulocyte colony-stimulating factor and granulocyte-monocyte colony-stimulating factor enhance neutrophil production and function in newborns and may have potential for prophylaxis, but data to date are inconclusive (436, 438). Further advances in the understanding of immune function in the newborn may lead to new strategies to strengthen neonatal defenses. REFERENCES 1. Goldmann DA, Freeman J, Durbin WA. Nosocomial infection and death in a neonatal intensive care unit. J Infect Dis 1983;147:635–641. 2. La Gamma EF, Drusin LM, Mackles AW, et al. Neonatal infections. An important determinant of late NICU mortality in infants less than 1000 g at birth. Am J Dis Child 1983;137:838–841. 3. Zafar N, Wallace CM, Kieffer P, et al. Improving survival of vulnerable infants increases neonatal intensive care unit nosocomial infection rate. Arch Pediatr Adolesc Med 2001;155:1098–1104. 4. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;15:128–140. 5. Goldmann DA. Prevention and management of neonatal infections. Infect Dis Clin North Am 1989;3:779–813. 6. Holbrook KF, Nottebart VF, Hameed SR, et al. Automated postdischarge surveillance for postpartum and neonatal nosocomial infections. Am J Med 1991;91(Suppl 3B):125S–130S. 7. Gaynes RP, Edwards JR, Jarvis WR, et al. Nosocomial infections among neonates in high-risk nurseries in the United States. Pediatrics 1996;98:357–361. 8. Jarvis WR. Epidemiology of nosocomial infections in pediatric patients. Pediatr Infect Dis J 1987;6:344–351. 9. Gaynes RP, Martone WJ, Culver DH, et al. Comparison of rates of nosocomial infections in neonatal intensive care units in the United States. Am J Med 1991:91(Suppl 3B):192S–196S. 10. Gray JE, Richardson DK, McCormick MC, et al. Coagulase-negative staphylococcal bacteremia among very low birth weight infants: relation to admission illness severity, resource use and outcome. Pediatrics 1995;95:225–230. 11. Beck-Sague CM, Azimi P, Fonesca SN, et al. Bloodstream infections in neonatal intensive care unit patients: results of a multicenter study. Pediatr Infect Dis J 1994;13:1110–1116. 12. Avila-Figueroa C, Goldmann DA, Richardson DK, et al. Intravenous lipid emulsions are the major determinant of coagulase-negative staphylococcal bacteremia in very low birth weight newborns. Pediatr Infect Dis J 1998:17:10–17. 13. Auriti C, Maccallini A, Di Liso G, et al. Risk factors for nosocomial infections in a neonatal intensive-care unit. J Hosp Infect 2003;53:25–30. 14. Haley RW, Cushion NB, Tenover FC, et al. Eradication of endemic methicillin-resistant Staphylcoccus aureus infections from a neonatal intensive care unit. J Infect Dis 1995;171:614–624. 15. Goldmann DA, Durbin WA, Freeman J. Nosocomial infections in a neonatal intensive care unit. J Infect Dis 1981;144:449–459. 16. CDC NNIS System. National Nosocomial Infections Surveillance (NNIS) System report, data summary from January 1992 to June 2002, issued August 2002. Am J Infect Control 2002;30:458–75. 17. Gladstone IM, Ehrenkranz RA, Edberg SC, et al. A ten year review of neonatal sepsis and comparison with the previous fifty-year experience. Pediatr Infect Dis J 1990;9:819–825. 18. Campins M, VaquГ J, RossellГі, et al. Nosocomial infections in pediatric patients: a prevalence study in Spanish hospitals. Am J Infect Control 1993;21:58–63. 19. Aavitsland P, Stormark M, Lystad A. Hospital-acquired infections in Norway: a national prevalence survey in 1991. Scand J Infect Dis 1992;24:477–483. 20. Sohn AH, Garrett DO, Sinkowitz-Cochran RL, et al. Prevalence of nosocomial infections in neonatal intensive care unit patients: results from the first national point-prevalence survey. J Pediatr 2001;139:821-7. 21. Scheckler WE, Peterson PJ. Nosocomial infections in 15 rural Wisconsin hospitals—results and conclusions from 6 months of comprehensive surveillance. Infect Control 1986;7:397–402. 22. Maguire GC, Nordin J, Myers MG, et al. Infections acquired by young infants. Am J Dis Child 1981;135:693–698. 23. Welliver RC, McLaughlin S. Unique epidemiology of nosocomial infection in a children's hospital. Am J Dis Child 1984;138:131–135. 24. Horan TC, White JW, Jarvis WR, et al. Nosocomial infection surveillance, 1984. MMWR CDC Surveill Summ 1986;35:17SS–29SS. 25. Bureau of Communicable Disease Epidemiology, Laboratory Centre for Disease Control, Health and Welfare Canada. Canadian Nosocomial Infection Surveillance Program. Annual summary June 1984–May 1985. Can Dis Wkly Rep 1986;12:S1. 26. Ford-Jones EL, Mindorff CM, Langley JM, et al. Epidemiologic study of 4684 hospital-acquired infections in pediatric patients. Pediatr Infect Dis J 1989;8:668–675. 27. Flenik LT, Bagatin AC, Castro ME, et al. Hospital infections in a high-risk nursery: 2-year analysis. Rev Soc Bras Med Trop 1990;23:91–95. 28. Josephson A, Karanfil L, Alonso H, et al. Risk-specific nosocomial infection rates. Am J Med 1991;91(Suppl 3B):131S–137S. 29. Boutte P, Berard E, Haas H, et al. L'infection nГonatale nosocomiale dans une unitГ de rГanimation pГdiatrique et une unitГ d'Гlevage. PГdiatrie 1990;45:889–893. 30. Orrett FA, Brooks PJ, Richardson EG. Nosocomial infections in a rural regional hospital in a developing country: incidence rates by site, service, cost and infection control practices. Infect Control Hosp Epidemiol 1998;19:136–140. 31. Drews MB, Ludwig AC, Leititis JU, et al. Low birth weight and nosocomial infection of neonates in a neonatal intensive care unit. J Hosp Infect 1995;30:65–72. 32. Ferguson JK, Gill A. Risk-stratified nosocomial infection surveillance in a neonatal intensive care unit: report on 24 months of surveillance. J Paediatr Child Health 1996;32:525–531. 33. Moro ML, De Toni A, Stolfi I, et al. Risk factors for nosocomial sepsis in newborn intensive care and intermediate care units. Eur J Pediatr 1996;155:315–322. 34. Raymond J, Aujard Y, the European Study Group. Nosocomial infections in pediatric patients: a European, multicenter prospective study. Infect Control Hosp Epidemiol 2000;21:260–263. P.875 35. Moore DL, Rodrigues R. Surveillance for nosocomial infections in a neonatal intensive care unit—a ten year perspective. Abstracts of the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington DC: American Society for Microbiology, 1997:312. 36. Kawagoe JY, Segre CA, Pereira CR, et al. Risk factors for nosocomial infections in critically ill newborns: a 5-year prospective cohort study. Am J Infect Control 2001;29:109–114. 37. Stover BH, Shulman ST, Bratcher DF, et al. Nosocomial infection rates in US children's hospitals' neonatal and pediatric intensive care units. Am J Infect Control 2001;29:152–157. 38. Leonard EM, van Saene HK, Shears P, et al. Pathogenesis of colonization and infection in a neonatal surgical unit. Crit Care Med 1990;18:264–269. 39. Khuri-Bulos NA, Shennak M, Agabi S, et al. Nosocomial infections in the intensive care units at a university hospital in a developing country: comparison with National Nosocomial Infections Surveillance intensive care unit rates. Am J Infect Control 1999; 27:547–552. 40. Long SS, Swenson RM. Development of anaerobic fecal flora in healthy newborn infants. J Pediatr 1977;91:298–301. 41. Graham JM, Taylor J, Davies PA. Some aspects of bacterial colonization in ill, low-birth-weight, and normal newborns. In: Stern L, Friis-Hansen B, Kildeberg P, eds. Intensive care in the newborn. New York: Masson, 1976:59–72. 42. Sprunt K, Leidy G, Redman W. Abnormal colonization of neonates in an intensive care unit: means of identifying neonates at risk of infection. Pediatr Res 1978;12:998–1002. 43. Bullen CL, Tearle PV, Willis AT. Bifidobacteria in the intestinal tract of infants: an in-vivo study. J Med Microbiol 1976;9:325–333. 44. Yoshioka H, Iseki K, Fujita K. Development and differences of intestinal flora in the neonatal period in breast-fed and bottle-fed infants. Pediatrics 1983;72:317–321. 45. Murono K, Fujita K, Yoshikawa M, et al. Acquisition of nonmaternal Enterobacteriaceae by infants delivered in hospitals. J Pediatr 1993;122:120–125. 46. Blakey JL, Lubitz L, Barnes GL, et al. Development of gut colonization in pre-term neonates. J Med Virol 1982;15:519–529. 47. Goldmann DA, Leclair J, Macone A. Bacterial colonization of neonates admitted to an intensive care environment. J Pediatr 1978;93:288–293. 48. Bennet R, Eriksson M, Nord C-E, et al. Fecal bacterial microflora of newborn infants during intensive care management and treatment with five antibiotic regimens. Pediatr Infect Dis J 1986;5:533–539. 49. Jarvis WR, Munn VP, Highsmith AK, et al. The epidemiology of nosocomial infections caused by Klebsiella pneumoniae. Infect Control 1985;6:68–74. 50. Toltzis P, Dul MJ, Hoyen C, et al. Molecular epidemiology of antibiotic-resistant gram-negative bacilli in a neonatal intensive care unit during a nonoutbreak period. Pediatrics 2001;108:1143–1148. 51. Almuneef MA, Baltimore RS, Farrel PA, et al. Molecular typing demonstrating transmission of gram-negative rods in a neonatal intensive care unit in the absence of a recognized epidemic. Clin Infect Dis 2001;32:220–227. 52. Singh N, Patel KM, Leger MM, et al. Risk of resistant infections with Enterobacteriaceae in hospitalized neonates. Pediatr Infect Dis J 2002;21:1029–1033. 53. Fonesca SN, Ehrenkrantz RA, Baltimore RS. Epidemiology of antibiotic use in a neonatal intensive care unit. Infect Control Hosp Epidemiol 1994;15:156–162. 54. Stoll BJ, Gordon T, Korones SB, et al. Early-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996;129:72–80. 55. Lee SK, McMillan DD, Ohlsson A, et al. Variations in practice and outcomes in the Canadian NICU network: 1996–1997. Pediatrics 2000;106:1070–1079. 56. Cordero L, Sananes M, Ayers LW. Bloodstream infections in a neonatal intensive-care unit: 12 years' experience with an antibiotic control program. Infect Control Hosp Epidemiol 1999;20:242–246. 57. de Man P, Verhoeven BA, Verbrugh HA, et al. An antibiotic policy to prevent emergence of resistant bacilli. Lancet 2000;355:973–978. 58. Calil R, Marba ST, von Nowakonski A, et al. Reduction in colonization and nosocomial infection by multiresistant bacteria in a neonatal unit after institution of educational measures and restriction in the use of cephalosporins. Am J Infect Control 2001;29:133–138. 59. Savey A, Fleurette J, Salle BL. An analysis of the microbial flora of premature neonates. J Hosp Infect 1992 21:275–289. 60. D'Angio CT, McGowan KL, Baumgart S, et al. Surface colonization with coagulase-negative staphylococci in premature neonates. J Pediatr 1989;114:1029–1034. 61. Huebner J, Pier GB, Maslow JN et al. Endemic nosocomial transmission of Staphylococcus epidermidis bacteremia isolates in a neonatal intensive care unit over 10 years. J Infect Dis 1994;169;526–531. 62. Sprunt K. Practical use of surveillance for prevention of nosocomial infection. Semin Perinatol 1985;9:47–50. 63. Uehara Y, Kikuchi K, Nakamura T, et al. Inhibition of methicillin-resistant Staphylococcus aureus colonization of oral cavities in newborns by viridans group streptococci. Clin Infect Dis 2001;32:1399–1407. 64. Hentges DJ. The anaerobic microflora of the human body. Clin Infect Dis 1993;16(Suppl 4):S175–180. 65. Mortimer EA, Lipsitz PJ, Wolinsky E, et al. Transmission of staphylococci between newborns. Importance of the hands of personnel. Am J Dis Child 1962;104:289–295. 66. Rammelkamp CH, Mortimer EA, Wolinksy E. Transmission of streptococcal and staphylococcal infections. Ann Intern Med 1964;60:753–758. 67. Yu WL, Cheng HS, Lin HC, et al. Outbreak investigation of nosocomial Enterobacter cloacae bacteraemia in a neonatal intensive care unit. Scand J Infect Dis 2000;32:293–298. 68. Moolenaar RL, Crutcher JM, San Joaquin VH, et al. A prolonged outbreak of Pseudomonas aeruginosa in a neonatal intensive care unit: did staff fingernails play a role in disease transmission? Infect Control Hosp Epidemiol 2000;21:80–85. 69. Foca M, Jakob K, Whittier S, et al. Endemic Pseudomonas aeruginosa infection in a neonatal intensive care unit. N Engl J Med 2000;343:695–700. 70. Huang YC, Su LH, Wu TL, et al. Outbreak of Acinetobacter baumannii bacteremia in a neonatal intensive care unit: clinical implications and genotyping analysis. Pediatr Infect Dis J 2002;21:1105–1109. 71. Hedberg K, Ristinen TL, Soler JT, et al. Outbreak of erythromycin-resistant staphylococcal conjunctivitis in a newborn nursery. Pediatr Infect Dis J 1990;9:268–273. 72. Reboli AC, John JF, Levkoff AH. Epidemic methicillin-gentamicin-resistant Staphylococcus aureus in a neonatal intensive care unit. Am J Dis Child 1989;143:34–39. 73. Coudron PE, Mayhall CG, Facklam RR, et al. Streptococcus faecium outbreak in a neonatal intensive care unit. J Clin Microbiol 1984;20:1044–1048. 74. Kinney JS, McCray E, Kaplan JE, et al. Risk factors associated with echovirus 11 infection in a hospital nursery. Pediatr Infect Dis J 1986;5:192–197. 75. Steere AC, Mallison GF. Handwashing practices for the prevention of nosocomial infections. Ann Intern Med 1975;83:683–690. 76. Parry MF, Hutchinson JH, Brown NA, et al. Gram-negative sepsis in neonates: a nursery outbreak due to hand carriage of Citrobacter diversus. Pediatrics 1980;65:1105–1109. 77. Belani A, Sherertz RJ, Sullivan ML, et al. Outbreak of staphylococcal infection in two hospital nurseries traced to a single nasal carrier. Infect Control 1986;7:487–490. 78. Coovadia YM, Bhana RH, Johnson AP, et al. A laboratory-confirmed outbreak of rifampicin-methicillin resistant Staphylococcus aureus (RMRSA) in a newborn nursery. J Hosp Infect 1989;14:303–312. 79. Anagnostakis D, Fitsialos J, Koutsia C, et al. A nursery outbreak of Serratia marcescens infection. Am J Dis Child 1981;135:413–414. 80. Archibald LK, Corl A, Shah B, et al. Serratia marcescens outbreak associated with extrinsic contamination of 1% chloroxylenol soap. Infect Control Hosp Epidemiol 1997;18:704–709. 81. Bobo RA, Newton EJ, Jones LF, et al. Nursery outbreak of Pseudomonas aeruginosa: epidemiologic conclusions from five different typing methods. Appl Microbiol 1973;25:414–420. 82. Nelson KE, Warren D, Tomasi AM, et al. Transmission of neonatal listeriosis in a delivery room. Am J Dis Child 1985;139:903–905. P.876 83. Hartstein AI, Rashad AL, Liebler JM, et al. Multiple intensive care unit outbreak of Acinetobacter calcoaceticus subspecies anitratus respiratory tract infection and colonization associated with contaminated reusable ventilator circuits and resuscitation bags. Am J Med 1988;85:624–631. 84. Neal TJ, Hughes CR, Rothburn MM, et al. The neonatal laryngoscope as a potential source of cross-infection. J Hosp Infect 1995;30:315–321. 85. Van Der Zwet WC, Parlevliet GA, Savelkoul PH, et al. Outbreak of Bacillus cereus infections in a neonatal intensive care unit traced to balloons used in manual ventilation. J Clin Microbiol 2000;38:4131–4136. 86. Drewett SE, Payne DJH, Tuke W, et al. Eradication of Pseudomonas aeruginosa infection from a special-care nursery. Lancet 1972;1:946–948. 87. Montanaro D, Grasso GM, Annino I, et al. Epidemiological and bacteriological investigation of Serratia marcescens epidemic in a nursery and in a neonatal intensive care unit. J Hyg (Camb) 1984;93:67–78. 88. Khan MA, Abdur-Rab M, Israr N, et al. Transmission of Salmonella worthington by oropharyngeal suction in hospital neonatal unit. Pediatr Infect Dis J 1991;10:668–672. 89. Gray J, George RH, Durbin GM, et al. An outbreak of Bacillus cereus respiratory tract infections on a neonatal unit due to contaminated ventilator circuits. J Hosp Infect 1999;41:19–22. 90. McAllister TA, Roud JA, Marshall A, et al. Outbreak of Salmonella eimsbuettel in newborn infants spread by rectal thermometers. Lancet 1986;1:1262–1264. 91. Abrahamsen TG, Finne PH, Lingaas E. Flavobacterium meningosepticum infections in a neonatal intensive care unit. Acta Paediatr Scand 1989;78:51–55. 92. Fleisch F, Zimmermann-Baer U, Zbinden R, et al. Three consecutive outbreaks of Serratia marcescens in a neonatal intensive care unit. Clin Infect Dis 2002;34:767–773. 93. Plotkin SA, McKitrick JC. Nosocomial meningitis of the newborn caused by a flavobacterium. JAMA 1966;198:194–196. 94. McCormack RC, Kunin CM. Control of a single source nursery epidemic due to Serratia marcescens. Pediatrics 1966;37:750–755. 95. Schuchat A, Lizano C, Broome CV, et al. Outbreak of neonatal listeriosis associated with mineral oil. Pediatr Infect Dis J 1991;10:183–189. 96. Mumford F, Hindes R, Given L. Nosocomial Xanthomonas meningitis/ventriculitis. Am J Infect Control 1993;21:85(abst). 97. Weist K, Wendt C, Petersen LR, et al. An outbreak of pyodermas among neonates caused by ultrasound gel contaminated with methicillin-susceptible Staphylococcus aureus. Infect Control Hosp Epidemiol 2000;21:761–764. 98. Welbel SF, McNeil MM, Kuykendall RJ, et al. Candida parapsilosis bloodstream infections in neonatal intensive care unit patients: epidemiologic and laboratory confirmation of a common source outbreak. Pediatr Infect Dis J 1996;15:998–1002. 99. McNaughton M, Mazinke N, Thomas E. Newborn conjunctivitis associated with triclosan 0.5% antiseptic intrinsically contaminated with Serratia marcescens. Can J Infect Control 1995;10:7–8. 100. Anonymous. Water bugs in the bassinet [Editorial]. Am J Dis Child 1961;101:273–277. 101. Moffet HL, Allan D, Williams T. Survival and dissemination of bacteria in nebulizers and incubators. Am J Dis Child 1967;114:13–20. 102. Muyldermans G, Desmet F, Pierard D, et al. Neonatal infections with Pseudomonas aeruginosa associated with a water-bath used to thaw fresh frozen plasma. J Hosp Infect 1998;39:309–314. 103. Verweij PE, Meis JF, Christmann V, et al. Nosocomial outbreak of colonization and infection with Stenotrophomonas maltophilia in preterm infants associated with contaminated tap water. Epidemiol Infect 1998;120:251–256. 104. Goldmann DA, Pier GB. Pathogenesis of infections related to intravascular catheterization. Clin Microbiol Rev 1993;6:176–192. 105. Matsaniotis NS, Syriopoulou VP, Theodoridou MC, et al. Enterobacter sepsis in infants and children due to contaminated intravenous fluids. Infect Control 1984;5:471–477. 106. Tresoldi AT, Padoveze MC, Trabasso P, et al. Enterobacter cloacae sepsis outbreak in a newborn unit caused by contaminated total parenteral nutrition solution. Am J Infect Control 2000;28:258–261. 107. Jarvis WR, Highsmith AK, Allen JR, et al. Polymicrobial bacteremia associated with lipid emulsion in a neonatal intensive care unit. Pediatr Infect Dis J 1983;2:203–208. 108. Fleer A, Senders RC, Visser MR, et al. Septicemia due to coagulase-negative staphylococci in a neonatal intensive care unit: clinical and bacteriological features and contaminated parenteral fluids as a source of sepsis. Pediatr Infect Dis J 1983;2:426–431. 109. Zaidi M, Sifuentes J, Bobadilla M, et al. Epidemic of Serratia marcescens bacteremia and meningitis in a neonatal unit in Mexico City. Infect Control Hosp Epidemiol 1989;10:14–20. 110. Sherertz RJ, Gledhill KS, Hampton KD, et al. Outbreak of Candida bloodstream infections associated with retrograde medication administration in a neonatal intensive care unit. J Pediatr 1992;120:455–461. 111. Archibald LK, Ramos M, Arduino MJ, et al. Enterobacter cloacae and Pseudomonas aeruginosa polymicrobial bloodstream infections traced to extrinsic contamination of a dextrose multidose vial. J Pediatr 1998;133:640–644. 112. de Beaufort AJ, Bernards AT, Dijkshoorn L, et al. Acinetobacter junii causes life-threatening sepsis in preterm infants. Acta Paediatr 1999;88:772–775. 113. Adler SP. Nosocomial transmission of cytomegalovirus. Pediatr Infect Dis J 1986;5:239–246. 114. Rosenblum LS, Villarino ME, Nainan OV, et al. Hepatitis A outbreak in a neonatal intensive care unit: risk factors for transmission and evidence of prolonged viral excretion among preterm infants. J Infect Dis 1991;164:476–482. 115. King EA, Alter AA, Schwartz O, et al. Postexchange transfusion hepatitis in the newborn infant. J Pediatr 1973;83:341–342. 116. Aach RD, Yomtovian RA, Hack M. Neonatal and pediatric posttransfusion hepatitis C: a look back and a look forward. Pediatrics 2000;105:836–842. 117. Saulsbury FT, Wykoff RF, Boyle RJ. Transfusion-acquired human immunodeficiency virus in twelve neonates: epidemiologic, clinical and immunological features. Pediatr Infect Dis J 1987;6:544–549. 118. Shulman IA, Saxena S, Nelson JM, et al. Neonatal exchange transfusions complicated by transfusion-induced malaria. Pediatrics 1984;73:330–332. 119. Piccoli DA, Perlman S, Ephros M. Transfusion-acquired Plasmodium malariae infection in two premature infants. Pediatrics 1983;72:560–562. 120. Oxtoby MJ. Human immunodeficiency virus and other viruses inhuman milk: placing the issues in broader perspective. Pediatr Infect Dis J 1988;7:825–835. 121. Dworsky M, Yow M, Stagno S, et al. Cytomegalovirus infection of breast milk and transmission in infancy. Pediatrics 1983;72:295–299. 122. Botsford KB, Weinstein RA, Boyer KB, et al. Gram-negative bacilli in human milk feedings: quantitation and clinical consequences for premature infants. J Pediatr 1986;109:707–710. 123. Donowitz LG, Marsik FJ, Fisher KA, et al. Contaminated breast milk: a source of Klebsiella bacteremia in a newborn intensive care unit. Rev Infect Dis 1981;3:716–720. 124. Gransden WR, Webster M, French GL, et al. An outbreak of Serratia marcescens transmitted by contaminated breast pumps in a special care baby unit. J Hosp Infect 1986;7:149–154. 125. Ryder RW, Crosby-Ritchie A, McDonough B, et al. Human milk contaminated with Salmonella kottbus. A cause of nosocomial illness in infants. JAMA 1977;238:1533–1534. 126. Stiver HG, Albritton WL, Clark J, et al. Nosocomial colonization and infection due to E. coli 0125:K70 epidemiologically linked to expressed breast-milk feedings. Can J Public Health 1977;68:479–482. 127. Parks YA, Noy MF, Aukett MA, et al. Methicillin resistant Staphylococcus aureus in milk. Arch Dis Child 1987;62:82–83. 128. Schreiner RL, Eitzen H, Gfell MA, et al. Environmental contamination of continuous drip feedings. Pediatrics 1979;63:232–237. P.877 129. Noriega FR, Kotloff KL, Martin MA, et al. Nosocomial bacteremia caused by Enterobacter sakazakii and Leuconostoc mesenteroides resulting from extrinsic contamination of infant formula. Pediatr Infect Dis J 1990;9:447–449. 130. Block C, Peleg O, Minster N, et al. Cluster of neonatal infections in Jerusalem due to unusual biochemical variant of Enterobacter sakazakii. Eur J Clin Microbiol Infect Dis 2002;21:613–616. 131. Himelright I, Harris E, Lorch V, et al. Enterobacter sakazakii infections associated with the use of powdered infant formula—Tennessee, 2001. MMWR Mortal Morb Wkly Rep 2002;51:298–300. 132. Graman PS, Hall CB. Epidemiology and control of nosocomial viral infections. Infect Dis Clin North Am 1989;3:815–841. 133. Hall CB, Kopelman AE, Douglas G, et al. Neonatal respiratory syncytial virus infection. N Engl J Med 1979;300:393–396. 134. Harpin VA, Rutter N. Barrier properties of the newborn infant's skin. J Pediatr 1983;102:419–425. 135. Lewis DB, Wilson CB. Developmental immunology and role of host defenses in fetal and neonatal susceptibility to infection. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant, 5th ed. Philadelphia: WB Saunders, 2001:25–138. 136. Haley RW, Bregman DA. The role of understaffing and overcrowding in recurrent outbreaks of staphylococcal infection in a neonatal special-care unit. J Infect Dis 1982;145:875–885. 137. Harbarth S, Sudre P, Dharan S, et al. Outbreak of Enterobacter cloacae related to understaffing, overcrowding, and poor hygiene practices. Infect Control Hosp Epidemiol 1999;20:598–603. 138. Cordero L, Anderson CW, Zuspan FP. Scalp abscess: a benign and infrequent complication of fetal monitoring. Am J Obstet Gynecol 1983;146:126–130. 139. Parvey LS, Ch'ien LT. Neonatal herpes simplex virus infection introduced by fetal-monitor scalp electrodes. Pediatrics 1980;65:1150–1153. 140. Asmar BI. Osteomyelitis in the neonate. Infect Dis Clin North Am 1992;6:117–132. 141. Berthelot P, Grattard F, Patural H, et al. Nosocomial colonization of premature babies with Klebsiella oxytoca: probable role of enteral feeding procedure in transmission and control of the outbreak with the use of gloves. Infect Control Hosp Epidemiol 2001;22:148–151. 142. Widdowson MA, van Doornum GJ, van der Poel WH, et al. Emerging group-A rotavirus and a nosocomial outbreak of diarrhoea. Lancet 2000;356:1161–1162. 143. Mehall JR, Kite CA, Gilliam CH, et al. Enteral feeding tubes are a reservoir for nosocomial antibiotic-resistant pathogens. J Pediatr Surg 2002;37:1011–1012. 144. Cronin WA, Germanson TP, Donowitz LG. Intravascular catheter colonization and related bloodstream infection in critically ill neonates. Infect Control Hosp Epidemiol 1990;11:301–308. 145. Nelson JD. The neonate. In: Donowitz LG, ed. Hospital-acquired infection in the pediatric patient. Baltimore: Williams & Wilkins, 1988:273–294. 146. Stamm WE, Kolff CA, Dones EM, et al. A nursery outbreak caused by Serratia marcescens—scalp-vein needles as a portal of entry. J Pediatr 1976;89:96–99. 147. Landers S, Moise AA, Fraley JK, et al. Factors associated with umbilical catheter-related sepsis in neonates. Am J Dis Child 1991;145:675–680. 148. Decker MD, Edwards KM. Central venous catheter infections. Pediatr Clin North Am 1988;35:579–612. 149. Sadiq HF, Devaskar S, Keenan WJ, et al. Broviac catheterization in low birth weight infants: incidence and treatment of associated complications. Crit Care Med 1987;15:47–50. 150. Chathas MK, Paton JB, Fisher DE. Percutaneous central venous catheterization. Am J Dis Child 1990;144:1246–1250. 151. Polberger S, Jirwe M, Svenningsen NW. Silastic central venous catheters for blood sampling and infusions in newborn infants. Prenat Neonat Med 1998;3:340–345. 152. Chien LY, Macnab Y, Aziz K, et al. Variations in central venous catheter-related infection risks among Canadian neonatal intensive care units. Pediatr Infect Dis J 2002;21:505–511. 153. Mahieu LM, De Dooy JJ, Lenaerts AE, et al. Catheter manipulations and the risk of catheter-associated bloodstream infection in neonatal intensive care unit patients. J Hosp Infect 2001;48:20–26. 154. Solomon SL, Alexander H, Eley JW, et al. Nosocomial fungemia in neonates associated with intravascular pressure-monitoring devices. Pediatr Infect Dis J 1986;5:680–685. 155. Weems JJ Jr, Chamberland ME, Ward J, et al. Candida parapsilosis fungemia associated with parenteral nutrition and contaminated blood pressure transducers. J Clin Microbiol 1987;25:1029–1032. 156. Steiner CK, Stewart DL, Bond SJ, et al. Predictors of acquiring a nosocomial bloodstream infection on extracorporeal membrane oxygenation. J Pediatr Surg 2001;36:487–492. 157. Kaisan GF, Elash JH, Tan LK. Bacteriologic surveillance of indwelling urinary catheters in pediatric intensive care unit patients. Crit Care Med 1988;16:679–682. 158. Pople IK, Bayston R, Hayward RD. Infection of cerebrospinal fluid shunts in infants: a study of etiological factors. J Neurosurg 1992;77:29–36. 159. Overturf GD, Marcy SM. Focal bacterial infections. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant, 5th ed. Philadelphia: WB Saunders, 2001:1047–1089. 160. Glaser JB, Engelberg M, Hammerschlag M. Scalp abscess associated with Mycoplasma hominis infection complicating intrapartum monitoring. Pediatr Infect Dis J 1983;2:468–470. 161. Nightingale LM, Eaton CB, Fruehan AE, et al. Cephalohematoma complicated by osteomyelitis due to Gardnerella vaginalis. JAMA 1986;256:1936–1937. 162. Faix RG, Kovarik SM, Shaw TR, et al. Mucocutaneous and invasive candidiasis among very low birth weight (<1500 grams) infants in intensive care nurseries: a prospective study. Pediatrics 1989;83:101–107. 163. McKenna H, Johnson D. Bacteria in neonatal omphalitis. Pathology 1977;9:111–113. 164. Cushing AH. Omphalitis: a review. Pediatr Infect Dis 1985;4:282–285. 165. Mason WH, Andrews R, Ross LA, et al. Omphalitis in the newborn infant. Pediatr Infect Dis J 1989;8:521–525. 166. Stoll BJ. The global impact of infection. Clin Perinatology 1997;24:1–21. 167. Gee WF, Ansell JS. Neonatal circumcision: a ten-year overview: with comparison of the Gomco clamp and the Plastibell device. Pediatrics 1976;58:824–827. 168. Wiswell TE, Geschke DW. Risks from circumcision during the first month of life compared with those for uncircumcised boys. Pediatrics 1989;83:1011–1015. 169. King S, Devi SP, Mindorff C, et al. Nosocomial Pseudomonas aeruginosa conjunctivitis in a pediatric hospital. Infect Control Hosp Epidemiol 1988;9:77–80. 170. Lohrer R, Belohradsky BH, Bacterial endophthalmitis in neonates. Eur J Pediatr 1987;146:354–359. 171. Baley JE. Neonatal candidiasis: the current challenge. Clin Perinatol 1991;18:263–280. 172. Birenbaum E, Linder N, Varsano N, et al. Adenovirus type 8 conjunctivitis outbreak in a neonatal intensive care unit. Arch Dis Child 1993;68:610–611. 173. Klein JO. Bacterial sepsis and meningitis. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant, 5th ed. Philadelphia: WB Saunders, 2001:943–998. 174. Gomez M, Alter S, Kumar ML, et al. Neonatal Streptococcus pneumoniae infection: case reports and review of the literature. Pediatr Infect Dis J 1999;18:1014–1018. 175. Stoll BJ, Gordon T, Korones SB, et al. Late-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996;129:63–71. 176. HervГЎs JA, Alomar A, SalvГЎ F, et al. Neonatal sepsis and meningitis in Mallorca, Spain, 1977–91. Clin Infect Dis 1993;16:719–724. 177. The WHO Young Infants Study Group. Serious infections in young infants in developing countries: rationale for a multicenter study. Pediatr Infect Dis J 1999;18(10 Suppl):S4–7. 178. Noel GJ, Laufer DA, Edelson PJ. Anaerobic bacteremia in a neonatal intensive care unit: an eighteen-year experience. Pediatr Infect Dis J 1988;7:858–862. P.878 179. Shah SS, Ehrenkranz RA, Gallagher PG. Increasing incidence of gram-negative rod bacteremia in a newborn intensive care unit. Pediatr Infect Dis J 1999;18:591–595. 180. Baltimore RS, Huie SM, Meek JI, et al. Early-onset neonatal sepsis in the era of group B streptococcal prevention. Pediatrics 2001;108:1094–1098. 181. Hyde TB, Hilger TM, Reingold A, et al. Trends in incidence and antimicrobial resistance of early-onset sepsis: population-based surveillance in San Francisco and Atlanta. Pediatrics 2002;110:690–695. 182. Isaacs D, Royle JA, Australasian Study Group for Neonatal Infections. Intrapartum antibiotics and early onset neonatal sepsis caused by group B Streptococcus and by other organisms in Australia. Pediatr Infect Dis J 1999;18:524–528. 183. The Investigators of the Vermont-Oxford Trials Network Database project. The Vermont-Oxford trials network: very low birth weight outcomes for 1990. Pediatrics 1993;91:540–545. 184. Stoll BJ, Hansen N, Fanaroff AA, et al. Changes in pathogens causing early-onset sepsis in very-low-birth-weight infants. N Engl J Med 2002;347:240–247. 185. Stoll BJ, Hansen N, Fanaroff AA, et al. Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics 2002;110:285–291. 186. Donowitz LG, Haley CE, Gregory WW, et al. Neonatal intensive care unit bacteremia: emergence of gram-positive bacteria as major pathogens. Am J Infect Control 1987;15:141–147. 187. Freeman J, Goldmann DA, Smith NE, et al. Association of intravenous lipid emulsion and coagulase-negative staphylococcal bacteremia in neonatal intensive care units. N Engl J Med 1990;323:301–308. 188. Stoll BJ, Temprosa M, Tyson JE, et al. Dexamethasone therapy increases infection in very low birth weight infants. Pediatrics 1999;104:e63. 189. Walsh MC, Simpser EF, Kliegman RM. Late onset of sepsis in infants with bowel resection in the neonatal period. J Pediatr 1988;112:468–471. 190. Weber TR. Enteral feeding increases sepsis in infants with short bowel syndrome. J Pediatr Surg 1995;30:1086–1089. 191. Unhanand M, Mustafa MM, McCracken GH, et al. Gram-negative enteric bacillary meningitis: a twenty-one-year experience. J Pediatr 1993;122:15–21. 192. Moreno MT, Vargas S, Poveda R, et al. Neonatal sepsis and meningitis in a developing Latin American country. Pediatr Infect Dis J 1994;13:516–520. 193. Cabrera HA, Davis GH. Epidemic meningitis of the newborn caused by Flavobacterium. Am J Dis Child 1961;101:289–295. 194. Willis J, Robinson JE. Enterobacter sakazakii meningitis in neonates. Pediatr Infect Dis J 1988;7:196–199. 195. KrcmГry V, Havlik J, VicianovГЎ L. Nosocomial meningitis caused by multiply resistant Pseudomonas cepacia. Pediatr Infect Dis J 1987;6:769. 196. Campbell JR, Diacovo T, Baker, CJ. Serratia marcescens meningitis in neonates. Pediatr Infect Dis J 1992;11:881–886. 197. Morgan MEI, Hart CA. Acinetobacter meningitis: acquired infection in a neonatal intensive care unit. Arch Dis Child 1982:57:557–559. 198. Dillon HC. 1966 Group A type 12 streptococcal infections in a newborn nursery. Am J Dis Child 1966;112:177–184. 199. Nelson JD, Dillon HC, Howard JB. A prolonged nursery epidemic associated with a newly recognized type of group A streptococcus. J Pediatr 1976;89:792–796. 200. Noel GJ, Kreiswirth BN, Edelson PJ, et al. Multiple methicillin-resistant Staphylococcus aureus strains as a cause for a single outbreak of severe disease in hospitalized neonates. Pediatr Infect Dis J 1992;11:184–188. 201. Morris JG, Lin F-YC, Morrison CB, et al. Molecular epidemiology of neonatal meningitis due to Citrobacter diversus: a study of isolates from hospitals in Maryland. J Infect Dis 1986;154:409–414. 202. Lin FC, Devoe WF, Morrison C, et al. Outbreak of neonatal Citrobacter diversus meningitis in a suburban hospital. Pediatr Infect Dis J 1987;6:50–55. 203. Graham DR, Band JD. Citrobacter diversus brain abscess and meningitis in neonates. JAMA 1981;245:1923–1925. 204. Morooka T, Takeo H, Yasumoto S, et al. Nosocomial meningitis due to Campylobacter fetus subspecies fetus in a neonatal intensive care unit. Acta Paediatr Jpn 1992;34:530–533. 205. Goossens H, Henocque G, Kremp L, et al. Nosocomial outbreak of Campylobacter jejuni meningitis in newborn infants. Lancet 1986;2:146–149. 206. James HE, Bejar R, Gluck L, et al. Ventriculoperitoneal shunts in high risk newborns weighing under 2000 grams: a clinical report. Neurosurgery 1984;15:198–202. 207. Webber S, Wilkinson AR, Lindsell D, et al. Neonatal pneumonia. Arch Dis Child 1990;65:207–211. 208. American Academy of Pediatrics. Summaries of infectious diseases. In: Pickering LK, ed. 2003 red book: report of the Committee on Infectious Diseases, 26th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2003:189–692. 209. Taylor-Robinson D. Infections due to species of Mycoplasma and Ureaplasma: an update. Clin Infect Dis 1996;23:671–684. 210. Hall SL. Coagulase-negative staphylococcal infections in neonates. Pediatr Infect Dis J 1991;10:57–67. 211. Mas-MuГoz RL, Udaeta-Mora E, Rivera-Rueda MA, et al. InfecciГіn nosocomial en reciГn nacidos con ventilaciГіn mecanica. Bol Med Hosp Infant Mex 1992;49:839–844. 212. Craven DE, Steger KA. Nosocomial pneumonia in the intubated patient. New concepts on pathogenesis and prevention. Infect Dis Clin North Am 1989;3:843–866. 213. Newport MT, John JF, Michel YM, Levkoff AH. Endemic Serratia marcescens infection in a neonatal intensive care nursery associated with gastrointestinal colonization. Pediatr Infect Dis J 1985;4:160–167. 214. Sherman MP, Goetzman BW, Ahlfors CE, et al. Tracheal aspiration and its clinical correlates in the diagnosis of congenital pneumonia. Pediatrics 1980;65:258–263. 215. Slagle TA, Bifano EM, Wolf JW, et al. Routine endotracheal cultures for the prediction of sepsis in ventilated babies. Arch Dis Child 1989;64:34–38. 216. Cordero L, Sananes M, Dedhiya P, et al. Purulence and gram-negative bacilli in tracheal aspirates of mechanically ventilated very low birth weight infants. J Perinatol 2001;21:376–381. 217. Berman SA, Balkany TJ, Simmons MA. Otitis media in the neonatal intensive care unit. Pediatrics 1978;62:198–201. 218. Rojas J, Flanigan TH. Postintubation tracheitis in the newborn. Pediatr Infect Dis 1986;5:714–715. 219. Jevon GP, Dunne WM, Hicks MJ, et al. Bacillus cereus pneumonia in premature neonates: a report of two cases. Pediatr Infect Dis J 1993;12:251–253. 220. Cleary TG, Guerrant RL, Pickering LK. Microorganisms responsible for neonatal diarrhea. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant, 5th ed. Philadelphia: WB Saunders, 2001:1249–1326. 221. Yankauer A. Epidemic diarrhea of the newborn, a nosocomial problem in developing countries [Editorial]. Am J Public Health 1991;81:415–417. 222. Seals JE, Parrott PL, McGowan JE, et al. Nursery salmonellosis: delayed recognition due to unusually long incubation period. Infect Control 1983;4:205–208. 223. Beers LM, Burke TL, Martin DB. Shigellosis occurring in newborn nursery staff. Infect Control Hosp Epidemiol 1989;10:147–149. 224. Terrier A, Altwegg M, Bader P, et al. Hospital epidemic of neonatal Campylobacter jejuni infection. Lancet 1985;2:1182. 225. Hershkowici S, Barak M, Cohen A, et al. An outbreak of Campylobacter jejuni infection in a neonatal intensive care unit. J Hosp Infect 1987;9:54–59. 226. Haffejee IE. Neonatal rotavirus infections. Rev Infect Dis 1991;13:957–962. 227. Kosloske AM. A unifying hypothesis for pathogenesis and prevention of necrotizing enterocolitis. J Pediatr 1990:117:S68–S74. 228. Caplan MS, Jilling T. New concepts in necrotizing enterocolitis. Curr Opin Pediatr 2001;13:111–115. 229. Scheifele DW. Role of bacterial toxins in neonatal necrotizing enterocolitis. J Pediatr 1990;117:S44–S46. P.879 230. Rotbart HA, Nelson WL, Glode MP, et al. Neonatal rotavirus-associated necrotizing enterocolitis: case control study and prospective surveillance during an outbreak. J Pediatr 1988;112:87–93. 231. Keller KM, Schmidt H, Wirth S, et al. Differences in the clinical and radiological pattern of rotavirus and non-rotavirus necrotizing enterocolitis. Pediatr Infect Dis J 1991;10:734–738. 232. Siu YK, Ng PC, Fung SC, et al. Double blind, randomised, placebo controlled study of oral vancomycin in prevention of necrotising enterocolitis in preterm, very low birthweight infants. Arch Dis Child Fetal Neonatal Med 1998;79:F105–109. 233. Lucas A, Cole TJ. Breast milk and necrotising enterocolitis. Lancet 1990:336;1519–1523. 234. Eibl MM, Wolf HM, FГјrnkranz H, et al. Prevention of necrotizing enterocolitis in low-birth-weight infants by IgA-IgG feeding. N Engl J Med 1988;319:1–7. 235. Hoyos AB. Reduced incidence of necrotizing enterocolitis associated with enteral administration of Lactobacillus acidophilus and Bifidobacterium infantis to neonates in an intensive care unit. Int J Infect Dis 1999;3:197–202. 236. Edelmann CM, Ogwo JE, Fine BP, et al. The prevalence of bacteriuria in full-term and premature infants. J Pediatr 1973;82:125–132. 237. Maherzi M, Guignard J-P, Torrada A. Urinary tract infection in high-risk newborn infants. Pediatrics 1978;62:521–523. 238. Davies HD, Ford-Jones EL, Sheng RY, et al. Nosocomial urinary tract infections at a pediatric hospital. Pediatr Infect Dis J 1992;11:349–354. 239. Lohr JA, Donowitz LG, Sadler JE III. Hospital-acquired urinary tract infection. Pediatrics 1989;83:193–199. 240. Langley JM, Hanakowski M, Leblanc JC. Unique epidemiology of nosocomial urinary tract infection in children. Am J Infect Control 2001;29:94–98. 241. Tullus K. Fecal colonization with P-fimbriated Escherichia coli in newborn children and relation to development of extraintestinal E. coli infections. Acta Paediatr Scand 1987;334(Suppl):1–35. 242. Doig CM, Wilkinson AW. Wound infection in a children's hospital. Br J Surg 1976;63:647–650. 243. Sharma LK, Sharma PK. Postoperative wound infection in a pediatric surgical service. J Pediatr Surg 1986;21:889–891. 244. Madden NP, Levinsky RJ, Bayston R, et al. Surgery, sepsis, and nonspecific immune function in neonates. J Pediatr Surg 1989:24:562–566. 245. Davenport M, Doig CM. Wound infection in pediatric surgery: a study of 1094 neonates. J Pediatr Surg 1993;28:26–30. 246. Moore D, Rodrigues R. Surgical site and other nosocomial infections after newborn surgery. J Hosp Infect 1998;40(Suppl A):Abstract p 2.2.1.15. 247. Bhattacharyya N, Kosloske AM. Postoperative wound infection in pediatric surgical patients: a study of 676 infants and children. J Pediatr Surg 1990;25:125–129. 248. Horwitz JR, Chawls WJ, Doski JJ, et al. Pediatric wound infections: a prospective mulicenter study. Ann Surgery 1998;227:553–558. 249. American Academy of Pediatrics. Antimicrobial prophylaxis. In: Pickering LK, ed. 2003 red book: report of the Committee on Infectious Diseases, 26th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2003:773–787. 250. Shinefield HR, St. Geme JW. Staphylococcal infections. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant, 5th ed. Philadelphia: WB Saunders, 2001:1217–1247. 251. Wise RI, Ossman EA, Littlefield DR. Personal reflections on nosocomial staphylococcal infections and the development of hospital surveillance. Rev Infect Dis 1989;11:1005–1019. 252. Back NA, Linnemann CC, Pfaller MA, et al. Recurrent epidemics caused by a single strain of erythromycin-resistant Staphylococcus aureus. JAMA 1993;270:1329–1333. 253. Gooch JJ, Britt EM. Staphylococcus aureus colonization and infections in newborn nursery patients. Am J Dis Child 1978;132:893–896. 254. Lockhart JD. How toxic is hexachlorophene? Pediatrics 1972:50:229–235. 255. Kaslow RA, Dixon RE, Martin SM, et al. Staphylococcal disease related to hospital nursery bathing practices—a nationwide epidemiologic investigation. Pediatrics 1973;51:418–429. 256. Johnson JD, Malachowski NC, Vosti KL, et al. A sequential study of various modes of skin and umbilical care and the incidence of staphylococcal colonization and infection in the neonate. Pediatrics 1976;58:354–361. 257. Shinefield HR, Ribble JC, Boris M. Bacterial interference between strains of Staphylococcus aureus, 1960 to 1970. Am J Dis Child 1971;121:148–152. 258. Moore EP, Williams EW. A maternity hospital outbreak of methicillin-resistant Staphylococcus aureus. J Hosp Infect 1991;19:5–16. 259. Rosenfeld CR, Laptook AR, Jeffery J. Limited effectiveness of triple dye in preventing colonization with methicillin-resistant Staphylococcus aureus in a special care nursery. Pediatr Infect Dis J 1990;9:290–291. 260. Farrington M, Ling J, Ling T, et al. Outbreaks of infection with methicillin-resistant Staphylococcus aureus on neonatal and burn units of a new hospital. Epidemiol Infect 1990;105:215–228. 261. Millar MR, Keyworth N, Lincoln C, et al. Methicillin-resistant Staphylococcus aureus in a regional neonatal unit. J Hosp Infect 1987;10:187–197. 262. Jernigan JA, Titus MG, GrГschel DHM, et al. Effectiveness of contact isolation during a hospital outbreak of methicillin-resistant Stapylococcus aureus. Am J Epidemiol 1996:143;496–504. 263. Zafar AB, Butler RC, Reese DJ, et al. Use of 0.3% triclosan (Bacti-Stat) to eradicate an outbreak of methicillin-resistant Stahylococcus aureus in a neonatal nursery. Am J Infect Control 1995;23:200–208. 264. Hitomi S, Kubota M, Mori N, et al. Control of a methicillin-resistant Staphylococcus aureus outbreak in a neonatal intensive care unit by unselective use of nasal mupirocin ointment. J Hosp Infect 2000;46:123–129. 265. Ish-Horowicz MR, McIntyre P, Nade S. Bone and joint infections caused by multiply resistant Staphylococcus aureus in a neonatal intensive care unit. Pediatr Infect Dis J 1992;11:82–87. 266. Freeman J, Platt R, Epstein MF, et al. Birth weight and length of stay as determinants of nosocomial coagulase-negative staphylococcal bacteremia in neonatal intensive care unit populations: potential for confounding. Am J Epidemiol 1990;132:1130–1140. 267. Freeman J, Epstein MF, Smith NE, et al. Extra hospital stay and antibiotic usage with nosocomial coagulase-negative staphylococcal bacteremia in two neonatal intensive care unit populations. Am J Dis Child 1990;144:324–329. 268. Van Der Zwet WC, Debets-Ossenkopp YJ, Reinders E, et al. Nosocomial spread of a Staphylococcus capitis strain with heteroresistance to vancomycin in a neonatal intensive care unit. J Clin Microbiol 2002;40:2520–2525. 269. Noya FJD, Baker CJ. Prevention of group B streptococcal infection. Infect Dis Clin North Am 1992;6:41–55. 270. Noya FJ, Rench MA, Metzger TG, et al. Unusual occurrence of an epidemic of type Ib/c group B streptococcal sepsis in a neonatal intensive care unit. J Infect Dis 1987;155:1135–1144. 271. Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease. MMWR Mortal Morb Wkly Rep 2002;51(no. RR-11):1–24. 272. Schrag SJ, Zell ER, Lynfield R, et al. A population-based comparison of strategies to prevent early-onset group B streptococcal disease in neonates. N Engl J Med 2002;347:233–239. 273. Dobson SR, Baker CJ. Enterococcal sepsis in neonates: features by age at onset and occurrence of focal infection. Pediatrics 1990;85:165–171. 274. McNeeley DF, Saint-Louis F, Noel GJ. Neonatal enterococcal bacteremia: an increasingly frequent event with potentially untreatable pathogens. Pediatr Infect Dis J 1996;15:800–805. 275. Luginbuhl LM, Rotbart HA, Facklam RR, et al. Neonatal enterococcal sepsis: case-control study and description of an outbreak. Pediatr Infect Dis J 1987;6:1022–1026. 276. Malik RK, Montecalvo MA, Reale MR, et al. Epidemiology and control of vancomycin-resistant enterococci in a regional neonatal intensive care unit. Pediatr Infect Dis J 1999;18:352–526. P.880 277. Geil CG, Castle WK, Mortimer EA. Group A streptococcal infections in newborn nurseries. Pediatrics 1970;46:849–854. 278. Lehtonen OP, Kero P, Ruuskanen O, et al. A nursery outbreak of group A streptococcal infection. J Infect 1987;14:263–270. 279. Campbell JR, Arango CA, Garcia-Prats JA, et al. An outbreak of M serotype group A Streptococcus in a neonatal intensive care unit. J Pediatr 1996;129:396–402. 280. Bortolussi R. Neonatal listeriosis. Semin Perinatol 1990;14(Suppl 1):44–48. 281. Melamed R, Greenberg D, Landau D, et al. Neonatal nosocomial pneumococcal infections acquired by patient-to-patient transmission. Scand J Infect Dis 2002;34:385–386. 282. Hart CA. Klebsiella and neonates. J Hosp Infect 1993;23:83–86. 283. Al-Rabea AA, Burwen DR, Eldeen MA, et al. Klebsiella pneumoniae bloodstream infections in neonates in a hospital in the Kingdom of Saudi Arabia. Infect Control Hosp Epidemiol 1998;19:674–679. 284. Macrae MB, Shannon KP, Rayner DM, et al. A simultaneous outbreak on a neonatal unit of two strains of multiply antibiotic resistant Klebsiella pneumoniae controllable only by ward closure. J Hosp Infect 2001;49:183–192. 285. Herruzo-Cabrera R, Garcia-Caballero J, Martin-Moreno JM, et al. Clinical assay of N-duopropenide alcohol solution on hand application in newborn and pediatric intensive care units: control of an outbreak of multiresistant Klebsiella pneumoniae in a newborn intensive care unit with this measure. Am J Infect Control 2001;29:162–167. 286. Hervas JA, Ballesteros F, Alomar A, et al. Increase of Enterobacter in neonatal sepsis: a twenty-two-year study. Pediatr Infect Dis J 2001;20:134–140. 287. Goering RV, Ehrenkranz J, Sanders CC, et al. Long term epidemiological analysis of Citrobacter diversus in a neonatal intensive care unit. Pediatr Infect Dis J 1992;11:99–104. 288. Finn A, Talbot GH, Anday E, et al. Vertical transmission of Citrobacter diversus from mother to infant. Pediatr Infect Dis J 1988;7:293–294. 289. Villari P, Crispino M, Salvadori A, et al. Molecular epidemiology of an outbreak of Serratia marcescens in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2001;22:630–634. 290. Leigh L, Stoll BJ, Rahman M, et al. Pseudomonas aeruginosa infection in very low birth weight infants: a case-control study. Pediatr Infect Dis J 1995;14:367–371. 291. Rapkin RH. Pseudomonas cepacia in an intensive care nursery. Pediatrics 1976;57:239–243. 292. Labarca JA, Trick WE, Peterson CL, et al. A multistate nosocomial outbreak of Ralstonia pickettii colonization associated with an intrinsically contaminated respiratory care solution. Clin Infect Dis 1999;29:1281–1286. 293. VanCowenbergh C, Cohen S. Analysis of epidemic and endemic isolates of Xanthomonas maltophilia by contour-clamped homogenous electric field gel electrophoresis. Infect Control Hosp Epidemiol 1994;15:691–696. 294. McDonald LC, Walker M, Carson L, et al. Outbreak of Acinetobacter spp. bloodstream infections in a nursery associated with contaminated aerosols and air conditioners. Pediatr Infect Dis J 1998;17:716–722. 295. Simor AE, Karmali MA, Jadavji T, et al. Abortion and perinatal sepsis associated with Campylobacter infection. Rev Infect Dis 1986;8:397–402. 296. Falla TJ, Dobson SRM, Crook DWM, et al. Population-based study of non-typable Haemophilus influenzae invasive disease in children and neonates. Lancet 1993;341:851–854. 297. Valenti WM, Pincus PH, Messner MK. Nosocomial pertussis: possible spread by a hospital visitor. Am J Dis Child 1980;134:520–521. 298. Linnemann CC, Ramundo N, Perlstein PH, et al. Use of pertussis vaccine in an epidemic involving hospital staff. Lancet 1975;2:540–543. 299. Levy I, Rubin LG. Legionella pneumonia in neonates: a literature review. J Perinatol 1998;18:287–290. 300. Franzin L, Scolfaro C, Cabodi D, et al. Legionella pneumophila pneumonia in a newborn after water birth: a new mode of transmission. Clin Infect Dis 2001;33:e103–e104. 301. Heggie AD, Bar-Shain D, Boxerbaum B, et al. Identification and quantification of ureaplasmas colonizing the respiratory tract and assessment of their role in the development of chronic lung disease in preterm infants. Pediatr Infect Dis J 2001;20:854–859. 302. Lee LH, LeVea CM, Graman PS. Congenital tuberculosis in a neonatal intensive care unit: case report, epidemiological investigation, and management of exposures. Clin Infect Dis 1998;27:474–477. 303. Saitoh M, Ichiba H, Fujioka H, et al. Connatal tuberculosis in an extremely low birth weight infant: case report and management of exposure to tuberculosis in a neonatal intensive care unit. Eur J Pediatr 2001;160:88–90. 304. Kossoff EH, Buescher ES, Karlowicz MG. Candidemia in a neonatal intensive care unit: trends during fifteen years and clinical features of 111 cases. Pediatr Infect Dis J 1998;17:504–508. 305. Saiman L, Ludington E, Dawson JD, et al. Risk factors for Candida species colonization of neonatal intensive care unit patients. Pediatr Infect Dis J 2001;20:1119–1124. 306. Weese-Mayer DE, Fondriest DW, Brouillette RT, et al. Risk factors associated with candidemia in the neonatal intensive care unit: a case-control study. Pediatr Infect Dis J 1987;6:190–196. 307. Leibovitz E, Iuster-Reicher A, Amitai M, et al. Systemic candidal infections associated with use of peripheral venous catheters in neonates: a 9-year experience. Clin Infect Dis 1992;14:485–491. 308. Botas CM, Kurlat I, Young SM, et al. Disseminated candidal infections and intravenous hydrocortisone in preterm infants. Pediatrics 1995;95:883–887. 309. Rowen JL, Rench MA, Kozinetz CA, et al. Endotracheal colonization with Candida enhances risk of systemic candidiasis in very low birth weight neonates. J Pediatr 1994;124:789–794. 310. Campbell JR, Zaccaria E, Baker CJ. Systemic candidiasis in extremely low birth weight infants receiving topical petrolatum ointment for skin care: a case-control study. Pediatrics 2000;105:1041–1045. 311. Faix RG. Invasive neonatal candidiasis: comparison of albicans and parapsilosis infections. Pediatr Infect Dis J 1992;11:88–93. 312. Fairchild KD, Tomkoria S, Sharp EC, et al. Neonatal Candida glabrata sepsis: clinical and laboratory features compared with other Candida species. Pediatr Infect Dis J 2002;21:39–43. 313. Yinnon AM, Woodin KA, Powell KR. Candida lusitaniae infection in the newborn: case report and review of the literature. Pediatr Infect Dis J 1992;11:878–880. 314. Saxen H, Virtanen M, Carlson P, et al. Neonatal Candida parapsilosis outbreak with a high case fatality rate. Pediatr Infect Dis J 1995;14:776–781. 315. Kaufman D, Boyle R, Hazen KC, et al. Fluconazole prophylaxis against fungal colonization and infection in preterm infants. N Engl J Med 2001;345:1660–1666. 316. Neely MN, Schreiber JR. Fluconazole prophylaxis in the very low birth weight infant: not ready for prime time. Pediatrics 2001;107:404–405. 317. Dankner WM, Spector SA, Fierer J, et al. Malassezia fungemia in neonates and adults: complication of hyperalimentation. Rev Infect Dis 1987;9:743–753. 318. Powell DA, Hayes J, Durrell DE, et al. Malassezia furfur skin colonization of infants hospitalized in intensive care units. J Pediatr 1987;111:217–220. 319. Gross GJ, MacDonald NE, Mackenzie AMR. Neonatal rectal colonization with Malassezia furfur. Can J Infect Dis 1992;3:9–13. 320. Richet HM, McNeil MM, Edwards MC, et al. Cluster of Malassezia furfur pulmonary infections in infants in a neonatal intensive-care unit. J Clin Microbiol 1989;27:1197–1200. 321. Welbel SF, McNeil MM, Pramanik A, et al. Nosocomial Malassezia pachydermatis bloodstream infections in a neonatal intensive care unit. Pediatr Infect Dis J 1994;13:104–108. 322. Chang HJ, Miller HL, Watkins N, et al. An epidemic of Malassezia pachydermatis in an intensive care nursery associated with colonization of health care workers' pet dogs. N Engl J Med 1998;338:706–711. 323. Salazar GE, Campbell JR. Trichosporonosis, an unusual fungal infection in neonates. Pediatr Infect Dis J 2002;21:161–165. 324. Aragao PA, Oshiro IC, Manrique EI, et al. Pichia anomala outbreak in a nursery: exogenous source? Pediatr Infect Dis J 2001;20:843–848. P.881 325. Chakrabarti A, Singh K, Narang A, et al. Outbreak of Pichia anomala infection in the pediatric service of a tertiary-care center in Northern India. J Clin Microbiol 2001;39:1702–1706. 326. Groll AH, Jaeger G, Allendorf A, et al. Invasive pulmonary aspergillosis in a critically ill neonate: case report and review of invasive aspergillosis during the first 3 months of life. Clin Infect Dis 1998;27:437–452. 327. Krasinski K, Holzman RS, Hanna B, et al. Nosocomial fungal infection during hospital renovation. Infect Control 1985;6:278–282. 328. Mitchell SJ, Gray J, Morgan MEI, et al. Nosocomial infection with Rhizopus microsporus in preterm infants: association with wooden tongue depressors. Lancet 1996;348:441–443. 329. Drusin LM, Ross BG, Rhodes KH, et al. Nosocomial ringworm in a neonatal intensive care unit: a nurse and her cat. Infect Control Hosp Epidemiol 2000;21:605–607. 330. Mossovitch M, Mossovitch B, Alkan M. Nosocomial dermatophytosis caused by Microsporum canis in a newborn department. Infect Control 1986;7:593–595. 331. Brady MT, Evans J, Cuartas J. Survival and disinfection of parainfluenza viruses on environmental surfaces. Am J Infect Control 1990;18:18–23. 332. Valenti WM, Clarke TA, Hall CB, et al. Concurrent outbreaks of rhinovirus and respiratory syncytial virus in an intensive care nursery: epidemiology and associated risk factors. J Pediatr 1982;100:722–726. 333. Snydman DR, Greer C, Meissner HC, et al. Prevention of nosocomial transmission of respiratory syncytial virus in a newborn nursery. Infect Control Hosp Epidemiol 1988;9:105–108. 334. Meissner HC, Murray SA, Kiernan MA, et al. A simultaneous outbreak of respiratory syncytial virus and parainfluenza virus type 3 in a newborn nursery. J Pediatr 1984;104:680–684. 335. Wilson CW, Stevenson DK, Arvin AM. A concurrent epidemic of respiratory syncytial virus and echovirus 7 infections in an intensive care nursery. Pediatr Infect Dis J 1989;8:24–29. 336. Singh-Naz N, Willy M, Riggs N. Outbreak of parainfluenza virus type 3 in a neonatal nursery. Pediatr Infect Dis J 1990;9:31–33. 337. Moisiuk SE, Robson D, Klass L, et al. Outbreak of parainfluenza virus type 3 in an intermediate care neonatal nursery. Pediatr Infect Dis J 1998;17:49–53. 338. Cunney RJ, Bialachowski A, Thornley D, et al. An outbreak of influenza A in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2000;21:449–454. 339. Sagrera X, Ginovart G, Raspall F, et al. Outbreaks of influenza A virus infection in neonatal intensive care units. Pediatr Infect Dis J 2002;1:196–200. 340. Abzug MJ, Levin MJ. Neonatal adenovirus infection: four patients and review of the literature. Pediatrics 1991;87:890–896. 341. Piedra PA, Kasel JA, Norton HJ, et al. Description of an adenovirus type 8 outbreak in hospitalized neonates born prematurely. Pediatr Infect Dis J 1992;11:460–465. 342. Modlin JF. Perinatal echovirus and group B coxsackievirus infections. Clin Perinatol 1988;15:233–246. 343. Rabkin CS, Telzak EE, Ho MS, et al. Outbreak of echovirus 11 infection in hospitalized neonates. Pediatr Infect Dis J 1988;7:186–190. 344. Chawareewong S, Kiangsiri S, Lokaphadhana K, et al. Neonatal herpangina caused by coxsackie A-5 virus. J Pediatr 1978;93:492–494. 345. Helin I, Widell A, Borulf S, et al. Outbreak of coxsackievirus A-14 meningitis among newborns in a maternity hospital ward. Acta Paediatr Scand 1987;76:234–238. 346. Sharma R, Hudak ML, Premachandra BR, et al. Clinical manifestations of rotavirus infection in the neonatal intensive care unit. Pediatr Infect Dis J 2002;21:1099–1105. 347. Watson JC, Fleming DW, Borella AJ, et al. Vertical transmission of hepatitis A resulting in an outbreak in a neonatal intensive care unit. J Infect Dis 1993;167:567–571. 348. Klein BS, Michaels JA, Rytel MW, et al. Nosocomial hepatitis A. A multinursery outbreak in Wisconsin. JAMA 1984;252:2716–2721. 349. American Academy of Pediatrics and American College of Obstetricians and Gynecologists. Perinatal infection. In: Gilstrap LC, Oh W, eds. Guidelines for perinatal care, 5th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2002:285–329. 350. Whitley RJ, Kimberlin DW. Treatment of viral infections during pregnancy and the neonatal period. Clin Perinatol 1997;24:267–283. 351. Brown ZA, Wald A, Morrow RA, et al. Effect of serologic status and cesarean delivery on transmission rates of herpes simplex virus from mother to infant. JAMA 2003;289:203–209. 352. Scott LL, Hollier LM, McIntyre D, et al. Acyclovir suppression to prevent clinical recurrences at delivery after first episode genital herpes in pregnancy: an open label trial. Infect Dis Obstet Gynecol 2001;9:75–80. 353. Hammerberg O, Watts J, Chernesky M, et al. An outbreak of herpes simplex virus type 1 in an intensive care nursery. Pediatr Infect Dis J 1983;2:290–294. 354. Sakaoka H, Saheki Y, Uzuki K, et al. Two outbreaks of herpes simplex virus type 1 nosocomial infection among newborns. J Clin Microbiol 1986;24:36–40. 355. Van Dyke RB, Spector SA. Transmission of herpes simplex virus type 1 to a newborn infant during endotracheal suctioning for meconium aspiration. Pediatr Infect Dis 1984;3:153–156. 356. Gershon AA, Raker R, Steinberg S, et al. Antibody to varicella-zoster virus in parturient women and their offspring during the first year of life. Pediatrics 1976;58:692–696. 357. Gustafson TL, Shehab Z, Brunell PA. Outbreak of varicella in a newborn intensive care nursery. Am J Dis Child 1984;138:548–550. 358. Friedman CA, Temple DM, Robbins KK, et al. Outbreak and control of varicella in a neonatal intensive care unit. Pediatr Infect Dis J 1994;13:152–154. 359. Huang YC, Lin TY, Lin YJ, et al. Prophylaxis of intravenous immunoglobulin and acyclovir in perinatal varicella. Eur J Pediatr 2001;160:91–94. 360. Demmler GJ. Summary of a workshop on surveillance for congenital cytomegalovirus disease. Rev Infect Dis 1991;13:315–329. 361. Yeager AS, Palumbo PE, Malachowski N, et al. Sequelae of maternally derived cytomegalovirus infections in premature infants. J Pediatr 1983;102:918–922. 362. Maschmann J, Hamprecht K, Dietz K, et al. Cytomegalovirus infection of extremely low-birth weight infants via breast milk. Clin Infect Dis 2001;33:1998–2003. 363. Spector SA. Transmission of cytomegalovirus among infants in hospital documented by restriction-endonuclease-digestion analysis. Lancet 1983;1:378–381. 364. Adler SP, Baggett J, Wilson M, et al. Molecular epidemiology of cytomegalovirus in a nursery: lack of evidence for nosocomial transmission. J Pediatr 1986;108:117–123. 365. American Academy of Pediatrics and American College of Obstetricians and Gynecologists. Infection control. In: Gilstrap LC, Oh W, eds. Guidelines for perinatal care, 5th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2002:331–353. 366. Centers for Disease Control and Prevention. U.S. Public Health Service Task Force recommendations for the use of antiretroviral drugs in pregnant HIV-1-infected women for maternal health and interventions to reduce perinatal HIV-1 transmission in the United States. MMWR Mortal Morb Wkly Rep 2002;51(no. RR-18):1–38. 367. Larson E. Rituals in infection control: what works in the newborn nursery? J Obstet Gynecol Neonatal Nurs 1987;16:411–416. 368. Horbar JD, Rogowski J, Plsek PE, et al. Collaborative quality improvement for neonatal intensive care. NIC/Q Project Investigators of the Vermont Oxford Network. Pediatrics 2001;107:14–22. 369. Kurlat I, Corral G, Oliveira F, et al. Infection control strategies in a neonatal intensive care unit in Argentina. J Hosp Infect 1998;40:149–154. 370. American Academy of Pediatrics and American College of Obstetricians and Gynecologists. Inpatient perinatal care services. In: Gilstrap LC, Oh W, eds. Guidelines for perinatal care, 5th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2002:17–55. 371. Bureau of Communicable Disease Epidemiology, Health Protection Branch and Health Services Directorate, Health Services and Promotion Branch. Infection control guidelines for perinatal care. Ottawa, Canada: Health and Welfare Canada, 1988. P.882 372. American Institute of Architects Academy of Architecture for Health. Guidelines for design and construction of hospital and health care facilities. Washington, DC: American Institute of Architects, 2001. 373. Bishop RF, Cameron DJS, Veenstra AA, et al. Diarrhea and rotavirus infection associated with differing regimens for postnatal care of newborn babies. J Clin Microbiol 1979;9:525–529. 374. Daschner F. Infectious hazards in rooming-in systems. J Perinatal Med 1984;12:3–6. 375. American Academy of Pediatrics. Active and passive immunization. In: Pickering LK, ed. 2003 red book: report of the Committee on Infectious Diseases, 26th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2003:1–98. 376. Campins-Marti M, Cheng HK, Forsyth K, et al. Recommendations are needed for adolescent and adult pertussis immunisation: rationale and strategies for consideration. Vaccine 2001;20:641–646. 377. Larson E. A causal link between handwashing and risk of infection? Examination of the evidence. Infect Control Hosp Epidemiol 1988;9:28–36. 378. De Carvalho M, Lopes JMA, Pellitteri M. Frequency and duration of handwashing in a neonatal intensive care unit. Pediatr Infect Dis J 1989;8:179–180. 379. Haque KN, Chagla AH. Do gowns prevent infection in neonatal intensive care units? J Hosp Infect 1989;14:159–162. 380. Brown J, Froese-Fretz A, Luckey D, et al. High rate of hand contamination and low rate of hand washing before infant contact in a neonatal intensive care unit. Pediatr Infect Dis J 1996;15:908–910. 381. Centers for Disease Control and Prevention. Guideline for hand hygiene in health-care settings: recommendations of the Healthcare Infection Control Practices Advisory Committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. MMWR Mortal Morb Wkly Rep 2002;51(no. RR-16):8–15. 382. Cloney DL, Donowitz LG. Overgown use for infection control in nurseries and neonatal intensive care units. Am J Dis Child 1986;140:680–683. 383. Birenbaum HJ, Glorioso L, Rosenberger C, et al. Gowning on a postpartum ward fails to decrease colonization in the newborn infant. Am J Dis Child 1990;144:1031–1033. 384. Rush J, Fiorino-Chiovitti R, Kaufman K, et al. A randomized controlled trial of a nursery ritual: wearing covergowns to care for healthy newborns. Birth 1990;17:25–30. 385. Pelke S, Ching D, Easa D, et al. Gowning does not affect colonization or infection rates in a neonatal intensive care unit. Arch Pediatr Adolesc Med 1994;148:1016–1020. 386. Agbayani M, Rosenfeld W, Evans H, et al. Evaluation of modified gowning procedures in a neonatal intensive care unit. Am J Dis Child 1981;135:650–652. 387. Health Canada, Steering Committee on Infection Control Guidelines. Infection control guidelines. Routine practices and additional precautions for preventing the transmission of infection in health care. Can Commun Dis Rep 1999;25S4:1–142. 388. American Academy of Pediatrics and American College of Obstetricians and Gynecologists. Care of the neonate. In: Gilstrap LC, Oh W, eds. Guidelines for perinatal care, 5th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2002:187–235. 389. Schwab F, Tolbert B, Bagnato S, et al. Sibling visiting in a neonatal intensive care unit. Pediatrics 1983;71:835–838. 390. Maloney MJ, Ballard JL, Hollister L, et al. A prospective controlled study of scheduled sibling visits to a newborn intensive care unit. J Am Acad Child Psychiatry 1983;22:565–570. 391. American Academy of Pediatrics. Recommendations for care of children in special circumstances. In: Pickering LK, ed. 2003 red book: report of the Committee on Infectious Diseases, 26th ed. Elk Grove Village, IL: American Academy of Pediatrics, 2003:99–188. 392. Rutala WA. APIC guideline for selection and use of disinfectants. Am J Infect Control 1996;24:313–342. 393. Meyer CL, Eitzen HE, Schreiner RL, et al. Should linen in newborn intensive care units be autoclaved? Pediatrics 1981;67:362–364. 394. Darmstadt GL, Dinulos JG. Neonatal skin care. Pediatr Clin North Am 2000;47:757–782. 395. Shuman RM, Leech RW, Alvord EC. Neurotoxicity of hexachlorophene in the human. I. A clinicopathologic study of 248 children. Pediatrics 1974;54:689–695. 396. Dixon RE, Kaslow RA, Mallison GF, et al. Staphylococcal disease outbreaks in hospital nurseries in the United States-December 1971 through March 1972. Pediatrics 1973;51:413–417. 397. Husak M, Wiltshire J, Carr H, et al. Effect of Hibiclens bathing on neonatal bactericidal colonization [Abstract 714]. Twenty-first Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, 1981. 398. Johnsson J, Seeberg S, Kjellmer I. Blood concentrations of chlorhexidine in neonates undergoing routine cord care with 4% chlorhexidine gluconate solution. Acta Paediatr Scand 1987;76:675–676. 399. Cowen J, Ellis SH, McAinsh J. Absorption of chlorhexidine from the intact skin of newborn infants. Arch Dis Child 1979;54:379–383. 400. Aggett PJ, Cooper LV, Ellis SH, et al. Percutaneous absorption of chlorhexidine in neonatal cord care. Arch Dis Child 1981;56:878–891. 401. Lund CH, Osborne JW, Kuller J, et al. Neonatal skin care: clinical outcomes of the AWHONN/NANN evidence-based clinical practice guideline. J Obstet Gynecol Neonatal Nurs 2001;30:41–51. 402. Nopper AJ, Horii KA, Sookdeo-Drost S, et al. Topical ointment therapy benefits premature infants. J Pediatr 1996;128:660–669. 403. Janssen PA, Selwood BL, Dobson SR, et al. To dye or not to dye: a randomized, clinical trial of a triple dye/alcohol regime versus dry cord care. Pediatrics 2003;111;15–20. 404. Smales O. A comparison of umbilical cord treatment in the control of superficial infections. N Z Med J 1988;101:453–455. 405. Arnold LDW, Larson E. Immunologic benefits of breast milk in relation to human milk banking. Am J Infect Control 1993;21:235–242. 406. Furman L, Taylor G, Minich N, et al. The effect of maternal milk on neonatal morbidity of very low-birth-weight infants. Arch Pediatr Adolesc Med 2003;157:66–71. 407. Larson E, Zuill R, Zier V, Berg B. Storage of human breast milk. Infect Control 1984;5:127–130. 408. Canadian Paediatric Society. Human milk banking and storage. Paediatr Child Health 1996;1:141–145. 409. Barry C, Verncombe M. Is the right breast milk being fed to infants? Can J Infect Control 1998;13:16–19. 410. American Dietetic Association. Preparation of formula for infants: guidelines for health care facilities. Chicago: American Dietetic Association, 1991. 411. Garland JS, Henrickson K, Maki DG. The 2002 Hospital Infection Control Practices Advisory Committee Centers for Disease Control and Prevention guideline for prevention of intravascular device-related infection. Pediatrics 2002;110:1009–1013. 412. Garland JS, Alex CP, Mueller CD, et al. A randomized trial comparing povidone-iodine to a chlorhexidine gluconate-impregnated dressing for prevention of central venous catheter infections in neonates. Pediatrics 2001;107:1431–1436. 413. Levy GJ, Strauss RG, Hume H, et al. National survey of neonatal transfusion practices. I. Red blood cell therapy. Pediatrics 1993;91:523–529. 414. Canadian Paediatric Society. Red blood cell transfusions in newborn infants: revised guidelines. Paediatr Child Health 2002;7:553–558. 415. Tierney AJ, Higa TE, Finer NN. Disseminated cytomegalovirus infection after extracorporeal membrane oxygenation. Pediatr Infect Dis J 1992;11:241–243. 416. Evans ME, Schaffner W, Federspiel CF, et al. Sensitivity, specificity, and positive predictive value of body surface cultures in a neonatal intensive care unit. JAMA 1988;259:248–252. 417. Fulginiti VA. Ray CG. Body surface cultures in the newborn infant. An exercise in futility, wastefulness, and inappropriate practice. Am J Dis Child 1988;142:19–20. 418. Lau YL, Hey E. Sensitivity and specificity of daily tracheal aspirate cultures in predicting organisms causing bacteremia in ventilated neonates. Pediatr Infect Dis J 1991;10:290–294. 419. Garner JS, Hospital Infection Control Practices Advisory Committee. Guideline for isolation precautions in hospitals. Infect Control Hosp Epidemiol 1996;17:53–80. P.883 420. Leclair JM, Freeman J, Sullivan BF, et al. Prevention of nosocomial respiratory syncytial virus infections through compliance with glove and gown isolation precautions. N Engl J Med 1987;317:329–334. 421. Gala CL, Hall CB, Schnabel KC, et al. The use of eye-nose goggles to control nosocomial respiratory syncytial virus infection. JAMA 1986;256:2706–2708. 422. Haas JP, Trezza LA. Outbreak investigation in a neonatal intensive care unit. Semin Perinatol 2002;26:367–378. 423. Ehrenkrantz NJ, Sanders CG, Eckert-Schollenberger D, et al. Lack of evidence of efficacy of cohorting nursing personnel in a neonatal intensive care unit to prevent contact spread of bacteria: an experimental study. Pediatr Infect Dis J 1992;11:105–113. 424. Wu F, Della-Latta P. Molecular typing strategies. Semin Perinatol 2002;26:357–366. 425. Sprunt K, Leidy G, Redman W. Abnormal colonization of neonates in an ICU: conversion to normal colonization by pharyngeal implantation of alpha hemolytic streptococcus strain 215. Pediatr Res 1980;14:308–313. 426. Cook LN, Davis RS, Stover, BH. Outbreak of amikacin-resistant Enterobacteriaceae in an intensive care nursery. Pediatrics 1980;65:264–268. 427. Reuman PD, Duckworth DH, Smith KL, et al. Lack of effect of Lactobacillus on gastrointestinal bacterial colonization in premature infants. Pediatr Infect Dis J 1986;5:663–668. 428. Millar MR, Bacon C, Smith SL, et al. Enteral feeding of premature infants with Lactobacillus GG. Arch Dis Child 1993;69:483–487. 429. Rastegar Lari A, Gold F, Borderon JC, et al. Implantation and in vivo antagonistic effects of antibiotic-susceptible Escherichia coli strains administered to premature newborns. Biol Neonate 1990;58:73–78. 430. LodinovГЎ-ZГЎdnГЎkovГЎ R, TlaskalovГЎ H, BartГЎkovГЎ Z. The antibody response in infants after colonization of the intestine with E. coli O83. Artificial colonization used as a prevention against nosocomial infections. Adv Exp Med Biol 1991;310:329–335. 431. Khalak R, Pichichero ME, D'Angio CT. Three-year follow-up of vaccine response in extremely premature infants. Pediatrics 1998;101:597–603. 432. Blake KD, Jellinek DC. Visiting and immunization policies in a regional neonatal unit. Lancet 1990;336:308–309. 433. Patel DM, Butler J, Feldman S, et al. Immunogenicity of hepatitis B vaccine in healthy very low birth weight infants. J Pediatr 1997;131:641–643. 434. Shinefield H, Black S, Ray P, et al. Efficacy, immunogenicity and safety of heptavalent pneumococcal conjugate vaccine in low birth weight and preterm infants. Pediatr Infect Dis J 2002;21:182–186. 435. Munoz FM, Englund JA. A step ahead. Infant protection through maternal immunization. Pediatr Clin North Am 2000;47:449–463. 436. Perez EM, Weisman LE. Novel approaches to the prevention and therapy of neonatal bacterial sepsis. Clin Perinatol 1997;24:213–229. 437. Hill HR. Additional confirmation of the lack of effect of intravenous immunoglobulin in the prevention of neonatal infection. J Pediatr 2000;137:595–597. 438. Parravicini E, van de Ven C, Anderson L, et al. Myeloid hematopoietic growth factors and their role in prevention and/or treatment of neonatal sepsis. Transfus Med Rev 2002;16:11–24.
Воскресенье, 06 Ноябрь 2011 | 1446 Просмотры | Печать | PDF
The obstetric unit in a hospital is one of the first examples of a characteristic of the modern hospital—a specialized patient population in a specialized hospital unit. “Lying-in” or obstetric hospitals were introduced in the eighteenth century, and some of the first significant studies on the epidemiology of hospital-acquired infections were made on obstetric services (1). Today's hospital is increasingly becoming a collection of specialized units containing unique patient populations characterized by specific disease entities and undergoing specialized treatments that either make the host more vulnerable to infection or facilitate transmission of infections. Obstetric care differs from other hospital infection problems in that both the primary patient and the newborn infant are at risk. A recent survey of postpartum infections reported an overall infection rate of 6%, as shown in Table 55.1 (2). Ledger (3) pointed out that the frequent empiric use of antibiotics for febrile patients on the obstetric service probably leads to lower reported infection rates and obscures the true frequency of hospital-acquired infections. The most important change in obstetric care in recent decades has been the increasingly shorter hospital stays mandated by cost concerns in managed care systems. Shortened hospital stays may decrease some risk factors for hospital-acquired infections, such as the use of catheters and intravenous lines, and may result in further decreases in infection rates. Unfortunately, shortened hospital stays also make surveillance of infections more difficult; thus, changes in infection rates cannot be confirmed by current data. TABLE 55.1. INFECTION RATES (CASES/100 DELIVERIES) ON OBSTETRIC SERVICES BY SITE OF INFECTION, 1993–1995 Type of delivery Site of infection UTI SSI Epi End Mast All sites Cesarean 1.1 3.4 NA 0.8 1.7 7.4 Vaginal 2.0 NA 0.3 0.2 3.0 5.5 End, endometritis; EPI, episiotomy; Mast, mastitis; SSI, surgical site infection (excluding endometritis); UTI, urinary tract infection. From Yokre DS, Christiansen CL, Johnson R, et al. Epidemiology of and surveillance for post partum infections. Emerg Infect Dis 2001;7:837–841, with permission HISTORY OF HOSPITAL-ACQUIRED OBSTETRIC INFECTIONS The establishment of obstetric hospitals in the mid-eighteenth century created the setting for epidemics of puerperal infections (1). The epidemics in turn provided the opportunity to demonstrate that the infections were contagious and to develop prevention methods. Alexander Gordon was one of the first to document this (4) and, later, so did Oliver Wendell Holmes (5), but the most famous was Ignaz Semmelweis in Vienna because of his extensive and carefully detailed observations (6). The “great free Vienna Lying-in Hospital” created a natural epidemiologic experiment that Semmelweis had the insight to appreciate. The hospital had two separate divisions: the First Division for teaching medical students and the Second Division for teaching midwives. The mortality was so much greater in the First Division that even the patients knew about it and tried to be admitted to the Second Division if at all possible. Semmelweis took advantage of this natural experiment to carefully collect data to document and determine the cause of the epidemics. Not only did he evaluate the data on hospital-acquired infections but also he made anecdotal observations that supported his conclusions, such as the low infection rate in women who delivered in the street on the way to the hospital as compared with those who delivered in the hospital. To explain the increased mortality in the First Division, he noted the similarity of the fatal illness in a pathologist who had been stuck in the finger by a medical student during an autopsy and the fatal infections in the obstetric patients. He concluded that material from the autopsies was being transmitted back to the patients and causing their illnesses. Hand washing with soap was done after autopsies, and thus he concluded that this was inadequate to remove all “cadaveric particles.” He added hand rinsing with chlorinated lime water after performing autopsies and after each patient contact (7). The result was a dramatic decrease in mortality in the First Division to rates similar to that of the Second Division. It took decades before Semmelweis' ideas became standard P.928 practice, stimulated by Lister's concept of antisepsis. Even then, obstetric infections and maternal mortality remained major problems into the 1930s. The appearance of the first antimicrobial agents and improvements in other aspects of obstetric care resulted in a major decrease in maternal mortality (8). Presumably, the reason for the persistence of high obstetric infection rates into the 1900s was that even though epidemics of puerperal fever transmitted by cross-infections were prevented by hand washing and antiseptic techniques, infection from patients' endogenous flora remained a problem. PATHOGENESIS OF OBSTETRIC INFECTIONS Most obstetric infections are caused by maternal vaginal and cervical flora, and, thus, increases in infections usually relate to risk factors other than the introduction of exogenous pathogens. Hospital pathogens are seldom problems on obstetric services. This is because obstetric patients have short stays, and obstetric units are separated from other hospital units. As Bartlett et al. (9, 10) pointed out, vaginal flora is a dynamic ecosystem, with some differences between vaginal and cervical flora. Anaerobic bacteria usually outnumber aerobes, with anaerobic and facultative lactobacilli predominating. Other anaerobes include Peptostreptococcus species, Bacteroides species, and Prevotella species. The aerobic gram-positive flora include coagulase-negative staphylococci, with varying amounts of streptococci, enterococci, and Staphylococcus aureus, and the gram-negative flora include Escherichia coli, Gardnerella vaginalis, Enterobacter species, Klebsiella pneumoniae, and Proteus mirabilis (11, 12). Both Mycoplasma and Ureaplasma are also found in the vagina. Sexually transmitted diseases may add Neisseria gonorrhoeae, Chlamydia trachomatis, or herpes simplex virus (HSV) to this flora. Vaginal flora may change during pregnancy. Some studies suggested that lactobacilli increase in pregnancy and that other anaerobes decrease (13). Antibiotics also change the flora, and the use of multiple doses of cephalosporins for prophylaxis has been reported to increase enterococci and perhaps Enterobacter species (14). As would be expected, most obstetric intrauterine infections are polymicrobial, representing contiguous spread from the vagina (15). Ascending infection is demonstrated both by the presence of routine cervicovaginal flora in endometritis and surgical site infections (SSIs) and the ability to predict postcesarean infections by intraoperative lower uterine cultures (16). Although a significant proportion of postcesarean SSIs is caused by staphylococci, as in other SSIs, most postcesarean infections are caused by endometrial contamination (17). Many risk factors have been proven or suggested to be associated with endometrial infection, including those that increase entry of vaginal flora such as prolonged rupture of the membranes, frequent vaginal and rectal examinations, or intrauterine monitoring; those that result in tissue injury such as soft tissue trauma, midforceps delivery, or inexperienced surgeons; and host factors that are associated with increased infections such as maternal age, lower socioeconomic status, or obesity. Understanding these risk factors may become important in evaluating infection rates on specific obstetric units. OBSTETRIC INFECTIONS (INFECTIONS RELATED TO PREGNANCY AND DELIVERY) Postpartum Endometritis The classic obstetric infection is postpartum endometritis. The postpartum patient develops fever that may be associated with abdominal pain, uterine tenderness, or foul-smelling discharge and in most cases is started on antibiotics without obtaining cultures (15). Endometritis can occur after either vaginal delivery or cesarean section but is diagnosed most often after cesarean section. Infection has been reported to occur after fewer than 3% of vaginal deliveries and 5% to 95% of cesarean sections (2, 15, 18). The variation in infection rates results from the prevalence of risk factors in the population studied and the way that patients are managed. Endometritis after cesarean section occurs earlier than after vaginal delivery, as shown by hospital readmissions for postpartum endometritis. Most women who were readmitted for endometritis had delivered vaginally (19). Infections may be caused by a single bacterial species but are often polymicrobial (20). Common etiologic agents include the gram-positive cocci such as streptococci and enterococci; gram-negative bacilli such as E. coli, K. Pneumoniae, and P. mirabilis; and anaerobes such as Bacteroides bivius and peptostreptococci (15, 20, 21). Some studies have distinguished between early and late postpartum endometritis; late infection is a milder disease that occurs after vaginal delivery (22, 23). It has been suggested that genital mycoplasma and Chlamydia are important etiologic agents in late endometritis and that erythromycin may be effective therapy. The most important risk factor for postpartum endometritis is cesarean section, and the risk of infection is greatest when it is a nonelective procedure after rupture of the membranes and the onset of labor (15, 18, 24). General anesthesia, long duration of surgery, intraoperative problems, and poor surgical technique may all be risk factors. Patients undergoing nonelective cesarean sections are routinely given prophylactic antibiotics, because this has been shown to reduce infection rates by 50% or more (24). In vaginal deliveries, prolonged rupture of the membranes, midforceps delivery, and soft tissue trauma increase the risk. With many other risk factors for postpartum endometritis, it is difficult to separate out relative risks, because the factors are interrelated. This applies to risk factors such as prolonged labor, frequent vaginal examinations, and internal monitoring. Host factors that increase risk include bacterial vaginosis, human immunodeficiency virus (HIV) infection, anemia, low socioeconomic status, maternal age, and obesity (24, 25, 26). Infection of the endometrium may extend into the myometrium and parametrial tissue, with abscess formation or sepsis. One clinical complication that should be suspected in a patient who does not respond to antibiotic therapy is septic pelvic thrombophlebitis (27). If the workup fails to identify another infection, computed tomography should be performed and a therapeutic trial of anticoagulation considered. In such cases, antibiotics would be continued in addition to heparinization, and a rapid defervescence may be observed. Laboratory evaluation of the febrile postpartum patient should include a complete blood count, chest x-ray, urine cul P.929 ture, and blood cultures. Leukocytosis is usually present but may also be seen in the noninfected postpartum patient. Uterine cultures are often not done because of the difficulty in interpreting the results. Because the microorganisms recovered are usually part of the normal maternal flora, these may either represent contamination during specimen collection or may be the cause of endometritis. Unless a blood culture is positive, there is no way to confirm that the isolates are significant. However, good aerobic and anaerobic cultures do show the range of potential pathogens and detect infections caused by unusual pathogens such as the rare group A ОІ-hemolytic streptococci (GABHS) infection. Uterine cultures can be collected with a cotton swab (28). Historically, GABHS has been a significant cause of postpartum endometritis but now is uncommon (3). These infections occur in previously colonized mothers or can be acquired by cross-infection from healthcare workers, other patients, or colonized infants. GABHS endometritis may differ from endometritis caused by the usual maternal vaginal flora, with an abrupt onset of high spiking fevers and diffuse tenderness. Diagnosis can be made by Gram stain and cultures of uterine discharge. The streptococcal toxic shock syndrome caused by GABHS has been reported in postpartum patients (29). Surgical Site Infections Episiotomy infections are uncommon and usually not serious, but severe complications such as necrotizing fasciitis can develop (30, 31). Episiotomy sites should be examined carefully to detect infection early and infections should be treated to prevent complications. A more serious problem is infection of the surgical site of a cesarean section. SSIs are reported to occur in about 3% to 4% of cesarean section patients, including both incisional and organ and space infections (2, 32). SSIs are usually caused by maternal flora in the endometrium but, as with any other surgical sites, can be caused by organisms from exogenous sources (17, 33). In the latter cases, S. aureus is the most frequent cause of infection. Although the pathogenicity of genital mycoplasma in SSIs has not been proved, a recent study reported these to be the most common bacteria isolated in infected postcesarean surgical sites (33). SSIs should be cultured before antibiotic therapy is begun. Urinary Tract Infection Urinary tract infections are a common problem in pregnancy and during the postpartum period (34). Risk factors for postpartum infections include urinary retention from anesthesia, trauma during delivery, and the need for catheterization. Urine cultures of the febrile patient should always be collected, although midstream samples may be contaminated by vaginal discharge. In those cases, the results are interpreted in the context of the clinical findings and the response to empiric antibiotic therapy. The major preventable risk factor in the postpartum period is catheterization. This is necessary with urinary retention but should be done only as needed. Another risk factor for postpartum urinary tract infection is bacteriuria during pregnancy. Detection and treatment of bacteriuria in pregnant women may decrease postpartum urinary tract infections. Chorioamnionitis (Intraamniotic Infection) Intrauterine infection during pregnancy, like postpartum endometritis, is usually caused by ascending infection from vaginal flora and is caused by similar bacteria (15, 35, 36). Most infections are also polymicrobial, and the major risk factor is prolonged rupture of the membranes. Infection is rare in women with intact membranes. Other risk factors are similar to those for postpartum endometritis, including duration of labor, number of vaginal examinations, internal monitoring, and possibly bacterial vaginosis. A variety of other obstetric procedures may introduce infection such as amniocentesis, chorionic villus sampling, and percutaneous umbilical blood sampling. Initially, the diagnosis may be difficult, because fever can be the only presenting sign. Specific diagnosis requires examination of amniotic fluid by Gram stain, culture, and amniotic fluid glucose level (37). Hospital-acquired chorioamnionitis can be suspected in patients who become febrile after vaginal examinations, internal fetal monitoring, or other such procedures, but there is no standardized definition for nosocomial infection. Once the diagnosis is suspected, the patient should be started on antibiotic therapy and delivered as soon as possible. Mastitis In a study of obstetric patients who were contacted after discharge from the hospital, mastitis was the most common infection reported (38). Very few breast infections were seen during hospitalization, because mastitis and breast abscess usually occur several weeks into the postpartum period. A slight fever can develop early with breast engorgement, but it is transient. Later in the postpartum period, infectious mastitis must be distinguished from milk stasis and noninfectious inflammation (39). Infection is associated with higher fevers and erythema and is usually unilateral. The most common cause of breast infection is S. aureus (40). Epidemics of staphylococcal mastitis occurred in the past but have not been reported in recent years. Therefore, the traditional classification of infectious mastitis into sporadic and epidemic forms is seldom useful. Both types are usually caused by S. aureus. Predisposing factors include the lack of nipple care, poor feeding technique, and inadequate emptying of the breasts. Infection can be confirmed by Gram stain and culture and responds to antistaphylococcal antibiotics and, if needed, surgical drainage. Continued breast drainage is important and can be accomplished by continued nursing, if appropriate, or pumping and discarding milk. NONOBSTETRIC INFECTIONS IN THE OBSTETRIC PATIENT There are many nonobstetric infections that must be considered in the evaluation and management of obstetric patients, P.930 not only for the sake of the patient but also for the safety of the fetus or neonate and the protection of others on the obstetric service. Selected infections of particular importance in the obstetric patient are described. Listeria Listeria monocytogenes causes a febrile illness in obstetric patients and may rarely result in severe diseases such as meningitis (41). It can be transmitted to the neonate and cause severe disease. Contaminated food, particularly soft cheeses and cold meats, can infect the obstetric patient. Transmission of Listeria to neonates in the delivery room has been reported on several occasions (42, 43). The collection of routine blood cultures during the evaluation of fever provides a diagnosis, allowing directed antibiotic therapy. Group B Streptococcal Infection Group B ОІ-hemolytic streptococci (GBS) are normal flora in the gastrointestinal and genitourinary tract and occasionally cause obstetric infections, including chorioamnionitis, endometritis, urinary tract infections, or SSIs (44). More often, the colonized mother may transmit GBS to the neonate, sometimes causing neonatal sepsis. The Centers for Disease Control and Prevention (CDC) recommends universal prenatal screening of pregnant women for GBS colonization of vagina and rectum (45). Antibiotic prophylaxis is recommended for colonized women delivering vaginally or who have had rupture of membranes (45). (See also Chapter 31.) Human Immunodeficiency Virus Infection One of the most important infections to identify in the pregnant patient is HIV, because prepartum diagnosis allows preventive treatment with antiretroviral drugs and possibly delivery by cesarean section (46). HIV antibody testing is offered to all pregnant patients, and if positive they should be treated according to current Public Health Service Task Force guidelines (46), which are frequently updated. HIV-infected mothers should not breast-feed their children. Hepatitis B Virus Infections All pregnant women should be tested for hepatitis B virus (HBV) surface antigen [and for hepatitis C virus (HCV) antibody if liver function tests are abnormal] to prevent transmission of HBV infection to neonates. If HBV infection is identified in the obstetric patient, the neonate can be treated with HBV immunoglobulin and vaccine (47). Healthcare workers with HBV also pose a risk to uninfected obstetric patients during high-risk procedures. Obstetricians with HBV have been reported to infect their patients during cesarean section and forceps deliveries (48). Every obstetrician should know his or her HBV status, including tests for HBV surface antigen, e antigen and antibody, and core antibody. If susceptible to HBV, obstetricians should be immunized (see also Chapters 45 and 78). Hepatitis C Virus Infection No prophylaxis is currently available to prevent transmission of HCV from infected mother to neonate. Herpes Simplex Virus Infection Genital HSV, both primary and recurrent infection, occurs in obstetric patients and on rare occasion may result in disseminated disease (49). HSV can be transmitted from mother to neonate intrapartum. Cesarean section is indicated for women with active HSV lesions at the time of delivery (50). Internal fetal monitoring should not be done if HSV is suspected (50). Postpartum, the mother with HSV lesions should be advised of potential risks of transmission to her newborn and educated about appropriate measures to limit contact transmission. (50) If healthcare workers with active lesions are allowed to continue working with patients, similar measures should be taken (51) (see also Chapter 43). Chickenpox (Varicella Zoster Virus Infection) Chickenpox in the obstetric patient may result in severe pneumonia, requiring hospitalization and antiviral therapy (52). The obstetric patient with chickenpox who is not in labor should be admitted to a nonobstetric unit and placed in a negative pressure room, because airborne transmission can occur (53, 54). A patient admitted to our obstetric unit and placed in a regular hospital room with the door closed still infected a susceptible nurse who walked past the closed door (54). After delivery of a pregnant patient with chickenpox, the mother and infant should be separated until all of the mother's lesions have crusted over (see also Chapters 42, 51, and 52). Smallpox Vaccine (Vaccinia Virus) The United States began smallpox vaccination of healthcare workers in 2003. Hospitals will need to ensure that vaccinated healthcare workers do not transmit vaccinia to pregnant women, because fetal vaccinia can occur (55). Rubella Virus Despite the availability of an effective vaccine, rubella outbreaks and congenital rubella continue to occur. Pregnant women should be screened for rubella antibody. If not immune, they should be advised and monitored for signs and symptoms of rubella (56) (see also Chapter 51). INFECTION CONTROL PROGRAM FOR OBSTETRICS Surveillance There are limited surveillance data available on hospital-acquired infections on obstetric units. The National Nosocomial Infections Surveillance (NNIS) system and other organizations P.931 do provide some data that are useful to hospitals doing surveillance on obstetric services (32). As would be expected, these reports show that SSIs are the major nosocomial infection problem in obstetrics, with urinary tract infections the next most frequent. However, the reported infection rates on obstetric services are lower than on medical and surgical services. The infection rates vary by the size and type of hospital, and the SSIs vary by the number of risk factors. More recent NNIS reports are limited to data on SSIs from in-hospital surveillance of patients who have had cesarean sections (32). None of these data provides a basis for comparing infection rates observed in routine surveillance with data from other hospitals (57). The best comparison is with rates in the same hospital at previous times. The CDC does provide standardized definitions for surveillance (Table 55.2) (58, 59). As discussed previously, many risk factors have been identified for obstetric infections. However, for most surveillance purposes it is sufficient to relate infections to the type of delivery, vaginal or cesarean, and to distinguish between elective and nonelective cesarean sections. TABLE 55.2. DEFINITIONS FOR SURVEILLANCE OF NOSOCOMIAL INFECTIONS ON OBSTETRIC UNITS: CENTERS FOR DISEASES CONTROL AND PREVENTION Endometritis must meet either of the following criteria: Microorganism isolated from culture of fluid or tissue from endometrium obtained during surgery, by needle aspiration, or by brush biopsy Purulent drainage from uterus and two of the following: fever (>38ВC), abdominal pain, or uterine tenderness Episiotomy site infection must meet either of the following criteria: Purulent drainage from episiotomy Episiotomy abscess Other infections (excluding surgical site infections) must meet one of the following criteria: Microorganism isolated from culture of tissue or fluid from affected site Abscess or other evidence of infection seen during surgery or by histopathologic examination Two of the following: fever (>38ВC), nausea, vomiting, pain, tenderness, dysuria, and either of the following: Microorganism isolated from blood culture Physician's diagnosis Postcesarean surgical site infections must meet the definitions used for all surgical site infections and are classified into the following categories: Superficial incisional: involves only skin or subcutaneous tissue and excludes stitch abscess or an episiotomy infection Deep incisional: involves deep tissues, e.g., fascial or muscle layers Organ/space: involves any part of the anatomy, other than an incision, opened or manipulated during surgery and includes postoperative endometritis From Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128–140; Horan TC, Gaynes RP, Martone WJ, et al. CDC definitions of nosocomial Surgical site infection, 1992; a modification of CDC definitions of surgical wound infections. Am J Infect Control 1992;20:271–274, with permission. The general value of surveillance and infection control programs in hospitals has been documented by the CDC Study on the Efficacy of Nosocomial Infection Control, and the effective use of surveillance data on an obstetrics and gynecology service has been demonstrated by a study at a Swedish hospital (60, 61). In the Swedish report, data collected on patients having cesarean sections showed that 15% of them were infected (urinary tract infections excluded). The infection rates decreased to 9% after the introduction of quarterly surveillance reports to obstetric personnel. These reports included surgeon-specific infection rates. This decrease was mainly in postpartum endometritis. Despite this success with surveillance of postcesarean infections, developing an overall obstetric surveillance program may be difficult. A traditional method of surveillance on obstetric services is to monitor fevers in all patients. Most infected patients will be detected by this approach, and a routine fever workup in the postpartum patient will identify many infectious causes. The limitation of fever surveillance is that half of the fevers are either due to noninfectious causes or the specific infection cannot be identified (62). Despite this limitation, fever surveillance on an obstetric unit is a good screening technique and can indicate the development of potential problems. Mead et al. (63) reported the use of a “sentinel list” technique on an obstetric unit, where the bedside nurse is involved in collecting information including fever and antimicrobial therapy and the collected information is reviewed for continuous surveillance. This method may be implemented and maintained by the obstetric staff, who would report to infection control when problems develop. Another technique to identify infections is to link computer data on patients who have cesarean sections with antibiotic utilization records and admission and discharge diagnoses (64). Short hospital stays and outpatient management of most postdischarge infections limit hospital-based surveillance of obstetric patients. Supplemental postdischarge surveillance systems are needed to provide an accurate picture of obstetric infections. Several different approaches to postdischarge surveillance have been tried, involving either the patients directly or their physicians. The gold standard of postdischarge surveillance is direct observation of patients after hospital discharge. Couto et al. (65) did this at a Brazilian hospital by having postcesarean patients return on the tenth to the fifteenth postoperative day (65). While in the hospital, 1.6% of the patients had SSIs, and this increased to 9.6% with postdischarge examination. A more practical approach was used by Holbrook et al. (38) who mailed one-page questionnaires to 19,650 women who delivered at their hospital. They received responses from only 36% of them (38). Ten percent of the patients who responded reported infections after discharge, including mastitis (6%), urinary tract infections (3%), and endometritis (1%). Postdischarge surveillance detected twice as many infections as in-hospital surveillance. The additional infections that were identified were mostly mastitis and urinary tract infections. Most cases of endometritis were reported by in-hospital surveillance, but an additional 1% of women reported endometritis after discharge. A major limitation of this approach is the poor response rate by the patients. Postdischarge surveillance using physician questionnaires to identify infections after cesarean section has been reported to be more successful (66). In a study by Hulton et al. (66), 90% of physicians completed questionnaires about their patients. These questionnaires indicated an infection rate of 6.3% compared with 1.6% observed by in-hospital surveillance before the intro P.932 duction of the postdischarge system. The increase occurred in incisional SSIs (0.3% to 3.9%) and endometritis (1.3% to 2.5%). Despite the success of this study, maintaining physician cooperation over a long period of time may be difficult. The low infection rates in obstetric patients may not warrant the resources needed for postdischarge surveillance. A reasonable approach is to do postdischarge surveillance for a limited period each year (67). Data mining of computerized records can augment traditional surveillance. Yokoe et al. (2) have used this method to determine postpartum infection rates among women in a large managed care organization (see also Chapter 94). Facilities on an Obstetric Unit Obstetric units vary greatly in design, ranging from birthing centers designed for low-risk deliveries to standard labor and delivery units including operating rooms for cesarean sections. The design needs are similar to other patient care areas in the hospital, including conveniently placed alcohol hand rubs and sinks for hand disinfection by staff, easily cleaned surfaces, and, in the case of complicated deliveries, a fully equipped operating room. The American College of Obstetricians and Gynecologists outlined basic standards for obstetric facilities (68). An isolation facility should be available for the rare delivery of an obstetric patient with airborne infectious diseases such as chickenpox or tuberculosis. A pregnant patient with such an infection who is not in labor can be isolated in a room with negative air pressure on other hospital units. The use of hydrotherapy to assist in labor raises additional environmental infection control concerns (69). Some obstetric units use baths, whirlpools, or Jacuzzi showers as an aid in delivery. This practice raises the same concerns about bacterial contamination as hydrotherapy in physical therapy (see Chapter 66). Very few studies have been reported that evaluate the potential infectious risks of obstetric hydrotherapy. In one nonrandomized study of 1,385 women with prelabor rupture of the membranes, 538 chose to use a warm tub bath during labor and 847 did not (70). Of those who used the bath, 1.1% developed chorioamnionitis and 0.6% developed endometritis. Of those who did not, 0.2% developed chorioamnionitis and 0.4% developed endometritis, suggesting no infectious risk. However, in a small study of 32 women, one infant developed a Pseudomonas infection that was isolated from the prelabor bath water (71). Presumably, this resulted from a lapse in cleaning technique and indicates a potential infectious risk. Whirlpool baths present even more complex maintenance problems. In a randomized controlled trial of whirlpool baths, 785 patients were studied (72). Benefits in regard to analgesics, instrumentation, and perineal conditions were reported, but no difference was observed in maternal and neonatal infections. These studies provide limited guidance in making an infection control decision regarding maternal hydrotherapy. If a facility decides to use hydrotherapy, a detailed protocol must be developed that includes both a procedure and a cleaning or maintenance protocol. Women with complicated pregnancies should be excluded, and many facilities require that the patient sign a consent form. Cleaning and maintenance protocols depend on the type of equipment used. To avoid these problems, some facilities use inflatable single-use tubs. Prevention Antepartum The goal of good medical care during pregnancy is to ensure that a healthy patient presents for delivery. Conditions that place the pregnant patient at risk for postpartum infection, including urinary tract infection and perhaps bacterial vaginosis, should be identified and treated during routine prenatal care (34, 73). Routine screening for infections, including sexually transmitted and blood-borne diseases such as HIV, HBV, syphilis, chlamydia, and gonorrhea should identify other infections in the pregnant patient. Dietary restrictions are appropriate to avoid infection with Listeria (74). Intrapartum Semmelweis' (6, 7) original observations on the value of good hand washing with an antibacterial agent remain the cornerstone of good obstetric infection control. The number of vaginal examinations should be limited, and internal monitoring with pressure catheters and scalp electrodes should be used only when necessary and should be introduced with aseptic technique. Fetal electrodes should be avoided in women with HSV, HIV, or HBV. Despite the effectiveness of traditional preventive measures, infection control measures will not prevent many intrapartum infections (35). Studies during obstetric procedures have clearly shown the high risk of exposure of the obstetric team to blood and body fluids during deliveries (75, 76, 77, 78). In one study, observers were placed in delivery rooms and directly recorded how often blood or amniotic fluid exposures were occurring (75). In 230 deliveries observed, blood or amniotic fluid exposure occurred in 39% of 202 vaginal deliveries and 50% of 28 cesarean sections. The highest rates of exposure occurred in obstetricians and midwives. Another study comparing different surgical procedures showed that the frequency of blood exposures during cesarean sections was exceeded only by cardiothoracic and trauma surgery (76). Tichenor et al. (77) demonstrated the need for good eye protection. They collected eye shields attached to surgical masks worn during deliveries and counted the visible splashes. This study found that 54% of the eye shields from the primary obstetricians had been splashed, including 30 of 68 shields from vaginal deliveries and 30 of 44 from cesarean sections. Perforation of surgical gloves is also common during deliveries and often goes unrecognized. Serrano et al. (78) collected 754 surgical gloves used by obstetricians during vaginal and cesarean deliveries and postpartum ligations. The gloves were examined for perforations by an air inflation-water submersion technique, and 13% of the gloves had been perforated. They noted that 62% of the perforations were not recognized during the surgical procedure; thus, the obstetricians were unaware of the potential exposure to blood or body fluids. Because all deliveries are associated with the splatter of blood and body fluids and exposures are common, the delivery team P.933 never knows for sure when they will be exposed. Also, they cannot be certain that any specific patient does not have a blood-borne disease at the time of delivery, even if screening was done during prenatal care. Therefore, appropriate protective equipment should always be worn, including gloves, long-sleeved impervious gowns, shoe covers, masks, and eye protection, and the obstetrician should always be aware of the possibility of glove perforation. Prophylactic antibiotics are given for all nonelective cesarean sections and will prevent 50% or more of postpartum endometritis (24). Despite this, as always, good surgical technique is critical to the prevention of surgical infections (79). Postpartum Good perineal, surgical site, and breast care is important in the postpartum period. Mother and newborn should be separated in the case of infections like tuberculosis or chickenpox. Masks should be worn during minor respiratory infections. The patient should be monitored for urinary retention, with urinary catheterization used only as needed. Epidemic Investigations The most common epidemic problem on obstetric units is an increase in postcesarean fevers, which is easy to investigate, because it is usually related to procedures in the surgical suite. One of the first things to review is the use of prophylactic antibiotics. Failure to give antibiotics or failure to give them appropriately can result in an increase in postpartum infections that is easily correctable. In a teaching hospital, another potential cause of increased infection rates is the presence of a resident with poor surgical technique. We appreciated this phenomenon on our own obstetric unit only when the increase in fevers disappeared at the end of the month when a resident rotated off service and reappeared when the same resident came back on service. The classic epidemic problems on obstetric units are outbreaks of streptococcal or staphylococcal infections. Streptococcus pyogenes (Group A ОІ-Hemolytic Streptococcal Infections) Despite good infection control practices, GABHS epidemics can still occur if a member of the obstetric team is a streptococcal carrier. The problem of the carrier is dramatically illustrated by an outbreak reported from a hospital in Washington state in the 1960s (80, 81). Eleven patients, nine obstetric and two gynecologic, developed GABHS infections and one died. Eight of nine obstetric patients developed signs of endometritis within 30 to 60 hours of delivery, and the ninth patient was readmitted on the sixth postpartum day. Although nasopharyngeal cultures were negative from all staff who had contact with the patients, epidemiologic investigation identified the only staff member who had contact with all infected patients. When he stopped practicing, the infections disappeared, and when he returned to practice, the infections reappeared. This pattern was seen on three separate occasions, despite empiric antibiotic treatment with penicillin. Finally, the physician was hospitalized for clinical and microbiologic studies and was found to be an anal carrier of GABHS. It was demonstrated that he disseminated streptococci when he was moving about, including undressing and exercising. Antibiotic treatment of the physician and his family cleared the carrier state and ended the epidemic. Such outbreaks of GABHS infections continue to be a problem, although rare, and must be recognized quickly because of the potential for severe disease (3, 82). The diagnosis of postpartum GABHS infection on an obstetric service should always be an infection control priority and should be investigated immediately (see also Chapter 31). Staphylococcus aureus Outbreaks of staphylococcal infections are also uncommon on the modern obstetric unit. Exceptions occur, which are of concern mainly because of the risk of introduction of methicillin-resistant S. aureus (MRSA) into other hospital units such as neonatal intensive care units. An outbreak of MRSA at a large regional maternity unit in England identified 37 patients who had MRSA and noted that perineal colonization was common in postpartum women but not in staff members (83). This was attributed to perineal injury from delivery that created a favorable site for colonization. The wards in this hospital differed from most American hospitals in that common toilet facilities with baths were provided for each ward rather than private bathrooms. Contamination of baths and bidets with MRSA was documented. Mattress covers were also contaminated and remained contaminated even after cleaning with detergent. Most mattress covers were found to be porous, and the core of some mattresses contained MRSA. The relative contribution of environmental transmission cannot be determined from this study because, as in most MRSA outbreaks, transient carriage by staff members was demonstrated. MRSA was eradicated from the maternity wards with infection control measures, including replacement of all mattresses. Staphylococcal infections do not present the same urgency as GABHS infections, but clustering of infections should be investigated thoroughly both for staphylococcal carriers and for breaks in technique that facilitate cross-infection on the obstetric ward (see also Chapters 28 and 29). CONCLUSIONS Hospital-acquired infections in obstetric patients have a long and dramatic history, but modern obstetric practices have produced low infection rates and extremely low maternal mortality. Good surveillance data on obstetric infection are limited because of difficulties in establishing specific diagnosis and the short hospital stays of most obstetric patients. It is important to appreciate that these infections still occur, produce significant maternal and neonatal morbidity, and require careful monitoring. Also, nonobstetric infections in the obstetric patient and infections in newborns remain a major concern for hospital staff, P.934 requiring good infection control practice and careful patient management. REFERENCES 1. Bridson EY. Iatrogenic epidemics of puerperal fever in the 18th and 19th centuries. Br J Biomed Sci 1996;53:134–139. 2. Yokoe DS, Christiansen CL, Johnson R, et al. Epidemiology of and surveillance for postpartum infections. Emerg Infect Dis 2001;7:837–841. 3. Ledger WJ. Puerperal endometritis. In: Bennet JV, Brachman PS eds. Hospital infections, 4th ed. Philadelphia: Lippincott-Raven Publishers, 1998:551–561. 4. Lowis GW. Epidemiology of puerperal fever: the contributions of Alexander Gordon. Med Hist 1993;37:399–410. 5. Holmes OW. The contagiousness of puerperal fever. N Engl Q J Med Surg 1843;1:503–530. 6. Semmelweis IP. Childbed fever. Rev Infect Dis 1981;3:808–811. 7. Semmelweis I. The etiology of childbed fever. In: The classics of medicine library. Birmingham: Gryphon Editions, 1981. 8. Loudon I. Puerperal fever, the streptococcus, and the sulphonamides, 1911–1945. BMJ 1987;295:485–490. 9. Bartlett JG, Onderdonk AB, Drude E, et al. Quantitative bacteriology of the vaginal flora. J Infect Dis 1977;136:271–277. 10. Bartlett JG, Moon NE, Goldstein PR, et al. Cervical and vaginal microflora: ecologic niches in the female lower genital tract. Am J Obstet Gynecol 1978;130:658–661. 11. Gibbs RS. Microbiology of the female genital tract. Am J Obstet Gynecol 1987;156:491–495. 12. Hammill HA. Normal vaginal flora in relation to vaginitis. Obstet Gynecol Clin North Am 1989;16:329–336. 13. Larsen B, Galask RP. Vaginal microbial flora: practical and theoretical relevance. Obstet Gynecol 1980;55(Suppl 5):100s–113s. 14. Faro S, Cox S, Phillips L, et al. Influence of antibiotic prophylaxis on vaginal microflora. J Obstet Gynecol 1986;6(Suppl 1):S4–S6. 15. Casey BM, Cox SM. Chorioamnionitis and endometritis. Infect Dis Clin North Am 1997;11:203–222. 16. Awadalla SG, Perkins RP, Mercer LJ. Significance of endometrial cultures performed at cesarean section. Obstet Gynecol 1986;68:220–225. 17. Emmons SL, Krohn M, Jackson M, et al. Development of wound infections among women undergoing cesarean section. Obstet Gynecol 1988;72:559–564. 18. Gibbs RS. Infection after cesarean section. Clin Obstet Gynecol 1985;28:697–710. 19. Atterbury JL, Groome LJ, Baker SL, et al. Hospital readmission for postpartum endometritis. J Matern Fetal Med 1998;7:250–254. 20. Faro S, Phillips LE, Martens MG. Perspectives on the bacteriology of postoperative obstetric-gynecologic infections. Am J Obstet Gynecol 1988;158:694–700. 21. Gibbs RS, Blanco JD, Bernstein S. Role of aerobic gram-negative bacilli in endometritis after cesarean section. Rev Infect Dis 1985;7(Suppl 4):S690–S695. 22. Watts DH. Eschenbach DA, Kenny GE. Early postpartum endometritis: the role of bacteria, genital mycoplasma and Chlamydia trachomatis. Obstet Gynecol 1989;73:52–60. 23. Hoyme UB, Kiviat N, Eschenbach DA. Microbiology and treatment of late postpartum endometritis. Obstet Gynecol 1986;68:226–232. 24. Hemsell DL. Prophylactic antibiotics in gynecologic and obstetrical surgery. Rev Infect Dis 1991;13(Suppl 10):S821–S841. 25. Newton ER, Prihoda TJ, Gibbs RS. A clinical and microbiologic analysis of risk factors for puerperal endometritis. Obstet Gynecol 1990;75:402–406. 26. Semprini AE, Castagna C, Ravizza M, et al. The incidence of complications after cesarean section in 156 HIV-positive women. AIDS 1995;9:913–917. 27. Calhoun BC, Brost B. Emergency management of sudden puerperal fever. Obstet Gynecol Clin North Am 1995;22:357–367. 28. Martens MG, Faro S, Hammill HA, et al. Transcervical uterine cultures with a new endometrial suction curette: a comparison of three sampling methods in postpartum endometritis. Obstet Gynecol 1989;74:273–276. 29. Noronha S, Yue CT, Sekosan M. Puerperal group A beta-hemolytic streptococcal toxic shock-like syndrome. Obstet Gynecol 1996;88:728. 30. Larsson PG, Platz-Christensen J, Bergman B, et al. Advantage or disadvantage of episiotomy compared with spontaneous perineal laceration. Gynecol Obstet Invest 1991;31:213–216. 31. Shy KK, Eschenbach DA. Fatal perineal cellulitis from an episiotomy site. Obstet Gynecol 1979;54:292–298. 32. CDC NNIS System. National Nosocomial Infections Surveillance (NNIS) system report, data summary from January 1992–June 2001, issued August 2001. Am J Infect Control 2001;29:404–421. 33. Roberts S, Maccato M, Faro S, et al. The microbiology of post-cesarean wound morbidity. Obstet Gynecol 1993;81:383–386. 34. Millar LK, Cox SM. Urinary tract infections complicating pregnancy. Infect Dis Clin North Am 1997;11:13–26. 35. Soper DE, Mayhall CG, Froggatt JW. Characterization and control of intraamniotic infection in an urban teaching hospital. Am J Obstet Gynecol 1996;175:304–309. 36. Riggs JW, Blanco JD. Pathophysiology, diagnosis, and management of intraamniotic infection. Semin Perinatol 1998;22:251–259. 37. Hussey MJ, Levy ES, Pombar X, et al. Evaluating rapid diagnostic tests of intraamniotic infection: gram stain, amniotic fluid glucose level, and amniotic fluid to serum glucose level ratio. Am J Obstet Gynecol 1998;179:650–656. 38. Holbrook KF, Nottebart VF, Hameed SR, et al. Automated postdischarge surveillance for postpartum and neonatal nosocomial infections. Am J Med 1991;91(Suppl 3B):S125–S130. 39. Marchant DJ. Inflammation of the breast. Obstet Gynecol Clin North Am 2002;29:89–102. 40. Marshall BR, Hepper JK, Zirbel CC. Sporadic puerperal mastitis: an infection that need not interrupt lactation. JAMA 1975;233:1377–1379. 41. Lorber B. State of the art clinical article: listeriosis. J Infect Dis 1997;24:1–11. 42. Campbell AN, Sill PR, Wardle JK. Listeria meningitis acquired by cross infection in a delivery suite. Lancet 1981:2:752–753. 43. Simmons MD, Cockcroft PM, Okubadejo OA. Neonatal listeriosis due to cross-infection in an obstetric theatre. J Infect 1986;13:235–239. 44. McKenna DS, Iams JD. Group B streptococcal infections. Semin Perinatol 1998;22:267–276. 45. Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: revised guidelines from the CDC. MMWR Mortal Morb Wkly Rep 2002;51(RR11):1–22. Available at http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5111a1.htm (accessed). 46. Public Health Service Task Force. Recommendations for use of antiretroviral drugs in pregnant HIV-1-infected women for maternal health and interventions to reduce perinatal HIV-1 transmission in the United States, 2002. Available at http://www.aidsinfo.nih.gov/guidelines/perinatal/perinatal/perinatal.pdf (accessed). 47. ACOG Educational Bulletin. Viral hepatitis in pregnancy. Int J Gynaecol Obstet 1998;63:195–202. 48. Weber DJ, Hoffmann KK, Rutala WA. Management of the healthcare worker with human immunodeficiency virus: lessons from nosocomial transmission of hepatitis B virus. Infect Control Hosp Epidemiol 1991;12:625–630. 49. Young EJ, Chafizadeh E, Oliveira VL, et al. Disseminated herpesvirus infection during pregnancy. Clin Infect Dis 1996;22:51–58. 50. ACOG Practice Bulletin. Management of herpes in pregnancy. Int J Gynaecol Obstet 2000; 68:165–174. 51. Bolyard EA, Tablan OC, Williams WW, et al. Guideline for infection control in healthcare personnel. Infect Control Hosp Epidemiol 1998;19:407–463. 52. Chandra PC, Patel H, Schiavello HJ, et al. Successful pregnancy outcome after complicated varicella pneumonia. Obstet Gynecol 1998;92:680–682. 53. Nathwani D, Maclean A, Conway S, et al. Varicella infections in pregnancy and the newborn. J Infect 1998;36(Suppl 1):59–71. P.935 54. Menkhaus NA, Lanphear B, Linnemann CC. Airborne transmission of varicella-zoster virus in hospitals [Letter]. Lancet 1990;336:1315. 55. Centers for Disease Control and Prevention. Vaccinia (smallpox) vaccine: recommendations of the Advisory Committee on Immunization Practices (ACIP), 2001. MMWR Mortal Morb Wkly Rep 2001;50(RR10):1–26. 56. Centers for Disease Control and Prevention. Control and prevention of rubella: evaluation and management of suspected outbreaks, rubella in pregnant women, and surveillance for congenital rubella syndrome. MMWR Mortal Morb Wkly Rep 2001;50(RR12):1–23. 57. Nosocomial infection rates for interhospital comparison: limitations and possible solutions. Infect Control Hosp Epidemiol 1991;12:609–621. 58. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128–140. 59. Horan TC, Gaynes RP, Martone WJ, et al. CDC definitions of nosocomial surgical site infection, 1992: a modification of CDC definitions of surgical wound infections. Am J Infect Control 1992;20:271–274. 60. Haley RW, Culver DH, White JW, et al. The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. Am J Epidemiol 1985;121:182–205. 61. Evaldson GR, Frederici H, Jullig C, et al. Hospital-associated infections in obstetrics and gynecology: effects of surveillance. Acta Obstet Gynaecol Scand 1992;71:54–58. 62. Klimek JJ, Ajemian ER, Gracewski J, et al. A prospective analysis of hospital-acquired fever in obstetric and gynecologic patients. JAMA 1982;247:3340–3343. 63. Mead PB, Hess SM, Page SD. Prevention and control of nosocomial infection in obstetrics and gynecology. In: Wenzel RP, ed. Prevention and control of nosocomial infections, 3rd ed. Baltimore: Williams & Wilkins, 1997:995–1016. 64. Baker C, Luce J, Chemoweth C, et al. Comparison of case-finding methodologies for endometritis after cesarean section. Am J Infect Control 1995;23:27–33. 65. Couto RC, Pedrosa TMG, Nogueria JM, et al. Post-discharge surveillance and infection rates in obstetric patients. Int J Gynaecol Obstet 1998;61:227–231. 66. Hulton LJ, Olmsted RN, Treston-Aurand J, et al. Effect of postdischarge surveillance on rates of infectious complications after cesarean section. Am J Infect Control 1992;20:198–201. 67. Gravel-Tropper D, Oxley C, Memish Z, et al. Underestimation of surgical site infection rates in obstetrics and gynecology. Am J Infect Control 1995;23:22–26. 68. Inpatient perinatal care services. In: American Academy of Pediatrics and American College of Obstetricians and Gynecologists: guidelines for perinatal care, 5th ed. Washington DC: American College of Obstetricians and Gynecologists, 2002:17–55. 69. McCandlish R, Renfrew M. Immersion in water during labor and birth: the need for evaluation. Birth 1993;20:79–85. 70. Eriksson M, Ladfors L, Mattsson LA, et al. Warm tub bath during labor. A study of 1385 women with prelabor rupture of the membranes after 34 weeks of gestation. Acta Obstet Gynaecol Scand 1996;75:642–644. 71. Hawkins S. Water vs conventional births: infection rates compared. Nursing Times 1995;91:38–40. 72. Rush J, Burlock S, Lambert K, et al. The effects of whirlpool baths in labor: a randomized controlled trial. Birth 1996;23:136–143. 73. ACOG Educational Bulletin. Antimicrobial therapy for obstetrical patients. Int J Gynaecol Obstet 1998;61:299–308. 74. Centers for Disease Control and Prevention. Foodborne listeriosis—United States, 1988–1990. JAMA 1992;267:2446–2448. 75. Panlilio AL, Welch BA, Bell DM, et al. Blood and amniotic fluid contact sustained by obstetric personnel during deliveries. Am J Obstet Gynecol 1992;167:703–708. 76. Popejoy SL, Fry DE. Blood contract and exposure in the operating room. Surg Gynecol Obstet 1991;172:480–483. 77. Tichenor JR, Miller RC, Wolf EJ. Risk of eye splash in obstetric procedures. Am J Perinatol 1994;11:359–361. 78. Serrano CW, Wright JW, Newton ER. Surgical glove perforation in obstetrics. Obstet Gynecol 1991;77:525–528. 79. Lyon JB, Richardson AC. Careful surgical technique can reduce infectious morbidity after cesarean section. Am J Obstet Gynecol 1987;157:557–562. 80. McIntyre DM. An epidemic of Streptococcus pyogenes puerperal and postoperative sepsis with an unusual carrier site the anus. Am J Obstet Gynecol 1968;101:308–314. 81. McKee WM, Di Caprio JM, Roberts CE Jr, et al. Anal carriage as the probable source of a streptococcal epidemic. Lancet 1966;2:1007–1009. 82. Mastro TD, Farley TA, Elliott JA, et al. An outbreak of surgical-wound infections due to group A streptococcus carried on the scalp. N Engl J Med 1990;323:968–972. 83. Moore EP, Williams EW. A maternity hospital outbreak of methicillin-resistant Staphylococcus aureus. J Hosp Infect 1991;19:5–16.
Воскресенье, 06 Ноябрь 2011 | 1318 Просмотры | Печать | PDF
Since the middle of the 20th century, patients with cancer have had an added risk of infection because of the use of immunosuppressive chemotherapy and radiation. In recent years the risk has been enhanced by the use of indwelling central venous catheters (CVCs) (1), more intensive therapy, and the use of bone marrow transplantation. Additionally, patients with neoplastic disease have a more prolonged and repeated contact with the hospital environment, which sometimes involves invasive procedures. Moreover, some patients now admitted with malignancy have an underlying infection with human immunodeficiency virus type 1, imposing a profound effect on the immune system (2). It must be kept in mind that the neoplastic process often has less of an effect on host susceptibility to nosocomial infections than the therapeutic process. One must also appreciate that the main priority is successful anticancer therapy, despite the iatrogenic side effects. Overall, advances in the field of oncology often test the fine balance between more aggressive therapy leading to improved survival and not losing ground because of complications, predominantly infections. Early diagnosis, treatment, containment, and prevention of nosocomial infections are of great importance to the management of neoplastic diseases. EPIDEMIOLOGY Cancer patients present the hospital epidemiologist with several special problems. Foremost is the similarity of nosocomial and community-acquired infections (CAIs) in this population. The etiology of an infection in a febrile neutropenic patient acquired at home may be the same as that acquired in the hospital. Many such infections are derived from microorganisms carried as usual flora of the skin and mucous membranes of the alimentary, respiratory, and genitourinary tracts. Under these circumstances, the incubation period cannot be used to determine the relationship of an acute infection to the time of admission to the hospital. Thus, when the cancer patient manifests signs and symptoms of infection while hospitalized, the causative microorganism (e.g., E. coli, P. aeruginosa, Staphylococcus species, Streptococcus species, Klebsiella species, Candida species, or Aspergillus species) might have been acquired before admission, during a previous hospitalization, or after a current admission to the hospital. The endogenous microbial flora may change after hospitalization. Especially during prolonged hospital courses, microorganisms of the hospital environment may be acquired that will increase the patient's risk for an infectious episode. Despite all this, although immunosuppression and malignancy may lead to some unique types of infections due to opportunistic microorganisms, cancer patients are also susceptible to the same infections encountered in the immunocompetent patient such as respiratory and gastrointestinal viral infections. The ability to diagnose viral infections with increased sensitivity using polymerase chain reaction (PCR)-based techniques has raised the question about differentiating reactivation of latent viral infections versus a new infection. Latent infections acquired early in life may become activated during immunosuppression and hospitalization. These must be differentiated from acute primary infections caused by the same microorganism that could have been acquired after hospitalization. Notable among these are the herpesvirus infections, including herpes simplex lesions and cytomegalovirus disease. Also, recurrent varicella-zoster virus (VZV) infection in the form of disseminated zoster is sometimes difficult to differentiate from primary varicella. Evidence suggests that some cases of Pneumocystis carinii pneumonitis may be acutely acquired infections in the hospital rather than the more usual reactivation of latent infection. Another problem is difficulty in establishing the etiology of an infectious episode in the immunocompromised cancer patient. Since most infections are due to commensal or opportunistic microorganisms of the normal microbial flora, the isolation of a microorganism by culture may not necessarily prove it to be the cause of the illness. For example, in the febrile neutropenic patient, a microorganism such as Corynebacterium species when isolated from a blood culture may be the causative agent or may merely be a skin commensal or contaminant of the culture. Additionally, often the recognition of an infected site is difficult. In the severely neutropenic and anemic patient with cancer, the key signs of infection may be absent because of a lack of inflammatory response. Staphylococcus aureus may be introduced nosocomially at the time of a finger stick for a blood count. P.970 Without neutrophils, no infiltration occurs, so swelling and pain of the affected finger may be absent; furthermore, the anemia does not allow the appearance of erythema, so the sole manifestation of the infected finger stick site may be local pain or fever. Meningeal infection in the neutropenic patient occurs without the typical signs of meningeal inflammation such as a stiff neck, and no neutrophils and hypoglycorrhachia will be found in the spinal fluid. Even with fairly extensive bacterial infection in the lung parenchyma, the neutropenic patient may not be able to muster sufficient inflammatory response to create a density recognizable by chest radiograph. Furthermore, because of the multidisciplinary management, some cancer patients may move through many sites during one hospitalization, such as the operating room, intensive care unit, medical service, physical rehabilitation units, and diagnostic imaging and irradiation departments. Under such circumstances, tracking the source of infection sometimes requires exhaustive epidemiologic investigation. Finally, standard meaningful definitions of hospital-acquired opportunistic infections are lacking and may vary from one institution to another. When reviewing the literature pertaining to surveillance of nosocomial infections in patients with cancer, one has to be cognizant of the criteria used to define such infections and the denominator used to quantify them; the characteristics of patient populations primarily in terms of risk factors, such as duration of neutropenia; the existing infection control policies, including antimicrobial prophylaxis; and the type of resources available to diagnose infections. In two prospective surveillance studies in adult and pediatric hematology-oncology patients from Bonn, Germany, the authors noted an overall nosocomial infection rate of 11 and 10.8 per 1,000 patient-days, respectively, with roughly 75% of the infections occurring in patients who were neutropenic (3, 4). To ensure comparability of surveillance data, these authors recommend that all surveillance studies in the cancer population should include infection rates based on number of patient-days at risk, where “at risk” may be defined as a period of neutropenia. Thus, the hospital epidemiologist must consider these and other nuances of the compromised host with cancer when searching for nosocomial infections. Molecular techniques to characterize microbes by subcellular and genetic components are evolving as powerful tools for hospital epidemiology. Analysis of chromosomal DNA by pulsed-field gel electrophoresis, ribotyping, and random primer PCR methods permits more precise characterization than more conventional phenotyping techniques. PATHOGENESIS OF NOSOCOMIAL INFECTIONS IN PATIENTS WITH NEOPLASTIC DISEASE The primary factor in the pathogenesis of infections in the cancer patient is a defect in host defense. In most patients, the defect is an iatrogenic impairment of the immune system such as suppression of B lymphocytes and antibody production, impairment of T lymphocytes impeding cell-mediated responses, or neutropenia due to intensive chemotherapy or irradiation. However, a breach in the integrity of the skin and mucous membranes is a frequent portal of entry for microbes composing the flora of these sites, especially in the neutropenic patient. A tumor mass may obstruct a vital organ, impair circulation, or invade adjacent tissue, providing a nidus for infection. Because of the extensive use of antibiotics, the normal microbial flora is deranged, and overgrowth of more pathogenic microorganisms occurs. The most frequent single defect predisposing to serious and life-threatening nosocomial infections in cancer patients is neutropenia. Easy access to the absolute neutrophil count plus available data on its predictability for infection make this marker useful in both surveillance and management of cancer patients in hospitals. The type, dose, and duration of use of anticancer drugs help identify patients at high risk for infection. Not all patients with cancer and not all anticancer drugs are associated with immunosuppression. Anticancer drugs associated with bone marrow suppression leading to neutropenia include cyclophosphamide, chlorambucil, ifosfamide, melphalan, thiotepa, busulfan, procarbazine, nitrosourea compounds, platinum complexes, triazenes, methotrexate, cytarabine, fluorouracil, mercaptopurine, thioguanine, hydroxyurea, dactinomycin, vincristine, vinblastine, and etoposide. Prednisone, although not a cause of neutropenia, is a potent inhibitor of both humoral and cell-mediated immune responses, especially T-lymphocyte activity. Cyclosporin specifically affects T lymphocytes by decreasing CD4+ lymphocytes and interleukin-2 synthesis. Irradiation and malnutrition cause decreases in T-lymphocyte function. Some anticancer drugs with little or no effect on the immune response and the neutrophil count are bleomycin, asparaginase, tamoxifen, diethylstilbestrol, testosterone, and megestrol. Indwelling central catheters and other intravascular access devices, prosthetic devices such as central nervous system shunts and artificial extremities, endotracheal intubation, surgery, and exposure to intensive care unit environments are also important in the pathogenesis of many infections in cancer patients. These are not discussed here because they are covered in Chapters Chapters 17, 18, 21, 22, 27, 49, and 67, and because there are no significantly unique features of these infections in the cancer patient. ETIOLOGIES OF INFECTION Nosocomial infections in cancer patients can be caused by a variety of infectious microorganisms, but the most common pathogens are bacterial. In previous reports from oncology centers, bacterial microorganisms were isolated in over 75% of nosocomial infections, fungal pathogens in approximately 3% to 10%, and viruses in only 2% (5, 6). Although in an oncology intensive care unit fungal infections comprised 22% of nosocomial infections (7), in another study of nosocomial infections in neutropenic patients 19% were due to fungi (8). As mentioned earlier when making comparisons between studies and centers, one has to keep in mind the differences in definitions and in patient and institute characteristics. Polymicrobial infections are not uncommon in this patient population. Robinson et al. (6) noted multiple isolates in one third of their infections. Both multiple bacterial isolates and mixed infections can occur. P.971 Table 58.1 lists most bacteria, fungi, viruses, and protozoa associated with infections in patients with malignancies. TABLE 58.1. CAUSES OF INFECTIONS IN THE PATIENT WITH CANCER Bacterial Gram-positive Bacillus species Corynebacterium speciesa Enterococcus fecalisa Listeria monocytogenes Staphylococcus aureusa Staphylococcus, coagulase neg ativea Streptococcus pneumoniaea Streptococcus pyogenesa Streptococcus, viridans groupa Gram-negative Aeromonas species Acinetobacter species Actinobacillus species Alcaligenes species Capnocytophagia species Chromobacterium species Citrobacter species Edwardsiella species Eikenella species Enterobacter speciesa Erwinia species Escherichia colia Flavobacterium species Gardnerella species Hatnia species Haemophilus influenzaea Kingella species Klebsiella speciesa Legionella species Moraxella species Morganella species Neisseria species Proteus speciesa Providencia species Pseudomonas aeruginosaa Pseudomonas, not aeruginosa Salmonella speciesa Serratia marcescens Shigella species Stenotrophomonas maltophilia Yersinia enterocolitica Zymononas species Anaerobic coccl and bacilli Bacteroides species Clostridium species Fusobacterium species Peptococcus species Peptostreptococcus species Propionibacterium species Veillonella species Other Mycobacteria Nocardia asteroides Nonbacterial Viral Adenovirusesa Cytomegalovirusa Epstein-Barr virus Enteroviruses Herpesvirus hominis (simplex)a Influenza viruses Parainfluenza viruses Parvovirus Respiratory syncytial virus Rotavirusa Varicella-zoster virusa Papillomavirus Fungi Pseudallescheria boydii Alternaria species Aspergillus speciesa Candida speciesa Coccidiodes immitis Dreshleria exohilum Fusarium species Geotrichum Histoplasma capsulatum Mucor species Penicillium species Rhizopus species Protozoa Cryptosporidium speciesa Pneumocystis carinii Toxoplasma gondii aMost frequent. Bacterial Infections The most important bacterial nosocomial pathogens are S. aureus, E. coli, P. aeruginosa, and coagulase-negative staphylococci (5, 6) (see Chapters 28, 30, 33, and 34). Together these four microorganisms account for over half of nosocomial bacterial infections in cancer patients. Gram-Negative Microorganisms As a family, Enterobacteriaceae are common nosocomial pathogens in cancer patients. E. coli and Klebsiella pneumoniae predominate (6, 9). These microorganisms, along with Serratia species (10), Enterobacter species (11), and Citrobacter species (12), have been isolated in sporadic infections and in epidemics. They are common causes of bacteremia, pneumonia, and urinary tract infections (UTIs). Frequently, patients are already receiving antibiotic therapy when these infections develop (9, 10, 11, 12). P. aeruginosa is the most notorious pathogen in patients with malignancies. It is associated with nosocomial bacteremia, pneumonia, and urinary tract and wound infection. Although a frequent nosocomial pathogen, it has a special predilection for granulocytopenic hosts. In a review of P. aeruginosa infections in cancer patients in the 1990s, Maschmeyer and Braveny (13) note that the proportion of these infections among cases of gram-negative bacteremia over the past two decades has not generally declined with marked local and regional differences in incidence of infections. Infections with P. aeruginosa account for approximately 10% of all nosocomial infections in cancer patients (5, 6, 14). In the hospital environment, P. aeruginosa is associated with respiratory equipment, sinks, and fresh fruit and vegetables. Colonization often precedes infection (14, 15). The case fatality rate for P. aeruginosa infections is reported to be as high as 65% to 70%, which is significantly higher than the rate for other gram-negative bacterial infections (15, 16). Newer antimicrobial agents with improved anti-Pseudomonas activity may lower the fatality rates (17). A variety of other gram-negative microorganisms have also been linked with nosocomial infections in cancer patients. The Legionella species are fastidious gram-negative bacilli. Approximately 42% of cancer patients with Legionnaire's disease are infected in a hospital setting. Use of steroids and neutropenia appear to have causal roles (18). Stenotrophomonas maltophilia (previously Xanthomonas maltophilia) has been reported as a cause of bacteremia, UTI, pneumonia, and wound infections in cancer patients. It is most often detected in patients who have received antibiotics and respiratory therapy. The microorganism has been isolated from hospital sinks and respirators. The association between the use of respiratory equipment and isolation of S. maltophilia from sputum suggests that the equipment may be a significant reservoir for the microorganism (19). Gram-Positive Microorganisms S. aureus was the most frequent bacterial isolate in two surveys of nosocomial pathogens in cancer patients, accounting for 14% to 18% of isolates (5, 6). Surgical sites were most often involved. Coagulase-negative staphylococcal infections have increased dramatically over the past decade; these microorganisms are the most common microorganisms isolated from bloodstream infections in some centers (20, 21). The rise of these relatively nonpathogenic bacteria has been linked to the use of tunneled CVCs such as the Hickman catheter. О-Hemolytic streptococci are normal inhabitants of the oropharynx that invade through damaged mucous membranes and cause bacteremia and pneumonia in cancer patients. A syndrome of severe shock and adult respiratory distress syndrome can result. There is a potential causal relationship with cytosine arabinoside administration (22, 23). Clusters of Corynebacterium jeikeium bacteremia have been P.972 reported from several cancer centers (24, 25, 26). Risk factors include immunosuppression and use of plastic devices such as intravenous catheters. Some evidence suggests that patient-topatient transmission does not occur (26). The microorganism is resistant to multiple antibiotics, and vancomycin is the suggested therapy. Anaerobes Anaerobes are infrequent nosocomial pathogens in the oncology patient and are isolated in less than 5% of infections. Usually, obvious disruption of normal gastrointestinal barriers are apparent when infections do occur (27). Antibiotic-Resistant Bacteria Widespread use of antibiotics, both prophylactic and empiric, have resulted in nosocomial infections caused by multiply resistant microorganisms. Methicillin-resistant S. aureus, vancomycin-resistant enterococci, and fluoroquinolone-resistant enteric microorganisms have been reported to cause significant problems in an oncology population (28, 29, 30). Prudent use of antibiotics and careful surveillance of this population is necessary to detect and control the spread of these pathogens. Fungal Infections Candida Candida albicans is the most common fungal pathogen in cancer patients (see Chapter 39). However, studies have noted increases in the frequency of other Candida species, including C. tropicalis, C. parapsilosis, and C. krusei(31). Within individual cancer centers, a significant species shift has been noted even within the non–C. albicans group, such as an increase in C. parapsilosis and decrease in C. tropicalis (32). Overall, these differences between institutions to some extent are influenced by institutional antifungal prophylaxis guidelines, use of indwelling catheters, as well as types of malignancies treated. Fungemia, pneumonia, UTI, or disseminated disease with involvement of the abdominal viscera may occur. Infections are usually preceded by colonization of the gastrointestinal tract with the offending microorganism, but common-source outbreaks have also been reported. Risk factors include the use of antibiotics, colonization with the microorganism, neutropenia, and the presence of tunneled CVCs. Aspergillus Although it is clear that the incidence of invasive aspergillosis has been increasing in patients with cancer, especially those with hematologic malignancies and bone marrow transplant recipients (33), controversy exists regarding the definition of nosocomial versus community-acquired infection. This is in part due to factors such as an unknown incubation period and size of “infectious” inoculum as well as lack of a uniform, reliable methodology for environmental sampling in studies that attempt to trace the source of infection (34). The overall case fatality rate of this disease is very high, with the highest being in bone marrow transplant recipients (35). Sites most often involved include the lungs and paranasal sinuses. Inhalation of conidia is requisite to the development of this infection. Direct inoculation of Aspergillus species spores from occlusive materials, such as tape, has also been reported. Other Fungal Microorganisms Historically, although C. albicans accounts for the majority of infections in compromised patients, recent epidemiological trends indicate a shift toward infections by Aspergillus species, non-albicans Candida species, as well as previously uncommon hyaline filamentous fungi (such as Fusarium species, Acremonium species, Pseudallescheria boydi), dematiaceous filamentous fungi (such as Bipolaris species and Alternaria species), and yeast-like pathogens (such as Trichosporon species and Malassezia species) (36). These emerging pathogens are increasingly encountered as causing life-threatening invasive infections that are often refractory to conventional therapies. Increasing use of antifungal prophylaxis may be linked with the emergence of these microorganisms. Viral Infections Overall, viruses account for relatively few nosocomial infections. This number is likely to increase as viral diagnostic technology improves. Known nosocomial pathogens include VZV (37, 38), respiratory syncytial virus (RSV) (39), influenza, and rotavirus (40). Hepatitis B and hepatitis C have also been reported from other countries as nosocomial pathogens in children with cancer (41, 42). CLINICAL MANIFESTATIONS Certain points regarding clinical manifestations of infections in patients with cancer are important to remember. Fever is the most frequent manifestation of nosocomial infection in the cancer patient. When fever occurs, especially in the setting of neutropenia, a diagnostic workup, including a careful history, physical examination, and bacterial and fungal cultures of blood, urine, stool, and any obvious sites of infection such as wounds, should be done before beginning therapy. Bloodstream infections most often present with fever with or without evidence of shock. Catheter-related bacteremias or fungemias may present with chills or rigors after flushing the catheter. If a tunneled CVC is in place, all lumens and ports should be cultured. In addition, if symptoms are present, a chest radiograph, and possibly sinus radiographs, should also be obtained. If no source of infection is identified by the diagnostic workup, the criteria for fever of unknown origin may be met. Although wound or other cutaneous infections may be diagnosed by the presence of erythema, induration, or purulence at a surgical site, intravenous catheter site, or previously uninvolved area, neutropenic cancer patients often may not be able to mount an inflammatory response, and local infection may be heralded only by pain or fever. This is often the case with perirectal celluli P.973 tis, a serious and potentially fatal infection in oncology patients. Rectal pain or pain on defecation should alert clinicians to this possibility. In the absence of neutrophils, extensive bacterial invasion of the meninges may occur without the typical signs of meningeal inflammation, or of the lung without a discernible infiltrate by radiograph, or of the skin without swelling, cellulitis, or formation of pus and abscesses. Alternatively, local or radiographic findings may worsen with the return of neutrophils, often in the presence of improving fever curve and clinical course. Differentiating an infection from side effects of chemotherapy or radiation can sometimes be very difficult, and not uncommonly empiric treatment for an infection is started while awaiting further information. SITES OF NOSOCOMIAL INFECTIONS Bloodstream Infections Overall, bloodstream infections account for approximately 20% of nosocomial infections in cancer patients (5, 6). In two recently published prospective surveillance studies of nosocomial infections in adult and pediatric hematology oncology patients, 43% to 52% of nosocomial infections were bloodstream infections (3, 4). The usual definition is the isolation of any pathogen in one or more blood cultures from a patient with clinical symptoms (5, 6). Even though microorganisms known to colonize the skin, such as coagulase-negative staphylococci and Corynebacterium species (but not C. jeikeium), are the most common contaminants of blood cultures, coagulase-negative staphylococcus, especially in the presence of an indwelling catheter, is also a common “true” pathogen. Currently, gram-positive microorganisms are isolated more often than gram-negative bacteria (8). Before 1960, S. aureus was the most common nosocomial pathogen. After adequate antistaphylococcal drugs were introduced, the gram-negative microorganisms gained prominence. More recently, the trend has been toward increasing numbers of gram-positive isolates. Most authors speculate that the increasing use of tunneled CVCs and the concomitant increase in the number of coagulase-negative staphylococcal infections are responsible for this shift (43). The widespread use of second- and third-generation cephalosporins for the empiric treatment of febrile neutropenia has also been speculated to explain this shift. Because these antibiotics have improved gram-negative coverage at the expense of gram-positive coverage, breakthrough nosocomial bacteremias are likely to be of gram-positive origin (44). Fungemia, most often with Candida species, is also increasing among oncology patients (45). Multiple reports note that patients with hematologic malignancies, such as leukemia and lymphoma, are at increased risk of nosocomial bloodstream infections when compared with patients with solid tumors (43, 46, 47). Mayo and Wenzel (47) noted that leukemics had an infection rate 15 times greater than that of patients with solid tumors. In their study, patients with solid tumors were at no greater risk for bloodstream infections than patients without malignancies. In addition to different rates of infection, the infecting microorganisms are different for patients with hematologic malignancies compared with those with solid tumors. E. coli and S. aureus are common pathogens in both groups. Patients with hematologic malignancies have more infections due to P. aeruginosa and K. pneumoniae, and patients with solid tumors have more infections due to Bacteroides species (46). Coagulase-negative staphylococcal infections are also more common in patients with hematologic malignancies (43). Patients who have undergone splenectomy as treatment for malignancy, such as those with Hodgkin's disease, have an increased risk for Streptococcus pneumoniae bacteremia (46). The portal of entry of nosocomial pathogens can often be identified in this population. The most common sites are the respiratory tract, the gastrointestinal tract, and the skin (43, 48). Surgical procedures, especially in patients with solid tumors, also provide a common portal of entry (46). The risk factors for developing nosocomial bloodstream infections include hematologic malignancies, prolonged hospitalization, and bone marrow transplantation (43). Patients with hematologic malignancies are at increased risk because of the intensive cytotoxic chemotherapy, which often renders them pancytopenic for long periods. Mayo and Wenzel (47) noted that over 75% of nosocomial bloodstream infections in leukemic patients occurred when the absolute neutrophil count was below 100 cells/mL. As the intensity of therapy for solid tumor patients increases, so may the rate of nosocomial bloodstream infections. The prognosis of nosocomial bloodstream infection is related to many factors, including the microorganism causing the sepsis, the source of infection, the absolute neutrophil count, the bone marrow status, and the presence or absence of shock (46). In general, the mortality rate of infections caused by gram-negative microorganisms is greater than that of infections caused by gram-positive microorganisms (43). P. aeruginosa and K. pneumoniae are often associated with high mortality rates. Bloodstream infections that are polymicrobial or that are associated with pulmonary or intraabdominal infections also carry a high mortality rate (46). Respiratory Tract Infections Nosocomial infections of the respiratory tract include pneumonia, sinusitis, pharyngitis, otitis, and rhinitis (see Chapters 22, 23, and 49). These infections account for slightly over 20% of nosocomial infections in cancer patients. Rotstein et al. (5) noted them to be the most common infection at Roswell Park Memorial Institute in 1983 and 1984, accounting for 30% of their reported nosocomial infections. The most frequently isolated microorganisms include S. aureus, P. aeruginosa, E. coli, and K. pneumoniae (5, 6). Fungal microorganisms, including Candida species and Aspergillus species, account for a smaller percentage of infections. Frequently, the infecting microorganism is unknown because an invasive procedure would be required to isolate the pathogen. This is common with upper respiratory tract infections such as otitis and sinusitis. Viral infections are also common undiagnosed pathogens of the respiratory tract. Singer et al. (46) reviewed 24 cases of nosocomial pneumonia in cancer patients. Despite invasive diagnostic procedures in 18, there were unclear etiolo P.974 gies in eight, including seven with chronic organizing pneumonia and one with diffuse interstitial pneumonia. Respiratory tract infections occur most commonly in patients with leukemia/lymphoma and those with solid tumors of the lung and head and neck regions (5, 6). Postoperative pneumonias are more often diagnosed in solid tumor patients, because surgical procedures are more often a part of their diagnosis or treatment (5). Urinary Tract Infections UTIs have been reported to cause between 17% and 28% of nosocomial infections in this population (5, 6) (see Chapter 20). A significant number of UTIs may be asymptomatic, especially in children. Gram-negative microorganisms, namely E. coli, K. pneumoniae, P. aeruginosa, and Proteus mirabilis, predominate. The most frequent gram-positive isolates are enterococci. Fungal microorganisms are unusual urinary tract pathogens (5, 6). Patients with cancer are at increased risk for nosocomial UTI when compared with other hospitalized patients (49). Those with underlying diagnoses of prostate, bladder, bone, joint, liver, ovarian, colorectal, or vulvar cancer are the most commonly afflicted (5). As with non-oncology patients, urinary catheterization and manipulation of catheters are often instrumental in the development of nosocomial UTI (49). Surgical Site Infections Surgical site infections (SSIs) account for approximately 20% of nosocomial infections in cancer patients (5, 6) (see Chapter 21). Identification of infections of surgical sites is usually based on the presence of purulent drainage from the sites. However, special consideration must be given to the neutropenic patient who may manifest infection by serous drainage or erythema and induration alone (6). Patients with solid tumors are most likely to develop SSIs (5, 6). This is due to the extensive surgical procedures involved in their diagnosis and treatment. Patients with carcinoma of the vulva or uterus, soft tissue sarcomas, or malignant melanoma are most susceptible. Other high-risk patients include those with gastrointestinal or head and neck malignancies (5, 6). S. aureus is the most common pathogen isolated from SSIs in the cancer patient (5, 6). Others frequently noted include E. coli, coagulase-negative staphylococci, enterococci, and anaerobes (5, 6). Gastrointestinal Infections Little information is available about nosocomial gastrointestinal infections in cancer patients (see Chapter 24). The diagnosis should be based on the development of clinical symptoms of diarrhea with or without the isolation of a known pathogen. It is often difficult to distinguish diarrhea associated with gastroenteritis from that caused by chemotherapy-related mucositis. Close communication between the infection control professional and clinical personnel may facilitate this distinction. Potential pathogens include Salmonella species, Shigella species, rotavirus, other viral agents, and Clostridium difficile. In general, surgical patients and those receiving antibiotics are at increased risk of developing C. difficile-associated diarrhea (see Chapter 36). These predisposing factors also apply to the oncology patient. Fever of Unknown Origin Although not officially recognized by the Centers for Disease Control and Prevention (CDC) as a reportable entity, fever of unknown origin (FUO) is common in the hospitalized cancer patient (3). Despite intensive diagnostic workups and clinical symptoms suggestive of sepsis, pathogens are often not isolated. At St. Jude Children's Research Hospital, FUO accounts for about one third of nosocomial infections in pediatric oncology patients. Rotstein et al. (5) noted that only about 10% of nosocomial infections were classified as FUO in their adult oncology population. Including FUO as a separate, defined clinical entity, Engelhart et al. (3) describe an overall rate of 8.2 per 1,000 days, with two thirds of these episodes occurring during periods of neutropenia. These authors recommend the inclusion of this entity routinely in studies of surveillance of nosocomial infections in patients with cancer. DIAGNOSIS OF NOSOCOMIAL INFECTIONS IN CANCER PATIENTS The establishment of a specific etiologic diagnosis in a cancer patient is often difficult because of the commensal nature of the causative microbes and limited capability of the host to respond with recognizable clinical features. Even more difficult is differentiation of nosocomial from CAI. Despite these difficulties in most cancer centers, the specific cause of an infectious episode is established more frequently in the cancer patient than in the noncancer patient. This stems from the urgent need for treatment and more aggressive approaches to diagnosis. Cultures for Bacteria, Fungi, and Viruses The most meaningful cultures for bacteria and fungi are those of otherwise sterile body fluids such as blood, spinal fluid, bone marrow, urine, and tissue biopsies. Cultures of specific surface lesions by swab, aspirate, or biopsy require correlation with clinical features, histology, and type of microorganism isolated. Cultures of stool, oropharynx, and normal skin usually provide information only on microbial colonization rather than on the etiology of disease. In certain clinical settings such as a patient with prolonged granulocytopenia with fever unresponsive to antibiotics, the isolation of Aspergillus species from the nares or Candida tropicalis from the stool or the urine may raise the index of suspicion for invasive fungal disease. Various techniques are used to diagnose catheter-related infections including paired quantitative CVC and peripheral venous blood cultures and difference in time to detection of blood cultures simultaneously drawn from these two sources (50). In terms of viral cultures, shell vial spin amplification cultures P.975 give a more rapid turnaround time than traditional viral cultures do for detection of cytomegalovirus (CMV) and the more common respiratory viruses including influenza A and B, parainfluenza 1, 2, and 3, RSV, and adenovirus. Respiratory viruses can be significant nosocomial pathogens, and further molecular techniques are sometimes required to differentiate a nosocomial versus a CAI. Although the significance of herpes simplex and VZV isolates is easily discernible because of the typical lesions and illness associated with the overt infections, CMV isolates may be difficult to assess, because the disease patterns associated with this infection are varied, may be nonspecific, and range from asymptomatic latency to life-threatening disease. As mentioned earlier, differentiating a new onset, nosocomial viral infection from reactivation of a latent infection can sometimes be a difficult and fruitless exercise. For many viruses, viral cultures may eventually be replaced by PCR-based tests. PCR-based assays are currently available or in development for numerous viruses including herpes simplex virus (HSV), CMV, adenovirus, and HHV-6. Although the majority of these tests are done by referral laboratories, centers with a high sample load such as ours are in the process of developing onsite testing facilities. This will facilitate the optimum utilization of this test in terms of turnaround time. Smears and Stains Material obtained from infected sites may contain enough of the causative microorganism to permit recognition of the microbe with selective stains and microscopy. Bacterial stains include Gram stain for most bacteria and acid-fast stain for mycobacteria, Nocardia species, and Cryptosporidium species. Although fungi are usually visualized directly by a Gram stain, a KOH (wet) mount, or a Calcofluor white stain, histopathologically or cytologically, special stains such as the methenamine silver, periodic acid-Schiff, or Papanicolaou stains are used. P. carinii can be visualized by a Grocott-Gomori methenamine silver nitrate, Calcofluor white, toluidine blue O, Giemsa, or monoclonal antibody stain. An India ink preparation is made for Cryptococcus neoformans. Rapid identification of viruses is based on tests that detect viral antigens such as direct fluorescent-antibody assay or enzyme immunoassay. These tests are currently used to detect respiratory viruses such as RSV, influenza, parainfluenza and adenovirus, HSV, VZV, and gastrointestinal pathogens such as rotavirus. Tissue Biopsy Biopsies of various tissues may be obtained for histopathologic examination, for microscopic examination of stained smears, and for culture, including skin biopsy (punch and excisional), lung biopsy (open biopsy and transbronchial), and liver and kidney biopsy (transcutaneous needle biopsy and open biopsy). In experienced hands, transthoracic needle biopsy of chest lesions is a relatively safe and noninvasive tool for the diagnosis of pulmonary fungal disease (51). Radiography and Imaging Radiography is most helpful in recognizing pneumonia. Serial chest radiographs are especially helpful in establishing nosocomially acquired pneumonia. Absence of discernible infiltrates does not exclude significant infection of the pulmonary parenchyma in the neutropenic patient. A more sensitive diagnostic test for early diagnosis of chest disease, especially fungal disease, is computed axial tomography (52, 53). Even this test has limitations in the setting of a profoundly granulocytopenic host. Computed axial tomography of the liver, spleen, and kidneys is useful in identifying systemic fungal infections. The hypodense distinct lesions are highly suggestive of systemic candidiasis and aspergillosis (54, 55). Newer Technologies Over the years, molecular techniques are increasingly being used to diagnose nosocomial infections and to investigate the source and spread of infection (39, 41, 56, 57, 58). Techniques being used to investigate outbreaks or in the context of epidemiologic surveillance include DNA-based methods such as plasmid profiling, restriction endonuclease analysis of plasmid and genomic DNA, Southern hybridization analysis using specific DNA probes, and chromosomal DNA profiling using either pulsed-field gel electrophoresis or PCR-based methods (59). Overall, although most of these tools are still in the development phase and have their own limitations, they can be useful supplements to traditional epidemiologic investigations. Criteria for Nosocomial (Hospital-Acquired) Infections in Cancer Patients Few studies have provided a perspective of nosocomial infections occurring exclusively in cancer patients (3, 4, 5, 6). Due to the epidemiologic nuances mentioned earlier in an immunosuppressed host, the criteria in use for general hospitals such as those used for the CDC's National Nosocomial Infections Surveillance (NNIS) system (60) are not adequate to evaluate nosocomial infections in the immunosuppressed host and require some modifications. Since 1970, an infection surveillance program has been in operation at St. Jude Children's Research Hospital, a 60-bed pediatric oncology center. Criteria were developed to specifically monitor cancer patients. These criteria are outlined here, and results of the surveillance are described to show the pattern of such infections over the past decade, the relationship to several malignancies, and current trends. General Infections that develop after 48 hours of hospitalization are classified as hospital-acquired infections (HAIs) when the interval between admission and onset of symptoms is greater than the incubation period of the disease and when there is no evidence that the infection was present or incubating at admission. Infection that develops in a patient who was febrile continuously from admission to onset of new symptoms is classified as CAI. Infections that develop in patients who are febrile at admission and become afebrile for 72 or more hours before onset of symptoms are classified as HAI. P.976 Infections present on admission and those that become evident during the first 48 hours of hospitalization are classified as CAI unless the infection is directly related to a recent admission. Reactivation of latent infection (e.g., herpes zoster, herpes simplex, P. carinii pneumonitis) is classified as CAI. Only the initial infection is counted in patients with multiple infections unless there is anatomic or temporal separation of the infections to suggest different origins. If a new and different microorganism is cultured from a previously infected site, it is counted as HAI only if deterioration in the patient's condition is evident or antibiotics are changed to provide specific coverage of the new pathogen. Specific Bacteremia: One or more positive blood cultures unrelated to an infection present at admission. Propionibacterium, О-hemolytic streptococci, and Staphylococcus epidermidis are excluded unless there are two or more positive blood cultures and specific antibiotic therapy is instituted. Postmortem blood cultures are excluded. Fungemia: One or more positive blood cultures resulting in antifungal therapy and unrelated to an infection present at admission. Postmortem blood cultures are excluded. Surgical site infections: Only infections resulting from procedures performed in the operating room are counted as SSIs (i.e., an infection at a central line site is classified as an SSI only if the line was inserted in the operating room. If the line was inserted at bedside or in the treatment room, the infection is counted as skin and subcutaneous or phlebitis). Any one or more of the following categories qualifies as an SSI: Nonneutropenic patients: purulent drainage with or without culture documentation or antibiotic therapy. Neutropenic patients: <500 neutrophils/mm3 with Purulent drainage, or Erythema and induration with a positive culture and/or antibiotic therapy. Fever (≥38.5ВC) lasting more than 72 hours postoperatively and resulting in antibiotic therapy when the patient was afebrile 2 or more days before surgery. When a patient is returned to surgery because of complications during the postoperative period and evidence of infection is present, it is counted as HAI if there was no evidence of infection during initial surgery. Skin and subcutaneous: may be considered procedure-related or procedure-unrelated. Procedure-related: infections resulting from an invasive diagnostic or therapeutic procedure not performed in surgery. Any one or more of the following categories qualifies: Purulent drainage; Cellulitis (clinical diagnosis) and antibiotic therapy; Erythema, tenderness, induration, and fever resulting in antibiotic therapy and/or culture of a microorganism not thought to be a contaminant; Semiquantitative culture of the catheter tip and/or insertion site section of the catheter resulting in a colony count of at least 15 bacterial colonies in a patient with fever; Do not include chemical inflammation from drugs known to cause phlebitis or necrosis unless purulent drainage or culture-documented infection is present. Procedure-unrelated: any of the following: Vesicles of viral origin that become secondarily infected and drain purulent material or when cultures of aspirates from these vesicles reveal a common pathogen (i.e., S. aureus). Purulent dermatitis or decubitus ulcers that drain purulent material or develop cellulitis. An ulcer in the perineal area that results in antibiotic therapy in a patient who had no signs of ulceration or irritation of the perineum on admission physical examination. Gastrointestinal: Clinical signs and symptoms and a positive culture for a known pathogen or diagnostic test for rotavirus, C. difficile, or similar microorganism not present on admission. Colonization with a known pathogen (e.g., Salmonella) not present on admission stool cultures. Respiratory Upper: nose, throat, or ear infection with signs and symptoms at the site involved. Findings such as coryza, streptococcal pharyngitis, otitis media, and mastoiditis qualify. Oral thrush, herpes simplex labialis, and chronic gingivitis are excluded. Lower: any of the following: Clinical signs and symptoms of lower respiratory tract infection (cough, pleuritic chest pain, purulent sputum, and fever) with or without culture or radiographic documentation. Radiographic evidence of pneumonia in a patient with fever when atelectasis can be ruled out. There should be no radiographic evidence or signs and symptoms of pneumonia at admission. Documentation of pneumonia at postmortem examination is included when the patient was admitted without radiographic evidence or signs and symptoms of pneumonia. Urinary tract Symptomatic: clinical signs and symptoms (such as fever, dysuria, or frequency) and A urine culture revealing a colony count exceeding 100,000 colonies of a single microorganism per milliliter on a properly collected and handled specimen. S. epidermidis is included in the symptomatic patient, or Culture of a suprapubic aspirate yielding an isolate not thought to be a contaminant. Asymptomatic A urine culture colony count exceeding 100,000 of a single microorganism per milliliter plus a negative urine culture at admission obtained while the patient was not on antibiotics, or P.977 The recovery of a new and different pathogen in pure culture with a colony count exceeding 100,000 when the patient was admitted with a UTI. S. epidermidis is excluded in the asymptomatic patient. FUO: All criteria listed below must be met: Fever of at least 38.5ВC lasting 24 or more hours. Developing 2 or more days after admission. These 2 or more days must be spent without fever or antibiotic therapy. No evidence of infection at any site on admission. No clinical evidence of infection except fever. No noninfectious cause for the fever can be determined (e.g., sickle cell crisis, aggressive chemotherapy in a patient with a large tumor load, rheumatic fever, systemic lupus erythematosus, rheumatoid arthritis). Postoperative cases with less than 2 days of fever and without antibiotic therapy are included unless a specific infection is identified. Nosocomial Infections at a Pediatric Cancer Hospital Over One Decade The aforementioned criteria for identification of nosocomial infections in cancer patients were used to determine the annual infection rates at St. Jude Children's Research Hospital from 1983 through 2001 (Fig. 58.1). During this time, although the mean length of stay in the hospital remained reasonably similar, there seems to be a trend toward a decreasing proportion of nosocomial infections per 1,000 patient-days as well as per number of hospital discharges. It should be pointed out, however, that the intensity of immunosuppressive therapy and the number of cancer patients undergoing bone marrow transplantation tended to increase. It should also be noted that noncontagious patients are kept in private rooms with high-efficiency particulate air (HEPA) filters and unidirectional air flow at portals (positive pressure). No special precautions are given for food, and standard nursing care is provided without protective gowns and masks. Standard universal precautions are practiced unless there are specific indications for transmission-based, airborne, droplet, or contact precautions. Obviously, these data cannot be accurately matched to those of other hospitals unless similar criteria are used for the identification of nosocomial infections, rates are adjusted for severity of illness, and medical practices are comparable. Thus, surveillance data are most valuable to the institution from which they are derived. Seemingly minor institutional practices may be reflected in such data. For example, at St. Jude, cultures of the urine, stool, and throat are routinely taken on admission from all cancer patients. Therefore, the likelihood of identifying true hospital acquisitions of urinary tract and enteric infections is greater than in hospitals wherein routine admission cultures are not done. In some hospitals, this practice would not be prudent. Ideally, each hospital should develop a system that best serves its specific needs for cancer patients. Figure 58.1. Annual nosocomial infection rates at a pediatric oncology center from 1983 to 2001. The nosocomial infection rates by malignancy type, by infection type, and for bone marrow transplant recipients at St. Jude P.978 during 1997 are summarized in Table 58.2. The rates are based on 2,139 patient discharges and are expressed as number of infections per 1,000 discharges. TABLE 58.2. NOSOCOMIAL INFECTION RATES BY MALIGNANCY TYPE BY INFECTION SITE AND FOR BONE MARROW TRANSPLANT RECIPIENTS DURING 1997 (2, 139 PATIENT DISCHARGES) Infection Category Infections per 1,000 Discharges ALL (n = 486) AML (n = 221) ST (n = 1,118) TRNS (n = 146) NHL (n = 74) Other (n = 94) Total (n = 2,139) Bacteremia 1 4 9 2 0 0 16 Surgical site 1 0 4 8 0 0 13 UTI 0 1 3 1 0 0 5 Skin and subcutaneous 2 0 0 0 0 0 2 Respiratory 0 1 9 8 1 0 19 Disseminated fungal 0 1 1 2 0 0 4 Fungemia 2 0 0 2 0 0 4 Gastrointestinal 0 0 0 1 1 0 2 Fever of undetermined etiology 2 4 12 13 1 0 32 Other 3 1 1 1 0 0 6 Total 11 12 39 38 3 0 103 ALL, acute lymphocytic leukemia; AML, acute myelocytic leukemia; ST, solid tumors; TRNS, bone marrow transplant; NHL, non-Hodgkin's lymphoma; UTI, urinary tract infection. It is strikingly obvious that the group at highest risk consists of oncology patients who have undergone bone marrow transplantation. Individuals with acute myelocytic leukemia are at second highest risk for nosocomial infections. The microbial etiology was identified in 52 of 103 nosocomial infections encountered in 1997 (Table 58.3). TABLE 58.3. ETIOLOGY OF NOSOCOMIAL INFECTIONS IN PEDIATRIC ONCOLOGY PATIENTS IN 1997 (ETIOLOGY IDENTIFIED IN 52 NOSOCOMIAL INFECTIONS) Etiology No. of Cases Bacterial causes (43 cases) Bacillus cereus 1 Corynebacterium species 1 Enterococcus 3 Staphylococcus aureus 7 Staphylococcus, coagulase neg. 2 Streptococcus 3 Ochrobactrum anthropi 1 Proteus mirabilis 1 Pseudomonas fluorescens 1 Serratia marcescens 1 Clostridium difficile 1 Escherichia coli 4 Enterobacter species 4 Klebsiella pneumoniae 11 Ureaplasma urealyticum 1 Pseudomonas aeruginosa 2 Fungal causes (10 cases) Aspergillus terreus 1 Candida albicans 5 Candida glabrata 1 Candida lusitaniae 1 Candida rugosa 1 Fungi not otherwise specified 1 Viral causes (3 cases) Adenovirus 3 Parainfluenza virus 1 Nosocomial Infections in Adult Neutropenic Cancer Patients An informative study of nosocomial infections in adults with cancer has been reported by Carlisle et al. (8) from the Albert Einstein Cancer Center in New York. This cancer center at the Montefiore Medical Center, like St. Jude, is a research-oriented 48-bed service. Their study included nosocomial infections in 920 cancer patients admitted over a 3.5-year period who were neutropenic (<1,000/mm3) for 2 or more consecutive days during hospitalization. Definitions were established to categorize infectious episodes. The investigators found a high rate of infection in these neutropenic patients, with 46 nosocomial infections per 1,000 neutropenic days. Table 58.4 shows the type of malignancy in relation to the number of days at risk. The sites of infection are shown in Table 58.5 and the causative microorganisms are listed in Table 58.6. Of the 124 bloodstream infections, coagulase-negative staphylococci were found in 42; E. coli in 19; streptococci in 19; Candida species in 11; S. aureus in ten; Klebsiella species in ten; and enterococci, Pseudomonas species, Enterobacter species, Lactobacillus species, Acinetobacter species, and other bacteria in five or less episodes. TABLE 58.4. NEUTROPENIC ADULTS WITH CANCER: DAYS AT RISK Malignancy Patients (%) Days at Risk (%) Hematologic 628 (68) 7,209 (75) Solid tumor 218 (24) 1,339 (14) Bone marrow transplant Hematologic 26 (3) 456 (5) Solid tumor 48 (5) 578 (6) From Carlisle PS, Gucalp R, Wiernik PH, Nosocomial infections in neutropenic cancer patients. Infect Control Hosp Epidemiol 1993; 14(6): 320–324, with permission. TABLE 58.5. SITE-SPECIFIC RATES OF NOSOCOMIAL INFECTIONS BY NUMBER OF NEUTROPENIC PATIENTS (N = 920) AND NUMBER OF DAYS AT RISK (N = 9,582) Site No. Rate/100 Patients Rate/1,000 Days at Risk Overall 444 48.3 46.3 Blood 124 13.5 12.9 Thrush 61 6.6 6.4 Urinary tract 52 5.7 5.4 Respiratory 51 5.5 5.3 Venous access site 43 4.7 4.5 Gastrointestinal tract 32. 3.4 3.3 Skin 31 3.4 3.2 Other 50 5.4 5.2 From Carlisle PS, Gucalp R, Wiernik PH, Nosocomial infections in neutropenic cancer patients. Infect Control Hosp Epidemiol 1993; 14(6):320–324, with permission. TABLE 58.6. NOSOCOMIAL INFECTIONS IN NEUTROPENIC ADULTS WITH CANCER (N = 392) Microorganisms No. Candida 70 Staphylococci, coagulase-negative 67 Escherichia coli 35 Staphylococcus aureus 26 Enterococci 23 Herpes 22 Clostridia 21 Klebsiella 21 Streptococci 21 Pseudomonas 16 Aspergillus 15 Acinetobacter 10 Enterobacter 9 Corynebacterium 8 Citrobacter 3 Proteus 3 Other gram-positive rods 8 Other gram-negative rods 8 Other 6 From Carlisle PS, Gucalp R, Wiernik PH, Nosocomial infections in neutropenic cancer patients. Infect Control Hosp Epidemiol 1993; 14(6):320–324, with permission. These studies, and earlier studies by Robinson et al. (6), clearly point out the patients within the cancer population at highest risk for nosocomial infections. These are neutropenic patients and patients who have undergone bone marrow transplantation. Other Unique Nosocomial Infections in Cancer Patients Perhaps the major serious infection threat to the cancer patient is that caused by the opportunistic fungi, especially candidiasis and aspergillosis. The secular trends in the epidemiology of nosocomial fungal infections in the United States from 1980 to P.979 1990 have been described by Beck-Sague and Jarvis (61). During this decade, the NNIS system hospitals reported 30,477 nosocomial fungal infections. During this time, the fungal infection rate increased from 2.0 to 3.8 infections per 1,000 patients discharged. The medical specialty with a high infection rate was oncology, within which rates varied from 8.9 to 10.6 infections per 1,000 discharges. C. albicans was the most frequently isolated fungal pathogen (59.7%) followed by other Candida species (18.6%). A study of candidemia in cancer patients from November 1992 to October 1994 found that of 249 episodes of candidemia, non-albicans candidemia accounted for 64% (101/159) of episodes in patients with hematologic malignancies and 30% (27/90) of the episodes in patients with solid tumors (62). Although Aspergillus causes a much lower rate of infection, it is the mycosis that has been most convincingly associated with the hospital environment. Outbreaks of nosocomial aspergillosis have been reported to be due to hospital construction and renovation activities (63, 64, 65, 66). Bone marrow transplant patients are especially susceptible. The source of infection is airborne conidia (spores) of Aspergillus species often associated with contaminated air-handling systems. Recent evidence suggests the hospital water distribution system as an additional indoor source for pathogenic fungi (67). PREVENTION AND CONTROL Prevention and control of nosocomial infections in patients with neoplastic disease is one of the most important contributors to the overall success of treatment in this patient population. Not only do nosocomial infections add to the morbidity, mortality, and overall cost of care, but, not uncommonly, infectious complications necessitate modifications in dose and scheduling of antineoplastic therapies, potentially compromising the successful treatment of the patient's malignancy. The general principles of infection control that are applied to any hospitalized patient remain the same for patients with neoplastic disease. These include, when indicated, specific transmission-based precautions in addition to standard precautions. Certain additional precautions are taken based on the immunosuppressed state of these patients and their susceptibility to infections by opportunistic pathogens. This is especially true in the very high risk host such as patients who have recently undergone bone marrow transplantation or those with prolonged granulocytopenia. Special precautions to prevent such patients from acquiring infection by filamentous fungi, especially Aspergillus species, are extremely important. Certain principles of prevention of nosocomial infections are discussed in the paragraphs that follow, keeping in mind the nuances of a host with a neoplastic disease. Although some preventive measures [e.g., sophisticated air-handling systems, total protected environment (TPE)] are labor intensive, consume a considerable amount of limited healthcare resources, and sometimes lack clear-cut supportive evidence, others (e.g., hand washing, appropriate aseptic technique) are simple, inexpensive, and require very little of the busy healthcare worker's time, and their efficacy is firmly established. Healthcare setups need to individualize their practices based on availability of resources and review of local problems. P.980 Hand Hygiene Ignaz Philipp Semmelweis, in his classic paper on the prevention of childbed fever in 1861 (68), clearly showed that hand washing is very effective in preventing the spread of nosocomial infection when used assiduously. His insistence, in the 19th century, that students and physicians clean their hands with a chlorine solution before seeing each patient in the clinic represents the first evidence that hand washing with plain soap and water in the setting of heavily contaminated hands may not be enough and the use of an antiseptic agent may be more effective. Since then, the importance of hand antisepsis in prevention of nosocomial infections is well accepted (69, 70), and numerous professional societies and committees have published guidelines for appropriate hand hygiene practices. The latest set of guidelines on this subject based on review of existent literature was recently published by the CDC in the Morbidity and Mortality Weekly Report (71). The recommendations include use of an alcohol-based hand rub for routinely decontaminating hands in various clinical care situations. Alternatively, the practice of hand washing with antimicrobial soap and water should be continued. Ready access at strategic locations of efficacious hand-hygiene products with low irritancy potential has been emphasized. Use of artificial fingernails or extenders by clinical care providers, especially those taking care of high-risk patients, is discouraged. It seems clear that the simple act of hand washing greatly reduces the likelihood of transmitting pathogenic microorganisms to hospitalized patients on the hands of healthcare workers. What is also clear is that despite the presence of published guidelines and policies, adherence of healthcare workers to recommended hand-hygiene procedures has been poor, with mean baseline rates of 5% to 81% (71). We recommend that institutions review these guidelines and, based on the resources available, select a hand-hygiene agent or agents and implement a hand-hygiene policy. Periodic monitoring for compliance and focused interventions to improve it, based on the feedback generated, is critical to the success of this intervention (see also Chapter 96). Infections Associated with Intravascular Devices Intravascular devices, particularly tunneled CVCs, are used extensively in patients with neoplastic diseases, and infections associated with them result in increased morbidity in the immunocompromised patient. Meticulous attention should be given to establishing effective infection control practices for the insertion and care of these devices. Guidelines for the prevention of intravascular catheter-related infections recommended by a working group including numerous professional organizations were recently published by the CDC (72). These include the use of antimicrobial or antiseptic-impregnated CVCs in adults, recommendations on selection of catheter insertion sites, catheter care, and surveillance for catheter-related infections. Consideration should be given to establishing special intravenous therapy teams to ensure a high level of aseptic technique during catheter insertion and follow-up care. Policies and procedures for infusion therapy should be comprehensive, and persons who perform manipulations of these devices should be thoroughly trained in appropriate infection control techniques (see also Chapters 17 and 18). Protective Isolation Although appropriate hand hygiene practices and general principles of antisepsis should be followed with any patient with a neoplastic disease, investigators have explored additional measures to protect certain high-risk patient populations, such as patients with prolonged granulocytopenia or with a recent bone marrow transplant, from acquiring nosocomial infections. These have included elaborate measures for protecting the patient from both extrinsic and intrinsic pathogens, including selective decontamination of the digestive tract, HEPA-filtered air, and food and water low in microbial content. The complexity of such studies makes it difficult to evaluate the effectiveness of simple protective isolation (routine use of gloves, gowns, and masks for all patient contact) as an independent variable. Studies suggest that most bacterial infections in patients with granulocytopenia arise from the patient's own flora, and that colonization by the causative microorganism in nearly half of the infections occurs only after admission to the hospital (73). Contaminated hands of healthcare workers are thought to play a major role in the colonization of these patients. Protective isolation, using only gloves and gowns, has been shown to be effective in reducing infection rates in a pediatric intensive care unit, but patients with immunologic dysfunction were excluded from the study (74). Protective isolation alone has been shown in one study to be of no value in protecting patients with severe granulocytopenia (75). In a randomized clinical trial comparing the role of gown and glove isolation and strict hand washing in the reduction of nosocomial infection in children with solid organ transplantation admitted to a pediatric intensive care unit, Slota et al. (76) found that although the rate of nosocomial infections in both intervention groups was significantly reduced compared to the baseline rate, there was a trend toward a higher reduction in the gown and glove group compared to the hand washing group. Although this study demonstrates the role of gown and glove isolation in certain specific clinical settings and indicates the possibility of some additional benefit of this intervention over simple hand washing, the latter intervention is undeniably relatively inexpensive and simple to implement. Until further studies demonstrating the efficacy and cost-effectiveness of protective isolation in routine care of neoplastic patients in various clinical settings are done, this intervention cannot be uniformly recommended. In the meantime its use as a component of standard infection control procedures such as respiratory or enteric isolation in the setting of documented infections should be continued. Total Protected Environment (TPE) Because the causative agents of nosocomial infections in patients with neoplastic diseases include endogenous and a wide variety of exogenous microorganisms, a comprehensive approach to preventing infection and colonization with hospital pathogens has been tried (77, 78). This comprehensive approach, total pro P.981 tected environment (TPE), has included the use of protective isolation, selective decontamination of the digestive tract, rigorous antisepsis of the skin and perirectal area, and HEPA-filtered air supplied to the patient in a laminar or turbulent fashion. TPE also includes provision of food and water low in microbial content, sterilization or high-level disinfection of objects before they are taken into the room, and frequent and thorough cleaning and disinfection of room surfaces. A sterile patient environment cannot be achieved and maintained. Because the patient's endogenous flora and microorganisms in the room and in food and water can only be suppressed, not totally eliminated, a labor-intensive decontamination regimen must be continued throughout the isolation period. TPE is expensive and is beyond the capabilities of many hospitals providing care for patients with neoplastic diseases. Although a reduction in the incidence of infection has been associated with the use of TPE, it must be recognized that since the performance of these studies more than two decades ago, the treatment of infectious complications in neutropenic patients has been improved considerably. Thus, a more relevant question is whether TPE would improve infection-related mortality with the current availability of better treatment options. At this time, a comprehensive approach of TPE is not a standard recommendation for patients with neoplastic disease. Various components of this approach are followed by individual centers, especially those with a bone marrow transplant program. Air-Handling System Although provision of clean air is important to any patient, additional measures to eliminate the risk of exposure to filamentous fungi such as Aspergillus species, are attempted especially in the high-risk host with prolonged granulocytopenia. Reported outbreaks of nosocomial invasive aspergillosis have been caused by concentration of conidia in hospital ventilation systems (79), contaminated fireproofing materials (80), and air contamination from construction (63, 64, 65, 66). One such measure is the use of HEPA filters. The modern HEPA filter, made of superfine spun-glass fibers less that 1 mm in diameter, was developed by the Army Chemical Corps and the Naval Research Laboratory in the years immediately after World War II. The maximum allowable penetration of a HEPA filter at any point in the media, frame, or gasket is 0.03% of the challenge concentration of monodispersed thermally generated dioctyl phthalate (DOP) droplets having a count-median diameter of 0.3 В 0.03 Вm. HEPA-filtered air has been used in many centers as a component of protective isolation to provide ultraclean air to patients during periods of granulocytopenia. In some studies, HEPA-filtered air was delivered to the patient in a unidirectional (formerly called “laminar flow”) fashion, and in other studies “life island” units that enclosed the patient's bed in a plastic canopy were used. Unidirectional airflow is not achieved in life island units. A concentration of about 2.12 microorganisms per cubic meter of air can be achieved in life island units and 0.21 microorganisms per cubic meter can be achieved in rooms with unidirectional flow, compared with approximately 106 microorganisms per cubic meter in conventional rooms. Although the efficacy of HEPA filters in preventing aspergillosis seems clear (64, 81, 82), the effect on preventing other infections is less certain. Optimum filtration efficiency using HEPA filters requires considerable effort and resources. HEPA filters should be purchased from a reputable company that scan tests each filter before shipment to ensure that the entire surface of the filter, gasket, and media-frame bonding is free of leaks at or above 0.03% of the challenge concentration of hot DOP. After installation, HEPA filters should be retested by skilled technicians, and leaks at or above 0.01% of the challenge concentration of cold DOP should be repaired. Portable HEPA filtration units have been associated with a decrease in incidence density of aspergillosis cases during a construction-associated outbreak when used in conjunction with other infection control measures (65). Because new air-handling units sometimes fail to meet design specifications, they should be DOP leak tested before they are placed in operation. We installed a new air handler with HEPA filters as part of a renovation project. DOP leak testing after construction of the unit revealed leaks of greater than 10% between the filter housing and the air handler walls. This represented a 3-log greater penetration of DOP than specified in design criteria and would not have been detected without proper testing. After repairs, no leaks at or above 0.01% were detected. With proper installation, testing, and maintenance, ultraclean air can be maintained in the patient room. Patients with neoplastic diseases, however, must periodically leave this protected environment for a wide range of diagnostic and therapeutic procedures. Despite the lack of clear-cut data, for a subset of immunocompromised patients identified at high risk for aspergillosis, it is desirable to protect the respiratory tract from opportunistic pathogens during these periods. For this purpose, various masks or respirators may be used. High-efficiency masks have been successfully used in high-risk patients during periods when they are outside their hospital rooms (83). A breakthrough in the manufacture of HEPA-filtered masks is the replacement of delicate fiberglass fibers with durable plastic fibers. This new technology has permitted the production of a durable, lightweight, comfortable mask that readily passes DOP leak tests and should provide the patient with air quality at least as good as that found in HEPA-filtered unidirectional rooms. Another mask in clinical use is the N95 respirator. In the guidelines for Preventing Opportunistic Infections Among Hematopoietic Stem Cell Transplant Recipients published by the CDC, use of the N95 mask (particulate respirator) has been mentioned to prevent mold exposure during transportation near hospital construction or renovation areas since these respirators are regarded as effective against any aerosol (84). For maximal efficacy of any face mask, whether it be the HEPA mask or the N95, proper fit testing and training of the patient is very important. In this regard, unavailability of small masks poses a limitation for their use in infants and small children. In conclusion, although HEPA filters are important, especially to prevent nosocomial aspergillosis in a high-risk host, the correct and optimal installation and use of HEPA filters is relatively complex and expensive. Other factors related to hospital air handling including appropriate air exchanges and pressurization are equally important in preventing nosocomial airborne infections, especially aspergillosis (85). In a recently published study assessing the ability of P.982 hospital air handling systems to filter Aspergillus, other fungi, and particles following the implosion of an adjacent building, an encouraging observation was that even standard hospital air handling systems with filtration exceeding minimum American Society of Heating, Refrigeration, and Air Conditioning Engineers (ASHRAE) standards have a significant safety buffer in filtering Aspergillus spores (86). HEPA filters likely provide an additional level of safety. Design and maintenance of hospital ventilation systems is discussed in detail in Chapter 89. In a recent publication, Anaissie et al. (67) submitted evidence to support the theory that hospital water distribution systems may be a potential indoor reservoir of Aspergillus species and other molds leading to aerosolization of fungal spores and potential patient exposure. In a high-risk population such as bone marrow transplant patients, these authors recommend the use of sterile (boiled) water for drinking and sterile sponges for bathing. Additionally they recommend cleaning of the floors of the patient shower facilities in order to reduce the air concentration of Aspergillus species and other pathogenic airborne molds (87). Antimicrobial Drugs The administration of antimicrobial prophylaxis in patients with cancer to prevent both community-acquired and hospital-acquired infections has been most widely studied in neutropenic patients. Initial trials of infection prophylaxis using combinations of nonabsorbable drugs such as aminoglycosides, polymyxins, and vancomycin were followed by studies of orally absorbable agents, primarily trimethoprim-sulfamethoxazole (TMP-SMX) and quinolones. A review of studies of prophylaxis with TMP-SMX by the Infectious Diseases Society of America (IDSA) Fever and Neutropenia Panel found that in most studies there was some benefit in terms of reduced infection rates in the TMP-SMX treated group compared to the placebo group (88). Studies have also shown some benefit of quinolone-based prophylaxis in reducing the rates of infection in neutropenic patients (89, 90). Finally with the increase in frequency of fungal infections in patients with neoplastic diseases, especially patients with hematologic malignancy, there has been considerable interest in the role of antifungal prophylaxis. A meta-analysis of studies of the efficacy of antifungal prophylaxis in neutropenic patients observed reductions in use of empirical antifungal therapy, superficial fungal infection, invasive fungal infection, and fungal-infection related mortality (91). Although the benefits of antifungal prophylaxis in terms of infections with Candida have been noted, there is no definitive data showing its impact on infections with filamentous fungi. Despite the data supporting the efficacy of prophylaxis with TMP-SMX, quinolones, fluconazole, and itraconazole in reducing the number of infectious episodes during the neutropenic period, the IDSA Fever and Neutropenia Guidelines Panel remarks “concern about the problem of emerging drug-resistant bacteria and fungi due to extensive antibiotic use, plus the fact that such prophylaxis has not been shown to consistently reduce mortality rates, leads to the recommendation that routine prophylaxis with these drugs in neutropenic patients should be avoided, with the exception of use of TMP-SMX for patients at risk for P. carinii pneumonitis” (92). IMPENDING CHANGES IN NOSOCOMIAL INFECTIONS IN CANCER PATIENTS Many developments are affecting, or will soon affect, the cancer patient and the pattern of nosocomial infections. Notable are the trends to treat certain patients with fever and neutropenia as outpatients rather than in the hospital and the emergence of antibiotic-resistant bacterial infections. Infection control personnel will need to expand their efforts to match the expansion of the healthcare delivery system and the overall trend to minimize in-hospital treatment of medical conditions. New hospital construction that provides improved physical facilities with better air-handling systems and new approaches to cancer treatment, such as improved bone marrow transplantation and gene therapy, are factors effecting change. In summary, patients with neoplastic diseases, for numerous reasons, are vulnerable to acquiring nosocomial infections. The challenges of diagnosing and managing nosocomial infections in a patient with cancer include difficulties in early diagnosis due to atypical presentation, difficulty in establishing pathogenicity versus mere colonization, problems in sorting out a community versus a hospital source, as well as differentiating a new infection from reactivation of a previous infection. Hospital epidemiologists should continue to reassess the incidence, type, and sources of nosocomial infections for possible changing trends. REFERENCES 1. Schwarz RE, Groeger JS, Coit DG. Subcutaneously implanted central venous access devices in cancer patients: a prospective analysis. Cancer 1997;79(8):1635–1640. 2. Johnson CC, Wilcosky T, Kvale P, et al. Cancer incidence among an HIV-infected cohort. Pulmonary Complications of HIV Infection Study Group. Am J Epidemiol 1997;146(6):470–475. 3. Engelhart S, Glasmacher A, Exner M, et al. Surveillance for nosocomial infections and fever of unknown origin among adult hematology-oncology patients. Infect Control Hosp Epidemiol 2002;23(5):244–248. 4. Simon A, Fleischhack G, Hasan C, et al. Surveillance for nosocomial and central line-related infections among pediatric hematology-oncology patients. Infect Control Hosp Epidemiol 2000;21(9):592–596. 5. Rotstein C, Cummings KM, Nicolaou AL, et al. Nosocomial infection rates at an oncology center. Infect Control 1988;9(1):13–19. 6. Robinson GV, Tegtmeier BR, Zaia JA. Brief report: nosocomial infection rates in a cancer treatment center. Infect Control 1984;5(6):289–294. 7. Velasco E, Thuler LC, Martins CA, et al. Nosocomial infections in an oncology intensive care unit. Am J Infect Control 1997;25(6):458–462. 8. Carlisle PS, Gucalp R, Wiernik PH. Nosocomial infections in neutropenic cancer patients. Infect Control Hosp Epidemiol 1993;14(6):320–324. 9. Bodey GP, Elting LS, Rodriquez S, et al. Klebsiella bacteremia. A 10-year review in a cancer institution. Cancer 1989;64(11):2368–2376. 10. Saito H, Elting L, Bodey GP, et al. Serratia bacteremia: review of 118 cases. Rev Infect Dis 1989;11(6):912–920. 11. Johnson MP, Ramphal R. Beta-lactam-resistant Enterobacter bacteremia in febrile neutropenic patients receiving monotherapy. J Infect Dis 1990;162(4):981–983. P.983 12. Samonis G, Anaissie E, Elting L, et al. Review of Citrobacter bacteremia in cancer patients over a sixteen-year period. Eur J Clin Microbiol Infect Dis 1991;10(6):479–485. 13. Maschmeyer G, Braveny I. Review of the incidence and prognosis of Pseudomonas aeruginosa infections in cancer patients in the 1990s. Eur J Clin Microbiol Infect Dis 2000;19(12):915–925. 14. Cross A, Allen JR, Burke J, et al. Nosocomial infections due to Pseudomonas aeruginosa: review of recent trends. Rev Infect Dis 1983;5(suppl 5):S837–S845. 15. Schimpff SC, Moody M, Young VM. Relationship of colonization with Pseudomonas aeruginosa to development of Pseudomonas bacteremia in cancer patients. Antimicrob Agents Chemother 1970;10:240–244. 16. Jackson MA, Wong KY, Lampkin B. Pseudomonas aeruginosa septicemia in childhood cancer patients. Pediatr Infect Dis 1982;1(4):239–241. 17. Fergie JE, Shema SJ, Lott L, et al. Pseudomonas aeruginosa bacteremia in immunocompromised children: analysis of factors associated with a poor outcome. Clin Infect Dis 1994;18(3):390–394. 18. Nunnink JC, Gallagher JG, Yates JW. Legionnaires' disease in patients with cancer. Med Pediatr Oncol 1986;14(2):81–85. 19. Khardori N, Elting L, Wong E, et al. Nosocomial infections due to Xanthomonas maltophilia (Pseudomonas maltophilia) in patients with cancer. Rev Infect Dis 1990;12(6):997–1003. 20. Wade JC, Schimpff SC, Newman KA, et al. Staphylococcus epidermidis: an increasing cause of infection in patients with granulocytopenia. Ann Intern Med 1982;97(4):503–508. 21. Patrick CC. Coagulase-negative staphylococci: pathogens with increasing clinical significance. J Pediatr 1990;116(4):497–507. 22. Cohen J, Donnelly JP, Worsley AM, et al. Septicaemia caused by viridans streptococci in neutropenic patients with leukaemia. Lancet 1983;2(8365–8366):1452–1454. 23. Sotiropoulos SV, Jackson MA, Woods GM, et al. Alpha-streptococcal septicemia in leukemic children treated with continuous or large dosage intermittent cytosine arabinoside. Pediatr Infect Dis J 1989;8(11):755–758. 24. Telander B, Lerner R, Palmblad J, et al. Corynebacterium group JK in a hematological ward: infections, colonization and environmental contamination. Scand J Infect Dis 1988;20(1):55–61. 25. Riebel W, Frantz N, Adelstein D, et al. Corynebacterium JK: a cause of nosocomial device-related infection. Rev Infect Dis 1986;8(1):42–49. 26. Khabbaz RF, Kaper JB, Moody MR, et al. Molecular epidemiology of group JK Corynebacterium on a cancer ward: lack of evidence for patient-to-patient transmission. J Infect Dis 1986;154(1):95–99. 27. Brown EA, Talbot GH, Provencher M, et al. Anaerobic bacteremia in patients with acute leukemia. Infect Control Hosp Epidemiol 1989;10(2):65–69. 28. Kern WV, Andriof E, Oethinger M, et al. Emergence of fluoroquinolone-resistant Escherichia coli at a cancer center. Antimicrob Agents Chemother 1994;38(4):681–687. 29. Nourse C, Murphy H, Byrne C, et al. Control of a nosocomial outbreak of vancomycin resistant Enterococcus faecium in a paediatric oncology unit: risk factors for colonisation. Eur J Pediatr 1998;157(1):20–27. 30. Rubin LG, Tucci V, Cercenado E, et al. Vancomycin-resistant Enterococcus faecium in hospitalized children. Infect Control Hosp Epidemiol 1992;13(12):700–705. 31. Marr KA. The changing spectrum of candidemia in oncology patients: therapeutic implications. Curr Opin Infect Dis 2000;13(6):615–620. 32. Safdar A, Perlin DS, Armstrong D. Hematogenous infections due to Candida parapsilosis: changing trends in fungemic patients at a comprehensive cancer center during the last four decades. Diagn Microbiol Infect Dis 2002;44(1):11–16. 33. Marr KA, Carter RA, Crippa F, et al. Epidemiology and outcome of mould infections in hematopoietic stem cell transplant recipients. Clin Infect Dis 2002;34(7):909–917. 34. Kontoyiannis DP, Bodey GP. Invasive aspergillosis in 2002: an update. Eur J Clin Microbiol Infect Dis 2002;21(3):161–172. 35. Lin SJ, Schranz J, Teutsch SM. Aspergillosis case-fatality rate: systematic review of the literature. Clin Infect Dis 2001;32(3):358–366. 36. Groll AH, Walsh TJ. Uncommon opportunistic fungi: new nosocomial threats. Clin Microbiol Infect 2001;7(suppl 2):8–24. 37. Meyers JD, MacQuarrie MB, Merigan TC, et al. Nosocomial varicella. Part I: outbreak in oncology patients at a children's hospital. West J Med 1979;130(3):196–199. 38. Gustafson TL, Lavely GB, Brawner ER Jr, et al. An outbreak of airborne nosocomial varicella. Pediatrics 1982;70(4):550–556. 39. Mazzulli T, Peret TC, McGeer A, et al. Molecular characterization of a nosocomial outbreak of human respiratory syncytial virus on an adult leukemia/lymphoma ward. J Infect Dis 1999;180(5):1686–1689. 40. Rogers M, Weinstock DM, Eagan J, et al. Rotavirus outbreak on a pediatric oncology floor: possible association with toys. Am J Infect Control 2000;28(5):378–380. 41. Widell A, Christensson B, Wiebe T, et al. Epidemiologic and molecular investigation of outbreaks of hepatitis C virus infection on a pediatric oncology service. Ann Intern Med 1999;130(2):130–134. 42. Styczynski J, Wysocki M, Koltan S, et al. Epidemiologic aspects and preventive strategy of hepatitis B and C viral infections in children with cancer. Pediatr Infect Dis J 2001;20(11):1042–1049. 43. Morrison VA, Peterson BA, Bloomfield CD. Nosocomial septicemia in the cancer patient: the influence of central venous access devices, neutropenia, and type of malignancy. Med Pediatr Oncol 1990;18(3):209–216. 44. Funada H, Machi T, Matsuda T. Bacteremia complicating acute leukemia with special reference to its incidence and changing etiological patterns. Jpn J Clin Oncol 1988;18(3):239–248. 45. Wright WL, Wenzel RP. Nosocomial Candida. Epidemiology, transmission, and prevention. Infect Dis Clin North Am 1997;11(2):411–425. 46. Singer C, Kaplan MH, Armstrong D. Bacteremia and fungemia complicating neoplastic disease. A study of 364 cases. Am J Med 1977;62(5):731–742. 47. Mayo JW, Wenzel RP. Rates of hospital-acquired bloodstream infections in patients with specific malignancy. Cancer 1982;50(1):187–190. 48. Kilton LJ, Fossieck BE Jr, Cohen MH, et al. Bacteremia due to gram-positive cocci in patients with neoplastic disease. Am J Med 1979;66(4):596–602. 49. Giongo F, Meroni MR, Busin S. Nosocomial urinary tract infections in neoplastic and non-neoplastic diseases. Chemioterapia 1987;6(2 suppl):503–504. 50. Mermel LA, Farr BM, Sherertz RJ, et al. Guidelines for the management of intravascular catheter-related infections. Infect Control Hosp Epidemiol 2001;22(4):222–242. 51. Hoffer FA, Gow K, Flynn PM, et al. Accuracy of percutaneous lung biopsy for invasive pulmonary aspergillosis. Pediatr Radiol 2001;31:144–152. 52. Barnes AJ, Oppenheim BA, Chang J, et al. Early investigation and initiation of therapy for invasive pulmonary aspergillosis in leukaemic and bone marrow transplant patients. Mycoses 1999;42(5–6):403–408. 53. Caillot D, Casasnovas O, Bernard A, et al. Improved management of invasive pulmonary aspergillosis in neutropenic patients using early thoracic computed tomographic scan and surgery. J Clin Oncol 1997;15(1):139–147. 54. Bartley DL, Hughes WT, Parvey LS, et al. Computed tomography of hepatic and splenic fungal abscesses in leukemic children. Pediatr Infect Dis 1982;1(5):317–321. 55. Flynn PM, Shenep JL, Crawford R, et al. Use of abdominal computed tomography for identifying disseminated fungal infection in pediatric cancer patients. Clin Infect Dis 1995;20(4):964–970. 56. Pfaller MA, Herwaldt LA. The clinical microbiology laboratory and infection control: emerging pathogens, antimicrobial resistance, and new technology. Clin Infect Dis 1997;25(4):858–870. 57. Peterson LR, Noskin GA. New technology for detecting multidrug-resistant pathogens in the clinical microbiology laboratory. Emerg Infect Dis 2001;7(2):306–311. 58. Liu Y, Davin-Regli A, Bosi C, et al. Epidemiological investigation of Pseudomonas aeruginosa nosocomial bacteraemia isolates by PCR-based DNA fingerprinting analysis. J Med Microbiol 1996;45(5):359–365. 59. Pfaller MA. Molecular epidemiology in the care of patients. Arch Pathol Lab Med 1999;123(11):1007–1010. P.984 60. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16(3):128–140. 61. Beck-Sague C, Jarvis WR. Secular trends in the epidemiology of nosocomial fungal infections in the United States, 1980–1990. National Nosocomial Infections Surveillance System. J Infect Dis 1993;167(5):1247–1251. 62. Viscoli C, Girmenia C, Marinus A, et al. Candidemia in cancer patients: a prospective, multicenter surveillance study by the Invasive Fungal Infection Group (IFIG) of the European Organization for Research and Treatment of Cancer (EORTC). Clin Infect Dis 1999;28(5):1071–1079. 63. Opal SM, Asp AA, Cannady PB Jr, et al. Efficacy of infection control measures during a nosocomial outbreak of disseminated aspergillosis associated with hospital construction. J Infect Dis 1986;153(3):634–637. 64. Cornet M, Levy V, Fleury L, et al. Efficacy of prevention by high-efficiency particulate air filtration or laminar airflow against Aspergillus airborne contamination during hospital renovation. Infect Control Hosp Epidemiol 1999;20(7):508–513. 65. Loo VG, Bertrand C, Dixon C, et al. Control of construction-associated nosocomial aspergillosis in an antiquated hematology unit. Infect Control Hosp Epidemiol 1996;17(6):360–364. 66. Oren I, Haddad N, Finkelstein R, et al. Invasive pulmonary aspergillosis in neutropenic patients during hospital construction: before and after chemoprophylaxis and institution of HEPA filters. Am J Hematol 2001;66(4):257–262. 67. Anaissie EJ, Stratton SL, Dignani C, et al. Pathogenic moulds (including Aspergillus spp.) in hospital water distribution systems: A three-year prospective study and clinical implications for patients with hematological malignancies. Blood 2003;101(7):2542–2546. 68. Classics in infectious diseases. Childbed fever by Ignaz Philipp Semmelweis. Rev Infect Dis 1981;3(4):808–811. 69. Larson E. A causal link between handwashing and risk of infection? Examination of the evidence. Infect Control 1988;9(1):28–36. 70. Larson E. Skin hygiene and infection prevention: more of the same or different approaches? Clin Infect Dis 1999;29(5):1287–1294. 71. Boyce JM, Pittet D. Guideline for hand hygiene in health-care settings: recommendations of the Healthcare Infection Control Practices Advisory Committee and the HICPAS/SHEA/APIC/IDSA Hand Hygiene Task Force. MMWR Recomm Rep 2002;51(RR-16):1–45. 72. O'Grady NP, Alexander M, Dellinger EP, et al. Guidelines for the prevention of intravascular catheter-related infections. Centers for Disease Control and Prevention. MMWR Recomm Rep 2002;51(RR-10):1–29. 73. Schimpff SC, Young VM, Greene WH, et al. Origin of infection in acute nonlymphocytic leukemia. Significance of hospital acquisition of potential pathogens. Ann Intern Med 1972;77(5):707–714. 74. Klein BS, Perloff WH, Maki DG. Reduction of nosocomial infection during pediatric intensive care by protective isolation. N Engl J Med 1989;320(26):1714–1721. 75. Nauseef WM, Maki DG. A study of the value of simple protective isolation in patients with granulocytopenia. N Engl J Med 1981;304(8):448–453. 76. Slota M, Green M, Farley A, et al. The role of gown and glove isolation and strict handwashing in the reduction of nosocomial infection in children with solid organ transplantation. Crit Care Med 2001;29(2):405–412. 77. Levine AS, Siegel SE, Schreiber AD, et al. Protected environments and prophylactic antibiotics. A prospective controlled study of their utility in the therapy of acute leukemia. N Engl J Med 1973;288(10):477–483. 78. Pizzo PA. The value of protective isolation in preventing nosocomial infections in high risk patients. Am J Med 1981;70(3):631–637. 79. Lentino JR, Rosenkranz MA, Michaels JA, et al. Nosocomial aspergillosis: a retrospective review of airborne disease secondary to road construction and contaminated air conditioners. Am J Epidemiol 1982;116(3):430–437. 80. Aisner J, Schimpff SC, Bennett JE, et al. Aspergillus infections in cancer patients. Association with fireproofing materials in a new hospital. JAMA 1976;235(4):411–412. 81. Sherertz RJ, Belani A, Kramer BS, et al. Impact of air filtration on nosocomial Aspergillus infections. Unique risk of bone marrow transplant recipients. Am J Med 1987;83(4):709–718. 82. Hahn T, Cummings KM, Michalek AM, et al. Efficacy of high-efficiency particulate air filtration in preventing aspergillosis in immunocompromised patients with hematologic malignancies. Infect Control Hosp Epidemiol 2002;23(9):525–531. 83. Raad I, Hanna H, Osting C, et al. Masking of neutropenic patients on transport from hospital rooms is associated with a decrease in nosocomial aspergillosis during construction. Infect Control Hosp Epidemiol 2002;23(1):41–43. 84. Guidelines for preventing opportunistic infections among hematopoietic stem cell transplant recipients. MMWR Recomm Rep 2000;49(RR-10):1–7. 85. Streifel AJ. In with the good air. Infect Control Hosp Epidemiol 2002;23(9):488–490. 86. Srinivasan A, Beck C, Buckley T, et al. The ability of hospital ventilation systems to filter Aspergillus and other fungi following a building implosion. Infect Control Hosp Epidemiol 2002;23(9):520–524. 87. Anaissie EJ, Stratton SL, Dignani MC, et al. Cleaning patient shower facilities: a novel approach to reducing patient exposure to aerosolized Aspergillus species and other opportunistic molds. Clin Infect Dis 2002;35(8):E86–E88. 88. Hughes WT, Armstrong D, Bodey GP, et al. 1997 guidelines for the use of antimicrobial agents in neutropenic patients with unexplained fever. Infectious Diseases Society of America. Clin Infect Dis 1997;25(3):551–573. 89. Engels EA, Lau J, Barza M. Efficacy of quinolone prophylaxis in neutropenic cancer patients: a meta-analysis. J Clin Oncol 1998;16(3):1179–1187. 90. Cruciani M, Rampazzo R, Malena M, et al. Prophylaxis with fluoroquinolones for bacterial infections in neutropenic patients: a meta-analysis. Clin Infect Dis 1996;23(4):795–805. 91. Bow EJ, Laverdiere M, Lussier N, et al. Antifungal prophylaxis for severely neutropenic chemotherapy recipients: a meta analysis of randomized-controlled clinical trials. Cancer 2002;94(12):3230–3246. 92. Hughes WT, Armstrong D, Bodey GP, et al. 2002 guidelines for the use of antimicrobial agents in neutropenic patients with cancer. Clin Infect Dis 2002;34(6):730–751.
Воскресенье, 06 Ноябрь 2011 | 1184 Просмотры | Печать | PDF
BACKGROUND Patients with spinal cord injury constitute a unique population with distinct multidisciplinary problems. According to a consensus statement by the National Institute on Disability and Rehabilitation Research, at least 8,000 new cases of spinal cord injury accrue each year in the United States and about 200,000 Americans suffer from the consequences of such an injury (1). The number of patients living with spinal cord injury is expected to continue to rise owing to the increase in their life expectancy. Most cases of spinal cord injury are traumatic, most notably due to motor vehicle accidents, gunshot wounds, and falls. Nosocomial infections are a major cause of morbidity and mortality in this population of patients (2, 3). The three most common infections in patients with spinal cord injury are urinary tract infections, infections of pressure sores and underlying bone, and respiratory tract infections. These three infections have certain characteristics when they occur in patients with spinal cord injury as compared with the able-bodied population. This chapter discusses the factors that predispose to nosocomial infections in relation to time of injury; delineates the interrelated pathogenesis and microbiology, unusual clinical manifestations, problematic diagnosis, and difficult prevention of infections involving the urinary tract, pressure sores and underlying bone, and respiratory tract; and addresses the issue of colonization and infection by multiresistant microorganisms. FACTORS THAT PREDISPOSE TO NOSOCOMIAL INFECTIONS IN RELATION TO TIME OF INJURY Patients with spinal cord injury are predisposed to nosocomial infections both in the acute and chronic settings after injury (Table 56.1). Immediately after the injury, patients are admitted to the hospital for management of injuries to the spinal cord and possibly other organs, and usually remain hospitalized for about 2 to 3 months for initial rehabilitation. Patients with spinal cord injury frequently have wounds of the chest, abdomen, and neck that may require surgical intervention, thereby imposing additional risks for postoperative infections. Critically ill patients have a particularly high risk of developing infections with resistant microorganisms, including multiresistant gram-negative bacilli, methicillin-resistant Staphylococcus aureus (MRSA), and vancomycin-resistant Enterococcus (VRE). Large doses of glucocorticosteroids, given immediately after traumatic injury, may also predispose these patients to infection. The majority of patients during the period of spinal cord shock suffer from neurogenic bladder that necessitates catheter drainage, often leading to development of urinary tract infection. Patients with high cervical injury usually require mechanical ventilation and can develop ventilator-associated pneumonia. TABLE 56.1. FACTORS THAT PREDISPOSE TO NOSOCOMIAL INFECTIONS IN RELATION TO TIME OF INJURY Soon after the injury Prolonged initial hospitalization Surgical management of injuries to the spinal cord and possibly other organs Admission to the intensive care unit Administration of glucorticosteroids after the injury Bladder catheterization Mechanical ventilation Long after the injury Bladder catheterization Pressure ulcers Tracheostomy in patients with high cervical lesions Surgical interventions for chronic complications of the injury Although the likelihood of developing infection per hospital day appears to be the highest in the acute postinjury period, the vast majority of infections occur long after the injury. This is due to the fact that most such patients sustain spinal cord injury when still young and have an almost normal life expectancy. Since most patients with spinal cord injury chronically rely on bladder catheters for urinary drainage, urinary tract infection is the most common infection long after the injury. Second in frequency are infections associated with pressure ulcers. Patients with high cervical lesions are predisposed to tracheostomy- and endotracheal tube–related respiratory tract infections. Surgical management of the chronic sequelae of spinal cord injury can be complicated by surgical site infections. P.938 URINARY TRACT INFECTIONS Pathogenesis and Epidemiology Urinary tract infection is the most common infection in patients with spinal cord injury. Unique factors that predispose this population to urinary tract infection include bladder catheterization and urinary stasis (4). Although the sterile technique of bladder catheterization can theoretically be safer, at least in hospitalized patients, than the clean technique, both catheterization techniques can introduce microorganisms into the urinary tract. Urinary stasis impairs the naturally occurring mechanisms that protect the urinary tract from infection, including the washout effect of voiding and the phagocytic capacity of bladder epithelial cells. Multiplication of bacteria in the urine and invasion of host tissues are promoted in the presence of reduced bladder emptying, increased residual urine, and high bladder pressure (5). Although more than 90% of episodes of urinary tract infection in this population appear to involve only the lower urinary tract, serious complications can still arise secondary to such infections. Ascending infection of the urinary tract may evolve in the presence of vesicoureteral reflux or as a consequence of manipulations used to empty the bladder. Kidney infection with loss of renal function is particularly worrisome, because renal failure was once the leading cause of death in patients with spinal cord injury. Additionally, urinary tract infection can be associated with a number of anatomic changes, such as renal calculi (occupying the bladder, ureters, or kidneys), bladder diverticula and fibrosis, penile and scrotal fistulas, epididymoorchitis, and abscesses. The frequency of these anatomic changes depends on the type and duration of bladder drainage; these changes are most commonly detected in patients with indwelling bladder catheters. The vast majority of episodes of urinary tract infection in patients with spinal cord injury are caused by commensal bowel flora, primarily gram-negative bacilli and enterococci. The microbiology of microorganisms residing in the bladder can be affected by patients' gender, the place where pathogens are acquired (i.e., nosocomial vs. community acquisition), and the method of urinary drainage. For instance, Klebsiella pneumoniae is a very common cause of urinary tract infection in hospitalized patients (6, 7). In contrast, Escherichia coli and Enterococcus species cause more than two thirds of cases of urinary tract infection in female patients undergoing intermittent bladder catheterization (8). The presence of condom catheters increases the likelihood of colonizing the urethra and perineal skin with Pseudomonas, Klebsiella, and other Gram-negative bacilli. Although the virulence of Pseudomonas aeruginosa when isolated from the urinary tract of patients with spinal cord injury has been questioned (9), tissue invasion and bacteremia due to this microorganism can still occur (10, 11). As in able-bodied subjects, the presence of renal calculi in patients with spinal cord injury suggests etiology by urease-producing bacteria (12). Spinal cord injury units are no different from other types of specialized care units as to the occurrence of outbreaks of urinary tract infection due to multiresistant gram-negative bacilli (13). Polymicrobial growth is detected in almost half of positive urine cultures obtained from patients with spinal cord injury, particularly those with chronic indwelling urethral catheters (14). Clinical Manifestations Urinary tract infection may manifest differently in patients with spinal cord injury than in the general population. For instance, infected patients with spinal cord injury may not complain of dysuria, frequency, and urgency, which usually exist in able-bodied patients with urinary tract infection. Furthermore, suprapubic and flank pain or tenderness are not felt in insensate patients. More common manifestations of urinary tract infection in patients with spinal cord injury include worsening spasm, increasing dysreflexia, and change in voiding habits. Fever is usually, but not always, present. Diagnosis Diagnosis of urinary tract infection in patients with spinal cord injury can be problematic for several reasons. First, by masking urinary-specific symptoms, absent sensations constitute the single most important obstacle in the diagnosis of urinary tract infection in this population. Second, the unusual manifestations of urinary tract infection in these patients are nonspecific and may be caused by a variety of other infectious or noninfectious conditions, including osteomyelitis beneath pressure ulcers, ingrown toe nails, and heterotopic bone ossification. Third, bacteriuria, the cornerstone for diagnosing urinary tract infection, is nonspecifically prevalent in this population. Bacteriuria is most frequent in patients who have chronic indwelling bladder catheters, as cultures of randomly obtained urine samples yield bacterial growth in about 98% of instances (15). Even patients who undergo intermittent bladder catheterization have a 70% likelihood of being bacteriuric (16). Most cases of bacteriuria in patients with spinal cord injury represent asymptomatic bladder colonization. Although asymptomatic bladder colonization can progress to symptomatic infection, often it does not. Fourth, the finding of pyuria, which can reflect inflammation of the uromucosal lining and signal the transition from bladder colonization to symptomatic urinary tract infection is not specific for infection. Pyuria can be caused by a variety of noninfectious conditions, including catheter-induced trauma, renal stone, recent urologic procedure, and interstitial nephritis. Because of these potential problems in establishing a diagnosis, there exists no universally accepted definition of symptomatic urinary tract infection in patients with spinal cord injury. A commonly used definition of symptomatic urinary tract infection in these patients requires the presence of significant bacteriuria [≥105 colony-forming units (CFU)/mL], pyuria [>104 white blood cells (WBC)/mL of uncentrifuged urine or >10 WBC/high power field (hpf) for spun urine], and fever (>100ВF) plus more than one of the following signs and symptoms—(a) suprapubic or flank discomfort, (b) bladder spasm, (c) change in voiding habits, (d) increased spasticity, and (e) worsening dysreflexia—provided that no other potential etiologies for these clinical manifestations could be identified (6). Most healthcare providers tend to distinguish upper from lower urinary tract infection based on clinical manifestations and labora P.939 tory data rather than imaging findings. For example, the presence of high fever (>102ВF), chills, systemic toxicity, high grade leukocytosis (>20 thousands per mm3), and/or leukocyte casts in urinary sediment supports the presence of pyelonephritis. Prevention Mechanical Approaches Since the indwelling transurethral and suprapubic catheters pose a higher risk of infection than intermittent bladder catheterization, the latter method of bladder drainage should always be considered, barring any anatomic or functional constraints. Increasing the frequency of intermittent bladder catheterization can decrease the risk of urinary tract infection. Although the technique of clean nonsterile intermittent self-catheterization is considered rather safe for use by outpatients, sterile intermittent catheterization is implemented by most hospitals owing to the fear of nosocomial introduction of multiresistant and virulent mircroorganisms into the urinary tract. In patients with persistent or recurrent urinary tract infections, the urinary tract should be investigated for anatomic abnormalities (including abscess, stone, obstruction, and stricture) and functional alterations (such as vesicoureteral reflux, high residual volume of urine in bladder, and elevated bladder pressure). The use of drugs and surgical procedures to reduce bladder pressure and aid bladder emptying can help alleviate the risk of urinary tract infection. Antimicrobial Approaches Since asymptomatic bacteriuria can progress to symptomatic infection, it is theoretically possible that prevention or eradication of asymptomatic bacteriuria may decrease the rate of symptomatic urinary tract infection (6). Although instillation of antibiotic solutions, such as kanamycin-colistin, into the bladder at the time of intermittent catheterization may alleviate bacteriuria at the expense of a switch in the microbiology of clinical isolates (17), there exists no evidence that this practice prevents urinary tract infection. Studies that examined the administration of systemic antimicrobial agents in patients with spinal cord injury (18, 19) have yielded either conflicting or disappointing results. For example, the administration of trimethoprim-sulfamethoxazole has been shown in some studies (19), but not others (20), to significantly reduce the overall rate of asymptomatic bacteriuria and symptomatic urinary tract infection. This approach, however, was associated with relatively common adverse effects (19) and antibiotic resistance (20, 21). Therefore, systemic antimicrobial use is generally discouraged for prevention of asymptomatic bacteriuria in patients with spinal cord injury. Exceptions may include patients with (a) enlarging struvite stones associated with urea-splitting microorganisms, such as Proteus mirabilis and Providentia stuartii; (b) conditions that enhance the likelihood of developing significant complications from having asymptomatic bacteriuria, such as premature birth in pregnant women; and (c) recurrent episodes of upper urinary tract infection that are complicated by sepsis or other clinical complications, particularly if the recurrent infections are caused by the same microorganism. Based on the results of a recent study (22), antimicrobial treatment of asymptomatic bacteriuria in women with diabetes mellitus is probably not warranted. Bacterial Interference The limited success of traditional antimicrobial prophylaxis prompted interest in exploring the novel approach of bacterial interference (23). This approach is based on the principle that nonpathogenic microorganisms may prevent colonization of the urinary tract by pathogenic microorganisms. A preliminary nonrandomized, open-label clinical trial in patients with spinal cord injury who had suffered from frequent episodes of infection (24, 25, 26, 27) indicated that intentional colonization of the neurogenic bladder by a nonpathogenic strain of Escherichia coli 83972 reduced the rate of symptomatic urinary tract infection. The efficacy and safety of this approach is currently being investigated in a prospective, randomized clinical trial. INFECTIONS OF PRESSURE SORES AND UNDERLYING BONE Pathogenesis and Epidemiology Although the incidence of pressure ulcers varies among medical centers and is affected by the level and completeness of spinal cord injury, about one third of patients develop clinically relevant pressure ulcers at one time or another after the injury. Pressure sores delay rehabilitation, prolong hospital stay, and incur excessive costs, particularly when infected. Although pressure sores may develop either at home or while residing at a medical institution, most patients get admitted for management of the infectious complications of the ulcers. Factors that contribute to skin and soft tissue infection in the vicinity of pressure sores include break in skin integrity and bacterial contamination due to soiling of the ulcer by stools or urine. The former factor predisposes to infection by skin microorganisms including staphylococci and streptococci, whereas the latter factor promotes infection by gram-negative bacilli and anaerobic bacteria (28). Infected pressure sores involve mostly the ischial tuberosities, trochanters, and sacrum—areas that are anatomically exposed to high pressure and likely to be exposed to fecal or urinary microorganisms. Pressure sores often harbor multiple aerobes and anaerobes. The type of microbes colonizing the pressure sores was shown to be affected by the presence of devitalized tissue (29). Pressure sores with tissue necrosis had comparably high concentrations (>105 CFU/g) of both aerobes and anaerobes in deep tissues. However, upon excision of the necrotic tissue and healing of the infected pressure sores, the bacterial density dropped to less than 104 CFU/g of tissue, accompanied by disappearance of anaerobes. Bacteroides species, Peptostreptococcus, E. coli, Proteus species, and enterococci were the most prominent microorganisms isolated from necrotic pressure sores. In contrast, P. aeruginosa and S. aureus were the two microorganisms that were most frequently isolated from healing pressure sores. There exists some variability in the culture results of deep tissue obtained from different parts of the pressure sore, and the value of obtaining repeated cultures of pressure ulcers remains in question. The P.940 polymicrobial spectrum of flora in pressure ulcers in children is rather similar to that in adults, but, additionally, includes Haemophilus influenzae (30). Candida infection of pressure sores in patients with spinal cord injury is unusual. In patients with spinal cord injury, most cases of osteomyelitis occur beneath pressure sores. Most such cases are caused by two or more bacterial species, including gram-positive cocci (particularly S. aureus and Streptococcus species), gram-negative bacilli (including the Enterobacteriaceae group and P. aeruginosa), and anaerobic bacteria (mainly Bacteroides species) (31). Vertebral and cranial osteomyelitis may also occur in association with spinal hardware and cervical halos, respectively. Clinical Manifestations Infection of pressure sores can be associated with cellulitis, abscess formation, osteomyelitis of underlying bone, septic arthritis, infected bursae, and septicemia. Local signs of infection include erythema, drainage, and foul-smelling or purulent drainage. Systemic manifestations of fever and leukocytosis commonly, but not invariably, occur. Septicemia is much rarer in the context of osteomyelitis beneath pressure ulcers in patients with spinal cord injury than in able-bodied adult patients with spinal osteomyelitis or children with long bone osteomyelitis. Clinically relevant blood cultures in patients with infected pressure ulcers suggests the presence of soft tissue abscess (28) or, less commonly, an infected hematoma. Diagnosis A number of factors can impede making a proper diagnosis of infection of pressure sores with or without underlying osteomyelitis in patients with spinal cord injury: Inadequate History Patients with spinal cord injury usually have no or altered sensations in the area of the pressure ulcers. Since most pressure ulcers occur in the trochanteric, ischial, and sacral regions, immobile patients cannot directly visualize the ulcers. Furthermore, such patients often complain of neurogenic or referred pain that may have no relation to the infection. These factors result in frequently obtaining an incomplete or inaccurate history from patients and help underscore the diagnostic importance of performing comprehensive physical examination by healthcare providers. Microbiologic Uncertainties Since pressure sores are universally colonized by bacteria, swab cultures of open ulcers should not be obtained unless infection is clinically evident. Sinus tract cultures are also usually unreliable. Cultures of material obtained by needle aspiration tend to overestimate the number of bacterial isolates (32). Although cellulitis adjacent to a pressure sore can theoretically be caused by a microorganism(s) present in the pressure sore, there is no evidence that skin biopsy in patients with spinal cord injury yields clinically relevant results. Cultures of biopsied deep soft tissue remains the most accurate means for determining the microbiologic cause of soft tissue infection. In patients with underlying osteomyelitis, swab cultures of pressure sores do not accurately predict the microorganisms causing bone infection (31). Moreover, since fibrotic tissue adherent to bone is usually colonized with bacteria, bone cultures are positive in at least two thirds of patients in whom histopathologic examination of bone tissue is incompatible with osteomyelitis (31). Therefore, osteomyelitis should not be diagnosed solely by positive cultures of biopsied bone. Radiologic Limitations Another diagnostic problem in patients with spinal cord injury arises from the limited ability to delineate the extent and depth of infection in association with pressure sores. Deep soft tissue abscesses can exist beneath apparently healed pressure sores. Although highly sensitive for detecting soft tissue abscesses, radionuclide scans can yield false-positive results in patients with spinal cord injury who have an infected pressure sore without an associated abscess (33). Computed tomography (CT) and magnetic resonance imaging (MRI) can detect abscesses in both soft tissue and muscle, as is the case with the infrequently diagnosed iliopsoas abscess (34). Since pressure necrosis affects subcutaneous and muscular tissues more than skin, the visualized skin opening of a sinus tract may seem deceptively small. Probing of the sinus tract, although generally helpful, may still not reveal the full depth of the sinus tract. Sinography can better delineate the full depth of the sinus tract and reveal potential communications with bone, joint, visceral organs, or deep-seated abscesses. In patients with nonhealing pressure sores who have persistent or recurrent infection, injection of dye into the bladder or intestines may help establish the presence of fistulous communications. Misinterpretation of the findings of imaging studies is particularly prominent when attempting to diagnose bone infection beneath pressure ulcers. Bone scan is very sensitive (almost 100%) but poorly specific (<33%) for diagnosing osteomyelitis beneath pressure sores (35). The low specificity of bone scan is attributed to the aggregation of technetium in areas of bone that are affected by pressure-induced changes and in foci of heterotopic bone ossification. Therefore, bone scan should be used primarily for its high negative predictive value (i.e., in an attempt to rule out osteomyelitis and, therefore, obviate the need for bone biopsy) rather than its low positive predictive value (i.e., to diagnose osteomyelitis). Neither clinical evaluation (duration of ulcer, bone exposure, purulent drainage, fever, peripheral WBC count, and erythrocyte sedimentation rate) nor radiologic examination (plain roentgenogram and bone scan) correlates well with the likelihood of finding histopathologic evidence for bone infection (31, 35, 36). Although the finding of bone changes by CT scan or MRI can be very helpful in supporting the diagnosis of osteomyelitis, there are no studies in patients with spinal cord injury that correlate the abnormal findings of these imaging studies with bone biopsy results. P.941 Multiple Sores Patients with spinal cord injury often have multiple pressure sores. In such patients, infection of soft tissue and/or bone may exist at some sites but not others. Furthermore, different sites may be infected by different microorganisms. Because of the above-described diagnostic limitations, definitive diagnosis of osteomyelitis beneath pressure sores requires histopathologic examination of bone tissue (35, 36). Percutaneous needle bone biopsy yields histopathologic evidence for infection of bone beneath nonhealing pressure sores in only one fifth to one third of cases (31, 35, 36). These findings support the clinical observation that nonhealing of pressure sores is much less likely to result from underlying osteomyelitis than from noninfectious conditions, such as pressure-related changes, malnutrition, anemia, heterotopic bone ossification, and spasticity. Prevention The process of preventing infection of pressure sores and underlying bone starts with preventing the development of pressure sores. This consists of quality nursing care, frequent turning of the patient for pressure relief, careful attention to bony prominences, avoidance of friction and shear forces, correction of anemia, adequate nutrition, and training patients and their attendants in skin care. The relationship between bacterial counts in wounds and delayed healing remains controversial, and their exists no evidence from prospective randomized studies that local or systemic antimicrobial agents enhance wound healing or prevent infection in patients with spinal cord injury. Systemic antibiotics, however, ought to be given perioperatively in patients undergoing myocutaneous flap surgery (37, 38). Although perioperatively administered antibiotics are typically active against the gram-positive skin flora, a broader spectrum regimen that provides additional coverage against gram-negative bacilli and anaerobes may be warranted if supported by the results of preoperative or intraoperative wound cultures. There exists no convincing evidence to support the prevailing practice of continuing perioperative antibiotics until wound drains are removed, usually 10 to 14 days after myocutaneous flap surgery. RESPIRATORY TRACT INFECTIONS Pathogenesis and Epidemiology The most serious respiratory tract infection is pneumonia, which constitutes the leading cause of death due to infection in this population (2). Pneumonia is the most common pulmonary complication in the immediate postinjury period (39), and is particularly likely to occur in the first few months after cervical or high thoracic injury and among quadriplegics and persons older than 55 years (2). A five-center study of respiratory complications following spinal cord injury found that almost one third of patients developed pneumonia while undergoing initial rehabilitation in the hospital (40). Factors that predispose patients with spinal cord injury to develop pneumonia or tracheitis include (a) weakness of the diaphragmatic and intercostal muscles in patients with cervical or high thoracic spinal cord injury, which would impair the capacity to clear respiratory secretions; (b) indwelling respiratory devices, such as endotracheal or tracheostomy tubes; and (c) aspiration that is promoted either by an abnormal state of consciousness due to illicit drug ingestion or associated head injury or by paralytic ileus that often occurs soon after spinal cord injury (40). The microbiology of nosocomial respiratory tract infections in this population is affected by the type of predisposing factor(s). For example, S. aureus (particularly MRSA) and P. aeruginosa are the two most common causes of pneumonia and tracheitis in patients with respiratory devices, whereas aspiration pneumonia is mostly caused by gram-negative and anaerobic bacteria. Clinical Manifestations Patients with cervical or thoracic spinal cord injury can have absent or altered sensations of chest pain and dyspnea. Infected patients with weakness of the diaphragmatic and intercostal muscles may also have no or minimal cough, and are unlikely to spontaneously produce sputum. In such patients, the only clinical manifestations of pneumonia may consist of physical findings (distressed appearance, fever, tachypnea, and tachycardia) and abnormal test results (leukocytosis, hypoxemia, and infiltrates on chest radiographs). Diagnosis Because of ineffective cough, patients with cervical or high thoracic lesions may not be able to provide adequate sputum samples for Gram stain and cultures. If tracheal secretions cannot be adequately suctioned, bronchoscopy may be required for both diagnostic and therapeutic purposes. The most prominent impediment to diagnosing pneumonia in patients with spinal cord injury arises from the limited ability to clinically distinguish pneumonia from a number of noninfectious pulmonary complications, including atelectasis, chemical pneumonitis, pulmonary embolism, and fat embolism (41). For instance, atelectasis, like pneumonia, commonly occurs in patients with cervical or high thoracic spinal cord injury who retain pulmonary secretions and can also manifest with fever. Furthermore, the site of pulmonary involvement may not help differentiate atelectasis from pneumonia since both conditions predominantly affect the left lung. Chemical pneumonitis due to aspiration can also mimic bacterial pneumonia. When adequate samples of respiratory secretions are available, microbiologic examination may help distinguish between these two clinical entities by showing a plethora of microorganisms (along with WBCs) in samples obtained from patients with bacterial pneumonia. Pulmonary embolism can also be clinically confused with pneumonia (42). This is partially attributed to the fact that the majority of patients with spinal cord injury disclose no thrombotic source for pulmonary embolism (43). Furthermore, since patients with spinal cord injury commonly display baseline roentgenographic changes in the lungs due to atelectasis or other causes that make it difficult to interpret ventilation-perfusion lung scans, a definitive diagnosis of pulmonary embolism often requires pulmonary angiography. P.942 Fat embolism, which can occur acutely after spinal cord injury in association with fracture of long bones, may be suspected if petechiae and cerebral dysfunction are present. Prevention Potential approaches for preventing pneumonia in patients with spinal cord injury include some that center around control of predisposing conditions and others that provide antimicrobial activity. The first group of approaches is intended to augment cough and lessen retention of secretions. Cough can be assisted by using abdominal binders or corsets. Adequate hydration, chest physical therapy, and postural drainage can enhance drainage of secretions, although it may be difficult to achieve certain optimal positions during the acute period following spinal cord injury. Antimicrobial approaches include antibiotics and immunization. In general, the use of systemic antibiotics for prevention of pneumonia in high-risk patients with spinal cord injury is not advocated. Because pneumonia can either occur more frequently or result in more serious complications in patients with spinal cord injury than in the general population, eligible patients should be immunized against potentially preventable causes of pneumonia. Almost two thirds of patients with spinal cord injury are eligible for vaccination against Streptococcus pneumoniae and influenza virus by virtue of old age, chronic respiratory disease, and/or residence in chronic-care facilities. The antibody response to pneumococcal (44) and influenza vaccination of patients with spinal cord injury appears adequate. Although there have been no prospective studies of the clinical efficacy of these vaccinations in patients with spinal cord injury, it is generally recommended that patients at risk receive influenza vaccine every year and pneumococcal vaccine every 5 years. COLONIZATION AND INFECTION BY MULTIRESISTANT MICROORGANISMS Patients in spinal cord injury units may acquire multiresistant microorganisms while residing at a referring institution (hospital or nursing home) or another unit (particularly the intensive care unit) within the same hospital. Alternatively, patients may acquire multiresistant microorganisms while hospitalized at the spinal cord injury unit either directly from already colonized patients or indirectly via the hands of healthcare providers (who care for colonized persons) or contaminated inanimate surfaces (in patients' rooms, rehabilitation areas, and whirlpools). Fortunately, most cases of growth of multiresistant microorganisms in clinical specimens represent colonization rather than clinical infection. The most commonly studied multiresistant microorganism in spinal cord injury units is MRSA, which accounts for at least half of all clinical isolates of S. aureus. The generally problematic diagnosis of infection in these insensate patients makes it sometimes difficult to distinguish between clinical infection and colonization. This microorganism most frequently infects the urinary tract, wounds, lungs, and blood. The sites that are most commonly colonized by MRSA include the anterior nares, wounds, urine, perineum, and stools (45). Patients may remain colonized with MRSA for months or years. Although the combination of an oral regimen of minocycline and rifampin and topical mupirocin was found to be effective in eradicating MRSA colonization, it is unwise to routinely attempt to eradicate MRSA colonization in this population of patients (46). Unfortunately, transfer of hospitalized patients to nursing homes may be delayed until MRSA colonization is eradicated. The prevalence of VRE in spinal cord injury units appears to have increased in recent years. For instance, preliminary findings from our center indicated that the gastrointestinal tract of one third to one half of patients residing in the spinal cord injury unit is colonized with VRE (47). In the vast majority of instances, isolation of VRE from stools was not associated with clinical infection. Molecular typing demonstrated that the majority of VRE isolates had distinctly different patterns, even in the case of patients sharing bedrooms. These findings suggested that nosocomial transmission of VRE within the spinal cord injury unit was rather unusual. Patients with spinal cord injury often harbor multiresistant gram-negative bacilli that produce extended spectrum ОІ-lactamases (ESBL). Such microorganisms are isolated mostly from the urine, wounds, and respiratory secretions. Most urinary ESBL-producing isolates belong to the Klebsiella–Enterobacter group of microorganisms that are fully susceptible only to carbapenems; some isolates are also susceptible to aminoglycosides. REFERENCES 1. National Institute on Disability and Rehabilitation Research (NIDRR) Consensus Statement. The prevention and management of urinary tract infection among people with spinal cord injuries. J Am Paraplegia Soc 1992;15:194–207. 2. DeVivo MJ, Kartus PL, Stover SL, et al. Cause of death for patients with spinal cord injuries. Arch Intern Med 1989;149:1761–1766. 3. Sugarman B, Brown D, Musher D. Fever and infection in spinal cord injury patients. JAMA 1982;248:66–70. 4. Stover SL, Lloyd LK, Waites KB, et al. Urinary tract infection in spinal cord injury. Arch Phys Med Rehabil 1989;70:47–54. 5. Merritt JL. Residual urine volume: correlate of urinary tract infection in patients with spinal cord injury. Arch Phys Med Rehabil 1981;62:558–561. 6. Darouiche R, Cadle R, Zenon G, et al. Progression from asymptomatic to symptomatic urinary tract infection in patients with SCI: a preliminary study. J Am Paraplegia Soc 1993:16:221–226. 7. Kil KS, Darouiche RO, Hull RA, et al. Identification of a Klebsiella pneumoniae strain associated with nosocomial urinary tract infection. J Clin Microbiol 1997;35:2370–2374. 8. Bennett CJ, Young MN, Darrington H. Differences in urinary tract infection in male and female spinal cord injury patients on intermittent catheterization. Paraplegia 1995;33:69–72. 9. Lindan R, Joiner E. A prospective study of the efficacy of low dose nitrofurantoin in preventing urinary tract infections in spinal cord injury patients with comments on the role of pseudomonads. Paraplegia 1984;22:61–65. 10. Montgomerie JZ, Guerra DA, Schick DG, et al. Pseudomonas urinary tract infection in patients with spinal cord injury. J Am Paraplegia Soc 1989;12:8–10. 11. Montgomerie JZ, Chan E, Gilmore DS, et al. Low mortality among patients with spinal cord injury and bacteremia. Rev Infect Dis 1991;13:867–871. 12. DeVivo MJ, Fine PR. Predicting renal calculus occurrence in spinal cord injury patients. Arch Phys Med Rehabil 1986;67:722–725. P.943 13. Simor AE, Ramage L, Wilcox L, et al. Molecular and epidemiologic study of multiresistant Serratia marcescens infections in a spinal cord injury rehabilitation unit. Infect Control Hosp Epidemiol 1988;9:20–27. 14. Darouiche RO, Priebe M, Clarridge JE. Limited vs full microbiological investigation for the management of symptomatic polymicrobial urinary tract infection in adult spinal cord-injured patients. Spinal Cord 1997;35:534–539. 15. Warren JW, Tenney JH, Hoopes JM, et al. A prospective microbiologic study of bacteriuria in patients with chronic indwelling urethral catheters. J Infect Dis 1982;146:719–723. 16. McGuire EJ, Savastano JA. Long-term followup of spinal cord injury patients managed by intermittent catheterization. J Urol 1983;129:775–776. 17. Pearman JW. The value of kanamycin-colistin bladder instillations in reducing bacteriuria during intermittent catheterization of patients with acute spinal cord injury. J Urol 1979;51:367–374. 18. Mohler JL, Cowen DL, Flanigan RC. Suppression and treatment of urinary tract infection in patients with an intermittently catheterized neurogenic bladder. J Urol 1987;138:336–340. 19. Gribble MJ, Puterman ML. Prophylaxis of urinary tract infection in persons with recent spinal cord injury: a prospective, randomized, double-blind, placebo-controlled study of trimethoprim-sulfamethoxazole. Am J Med 1993;95:141–152. 20. Sandock DS, Gothe BG, Bodner DR. Trimethoprim-sulfamethoxazole prophylaxis against urinary tract infection in the chronic spinal cord injury patient. Paraplegia 1995;33:156–160. 21. Jiminez EM, Schick DG, Canawati HN, et al. Klebsiella pneumoniae colonization of the bowel associated with the use of trimethoprim-sulfamethoxazole. Eur J Clin Microbiol 1982;1:253–254. 22. Harding GKM, Zhanel GG, Nicolle LE, et al. Antimicrobial treatment in diabetic women with asymptomatic bacteriuria. N Engl J Med 2002;347:1576–1583. 23. Darouiche RO, Hull RA. Bacterial interference for prevention of urinary tract infection: an overview. J Spinal Cord Med 2000;23:136–141. 24. Hull RA, Rudy DC, Donovan WH, et al. Virulence properties of Escherichia coli 83972, a prototype strain associated with asymptomatic bacteriuria. Infect Immun 1999;67:429–432. 25. Hull RA, Rudy DC, Donovan WH, et al. Urinary tract infection prophylaxis using Escherichia coli 83972 in spinal cord injured patients. J Urol 2000;163:872–877. 26. Hull RA, Donovan WH, del Terzo M, et al. Role of type 1 fimbria- and P fimbria-specific adherence in colonization of the neurogenic human bladder by Escherichia coli. Infect Immun 2002;70:6481–6484. 27. Darouiche RO, Donovan WH, del Terzo M, et al. Pilot trial of bacterial interference for preventing urinary tract infection. Urology 2001;58:2339–2344. 28. Sugarman B. Infection and pressure sores. Arch Phys Med Rehabil 1985;66:177–179. 29. Sapico FL, Ginunas VJ, Thornhill-Joynes M, et al. Quantitative microbiology of pressure sores in different stages of healing. Diagn Microbiol Infect Dis 1986;5:31–38. 30. Brook I. Microbiological studies of decubitus ulcers in children. J Pediatr Surg 1991;26:207–209. 31. Thornhill-Joynes M, Gonzales F, Stewart CA, et al. Osteomyelitis associated with pressure ulcers. Arch Phys Med Rehabil 1986;67:314–318. 32. Rudensky B, Lipschits M, Isaacsohn M, et al. Infected pressure sores: comparison of methods for bacterial identification. South Med J 1992;85:901–903. 33. Firooznia H, Rafii M, Golimbu C, et al. Computerized tomography of pelvic osteomyelitis in patients with spinal cord injuries. Clin Orthop 1983;126–131. 34. Rubayi S, Soma C, Wang A. Diagnosis and treatment of iliopsoas abscess in spinal cord injury patients. Arch Phys Med Rehabil 1993;74:1186–1191. 35. Sugarman B. Pressure sores and underlying bone infection. Arch Intern Med 1987;147:553–555. 36. Darouiche RO, Landon GC, Klima M, et al. Osteomyelitis associated with pressure sores. Arch Intern Med 1994;154:753–758. 37. Salzberg CA, Gray BC, Petro JA, et al. The perioperative antimicrobial management of pressure ulcers. Decubitus 1990;3:24–26. 38. Garg M, Rubayi S, Montgomerie JZ. Postoperative wound infections following mycocutaneous flap surgery in spinal injury patients. Paraplegia 1992;30:734–739. 39. Fishburn MJ, Marino RJ, Ditunno JF Jr. Atelectasis and pneumonia in acute spinal cord injury. Arch Phys Med Rehabil 1990;71:197–200. 40. Jackson AB, Groomes TE. Incidence of respiratory complications following spinal cord injury. Arch Phys Med Rehabil 1994;75:270–275. 41. Reines HD, Harris RC. Pulmonary complications of acute spinal cord injuries. Neurosurgery 1987;21:193–196. 42. Dee PM, Suratt PM, Bray ST, et al. Mucous plugging simulating pulmonary embolism in patients with quadriplegia. Chest 1984;85:363–366. 43. Waring WP, Karunas RS. Acute spinal cord injuries and the incidence of clinically occurring thromboembolic disease. Paraplegia 1991;29:8–16. 44. Darouiche RO, Groover J, Rowland J, et al. Pneumococcal vaccination for patients with spinal cord injury. Arch Phys Med Rehabil 1993;74:1354–1357. 45. Darouiche R, Wright C, Hamill R, et al. Eradication of methicillin-resistant Staphylococcus aureus by using oral minocycline-rifampin and topical mupirocin. Antimicrob Agents Chemother 1991;35:1612–1615. 46. Maeder K, Ginunas VJ, Montgomerie JZ, et al. Methicillin-resistant Staphylococcus aureus (MRSA) colonization in patients with spinal cord injury (SCI). Paraplegia 1993;31:639–644. 47. Byers PA, Koza MA, Abraham FP, et al. Prevalence of vancomycin resistant Enterococcus (VRE) in spinal cord injury patients. The 42nd Interscience Conference on Antimicrobial Agents and Chemotherapy (abstract K-1951), San Diego, CA, 2002.
Воскресенье, 06 Ноябрь 2011 | 961 Просмотры | Печать | PDF
Infections remain a significant complication and a leading cause of mortality, particularly within the first year after transplantation. Most infections in transplant recipients are nosocomially acquired and represent either opportunistic infections resulting from iatrogenic immunosuppression or infections resulting from conventional nosocomial pathogens. Within the last decade, the incidence of several opportunistic infections [e.g., cytomegalovirus (CMV) and Pneumocystis carinii pneumonia (PCP)] has declined dramatically, largely because of the advent of effective prophylaxis. Instead, nosocomial infections (primarily resulting from bacteria) transmitted from environmental reservoirs or harbored as a result of endogenous colonization in nosocomial settings have emerged as leading infections in organ transplant recipients. In a recent study in liver transplant recipients, 82% of the episodes of fever documented in consecutive patients over a 2-year period were nosocomial, of which 62% were bacterial in origin (1). Fifty-three percent of all infections in heart transplant recipients in another study were nosocomially acquired, and of these 63% were bacterial (2). Paralleling the trends in nosocomially acquired infections, antimicrobial resistance is increasingly recognized as a problem in the transplant setting. It is noteworthy, however, that the emergence of several of the antibiotic-resistant pathogens was first documented in transplant recipients (3). For example, vancomycin-resistant enterococci (VRE) were initially discovered in liver transplant recipients at several institutions where they eventually became a more widespread problem. Transplant recipients are uniquely vulnerable to colonization and infection resulting from nosocomial pathogens. Within the same institution, transplant recipients have been shown to have a significantly higher incidence of nosocomial infections than nontransplant patients (4). The predilection of immunocompromised patients to Legionella infection is well recognized. However, it is notable that within this subgroup, transplant recipients have the highest risk (5). Among patients undergoing surgical procedures at one institution where legionellosis was documented, renal transplant recipients had an attack rate of 50%, whereas the general hospital population experienced an attack rate of only 0.4% (6). Transplant recipients exposed to tuberculosis during an institutional outbreak were more likely to contract Mycobacterium tuberculosis as compared with nontransplant contacts of the source case (7). During a nosocomial outbreak of extended-spectrum ОІ-lactamase–producing Escherichia coli, 67% of patients on the liver transplant service, but no other surgical patients on the same floor, were shown to be colonized or infected with the outbreak isolate (8). This chapter discusses the potential sources, unique risk factors, prevention, and infection control implications for nosocomial infections that may be acquired during or after transplantation. SOURCES OF NOSOCOMIAL INFECTIONS Donor-Transmitted Infections Latent Infections in the Donor Viral infections latent in the donor have by far the greatest potential for transmission by the transplanted organ and exert a more profound clinical impact in the allograft recipient compared with many other donor-transmitted infections. Thus, serologic screening of the donor for hepatitis B virus (HBV), hepatitis C virus (HCV), CMV, human immunodeficiency virus (HIV), and human T-cell leukemia virus type I (HTLV-I) is routinely recommended (9). Hepatitis B Virus The risk of transmission of HBV varies according to the HBV serologic profile of the donor and the recipient and the type of organ transplanted. Allografts from hepatitis B surface antigen (HBsAg)-positive donors can transmit HBV infection not only to recipients who are HBsAg negative but also to the ones who are HBsAg positive. Thus, organ donation from HBsAg-positive donors is recommended only in life-threatening situations, particularly if the recipient has no antibody to hepatitis B surface antigen (anti-HBs). Anti-HBs–positive donors who are negative for HBsAg and antibody to hepatitis B core antigen (anti-HBc) are generally considered unlikely to transmit HBV. HBV DNA, however, may be detectable in these patients by the polymerase chain reaction (PCR). Therefore, the potential for HBV transmission exists for donors with isolated anti-HBs positivity, unless anti-HBs positivity in the donor is the result of HBV vaccination or administration of hepatitis B immunoglobulin (HBIG) (10). Anti-HBc positivity in the absence of HBsAg poses a low but not negligible likelihood of transmission of HBV. The liver may P.986 continue to harbor the replicative form of HBV in such donors with the potential of HBV transmission, particularly with the hepatic allograft. Donors with isolated anti-HBc positivity should be considered infectious, although the risk of transmission, especially for the recipients of extrahepatic organs, is low. None of the seven heart transplant recipients, 2.3% (1/42) of the renal, and 50% (3/6) of the liver transplant recipients who received organs from isolated anti-HBc positive donors became infected (11). General consensus is that the organs from anti-HBC positive donors should be used for recipients who are HBsAg positive or who have received HBV vaccine. Transplantation of an anti-HBc positive liver into a nonimmune recipient should be performed only if deemed medically urgent and under a prophylactic regimen of lamivudine with or without HBIG (12). Hepatitis C Virus Approximately 5% of all cadaveric organ donors are positive for antibody to HCV (anti-HCV), and 50% of these have detectable HCV viremia by PCR (9). Although anti-HCV–positive donor organs may transmit HCV infection, discarding all HCV-positive organs is not feasible. Transplantation of livers from HCV-positive donors into HCV-positive recipients has not been associated with a decrease in graft or patient survival at 1 to 5 years (13, 14). Most transplant centers use HCV-positive donor kidneys only for HCV-positive recipients. The use of anti-HCV–positive organs in anti-HCV–negative recipients may be considered in emergency or life-threatening situations or in selected circumstances, for example, the elderly kidney transplant candidates with limited lifespan or diabetics with end-stage renal disease in whom the gain in quality of life after kidney transplantation might outweigh the risk of HCV, particularly because the long-term outcome of HCV infection in transplant recipients has not been well defined. Herpesviruses The donated allograft is a significant and an efficient source of transmission of CMV (15, 16). The morbidity from infection is greatest in CMV-seronegative recipients of CMV-positive allografts. Superinfection (i.e., infection with an exogenous strain of CMV in patients with prior evidence of CMV infection) has also been documented. Symptomatic CMV disease occurred more frequently in patients infected with the new CMV strain compared with those with reactivation of the latent virus (17). Donor transmission (documented by molecular typing) has also been demonstrated with other herpesviruses, including herpes simplex virus (HSV), Epstein-Barr virus, and human herpesvirus-6 (HHV-6) (18, 19, 20, 21). Human Immunodeficiency Virus Donor positivity for HIV by enzyme-linked immunosorbent assay (ELISA) is considered an absolute contraindication to organ donation. Living donors who test negative for HIV antibody may be infectious during the window period between the acquisition of HIV infection and seroconversion. Although HIV p24 antigen may potentially detect HIV infection in such cases, this test is not yet routinely available for organ donors (9). According to the Centers for Disease Control and Prevention's guidelines for HIV screening practices for organ donation, social history suggestive of HIV exposure should be sought in all potential organ donors and may indicate an increased risk of HIV transmission, even in the presence of negative HIV serology (9). Human T-Cell Leukemia Virus Type I Although HTLV-I has been documented to be transmitted by blood, the risk of transmission by organ transplantation has not been clearly discerned. Except in life-threatening circumstances, HTLV-I positivity is considered a contraindication to organ donation (9). Other Pathogens Transmission of syphilis by organ donation has never been conclusively demonstrated. However, routine serologic screening of all potential donors for syphilis is recommended; recipients of such organs should receive penicillin as prophylaxis posttransplant (22). Transmission of M. tuberculosis to recipients receiving allografts from donors with active tuberculosis has been documented. Transmission of M. tuberculosis to two renal transplant recipients from a donor with unrecognized tuberculous meningitis at the time of organ retrieval has been reported (23). Tuberculin-positive donors without clinically overt tuberculosis may also transmit tuberculosis. It is recommended that the recipients of allografts from donors with tuberculin reactivity or a history of tuberculosis receive chemoprophylaxis for tuberculosis after transplantation (24). Toxoplasma gondii, because of its predilection for latency in muscle tissue, poses a considerable risk for transmission in heart transplant recipients; 50% to 75% of the seronegative recipients of T. gondii antibody-positive allografts have developed toxoplasmosis. Because of the paucity of Toxoplasma cysts in noncardiac tissue, toxoplasmosis is rarely transmitted by the transplanted organs in renal and liver transplant recipients. Acquired Infections in the Donor Life-sustaining measures in critically ill donors may render them susceptible to nosocomially acquired infections with the potential for transmission to the allograft recipients. Two recent studies comprising a large number of patients have shown that donor bacteremia did not portend a higher risk of infectious complications or compromise graft or patient survival (25, 26). The most frequent cause of the donor bacteremias in these studies was gram-positive bacteria, of which Staphylococcus aureus was the predominant pathogen. Most recipients of organs retrieved from bacteremic donors in the aforementioned studies received antimicrobial therapy. In the study by Lumbreras et al. (25), specific antibiotics were administered to the recipients for 7 to 10 days on receipt of donor blood culture results. In the report by Freeman et al. (26), 91% of the recipients received antibiotics for a mean of 3.8 days. These data suggest that with appropriately administered antibiotic therapy, organs from bacteremic donors can be successfully transplanted without incurring an additional risk for infection or allograft dysfunction in the recipient. A similar dilemma exists regarding the feasibility of using P.987 organs from donors with bacterial meningitis (27). Lopez-Navidad et al. (27) described the outcome in 16 recipients who had received organs from five patients with bacterial meningitis. The pathogens included Neisseria meningitidis, Streptococcus pneumoniae, and E. coli. With antibiotic administration ranging from 5 to 10 days, infection caused by the aforementioned bacteria were not documented in any of the recipients. Thus, patients with brain death attributable to bacterial meningitis caused by these bacteria can also be suitable organ donors, if the donor and the recipient receive appropriate antibiotic therapy. An exception, however, is donors with a less commonly encountered bacterial infection, that is, M. tuberculosis. Unrecognized active M. tuberculosis infection in the donor can be efficiently transmitted to the recipient with deleterious sequelae. Donor organs colonized with Candida or Aspergillus may transmit the fungi to lung and heart-lung transplant recipients. Karyotypic analysis of the Candida albicans isolates demonstrated identical strains from the donor lung and C. albicans isolates causing disseminated infection in a lung transplant recipient (28). Donor organs have also been documented to transmit other fungal infections (e.g., Cryptococcus neoformans and Histoplasma capsulatum) (29). Contamination during Organ Procurement Contamination during harvesting and preservation of the allograft has been reported to occur in 2% to 23% of the kidney allografts. Although some bacteria (e.g., Staphylococcus epidermidis, diphtheroid species, and Propionibacterium acnes) present little risk of infection to the allograft recipient, more virulent pathogens (e.g., gram-negative rods, particularly Pseudomonas aeruginosa; S. aureus; Bacteroides species; and fungi) cultured from the donor or preservation fluid can lead to serious infections (e.g., mycotic aneurysm and anastomotic rupture) in kidney transplant recipients (29). Blood Products Although CMV infection has been shown to be transmitted by blood products in organ transplant recipients, the risk is small and has not been shown to correlate with the number of blood products transfused (21). Over a 13 year-period, only 2.6% (3/112) of CMV-seronegative recipients who received CMVnegative renal, heart, lung, or liver allografts were documented to develop transfusion-associated CMV infection (30). Furthermore, transfusion, compared with donor-transmitted CMV infection, has been associated with a lower frequency of symptomatic disease and, therefore, has a less profound clinical impact (31). Although some centers use CMV-seronegative blood products for recipients who are seronegative for CMV, routine testing of blood products for CMV is not recommended. Since May 1990, all blood products in the United States have been routinely screened for HCV. Consequently, the risk of posttransfusion HCV has declined from 8% to 10% to less than 1% currently. Environmental Reservoirs and Sources Environmental sources are significant sites for acquisition of a number of infectious agents, particularly nosocomial pathogens in transplant recipients (Table 59.1). Most cases of Legionella in solid organ transplant recipients are nosocomially acquired (32). The source of posttransplant legionellosis in all studies where an environmental link was sought was the hospital's potable water distribution system (5). Restriction fragment length polymorphism patterns documented that the hospital's central hot water supply was the source of legionellosis in a hospital where 14 cases were documented in transplant recipients over an 8-year period (33). Nosocomial legionellosis in heart-lung transplant recipients at one institution was linked to a contaminated ice machine (34). TABLE 59.1. MODE OF ACQUISITION OF MAJOR PATHOGENS IN TRANSPLANT RECIPIENTS Pathogen Mode of acquisition Viruses Cytomegalovirus Seronegative recipient Donor transmission, rarely transfusions Seropositive recipient Reactivation and donor transmission Herpes simplex virus Reactivation, rarely donor transmission Varicella zoster virus Reactivation, rarely donor transmission Human herpesvirus-6 Reactivation and donor transmission Hepatitis C virus Reactivation, unless donor antihepatitis C virus positive Hepatitis B virusaa Rarely donor transmission Adenovirus Donor and nosocomial transmission Respiratory viral infections Nosocomial and community acquisition Bacteria Staphylococcus aureus Endogenous nasal colonization, nosocomial transmission Vancomycin-resistant enterococci Nosocomial transmission, endogenous gastrointestinal colonization Pseudomonas aeruginosa Nosocomial environmental acquisition Enterobacteriaceae Endogenous infection, nosocomial transmission Environmental acquisition Legionella Environmental acquisition Mycobacterium tuberculosis Reactivation, donor transmission, nosocomial transmission Fungi Candida Endogenous infection (liver transplants), donor transmission (lung transplants) Aspergillus Environmental acquisition Pneumocystis carinii Reactivation, possibly nosocomial transmission Cryptococcus neoformans Unknown Protozoa Toxoplasma gondii Donor transmission, rarely reactivation aHBsAg-positive donors can transmit hepatitis B virus (HBV) but are not considered acceptable organ donors. Rarely anti-HBs-positive donors (particularly of hepatic allografts) can transmit HBV. Outbreaks of invasive aspergillosis in transplant recipients have been linked to construction or demolition activity within or near a hospital; contaminated or poorly maintained ventilation ducts, grids, or air filters; and other dust-generating activities that may aerosolize Aspergillus spores. Accommodation of marrow transplant recipients outside of rooms with laminar air flow and high-efficiency particulate air filters during periods of neutropenia have been shown to be a risk factor for invasive aspergillosis (35). P.988 A seasonal variation in the incidence of invasive aspergillosis, coinciding with a high outdoor concentration of airborne spores in late summer or fall and a lower concentration in the winter months, has also been observed. The prevailing belief that Aspergillus is predominantly an airborne pathogen acquired via inhalation has recently been challenged. It has been proposed that Fusarium and Aspergillus can be detected in hospital water systems, and aspiration, as opposed to inhalation of Aspergillus, may be the mode of acquisition of nosocomial invasive aspergillosis in susceptible hosts (36). VRE and methicillin-resistant S. aureus (MRSA) have become established as endemic pathogens in many institutions and are increasingly recognized as significant microorganisms in transplant recipients. At many centers, VRE, MRSA, or Clostridium difficile are currently the most frequent etiologic agents of infections in transplant recipients. Although patient specific variables (e.g., severity of illness, intensity of antimicrobial use, and length of hospital stay) are risk factors for acquisition, environmental contamination and, more importantly, person-to-person transmission are also considered significant factors in the nosocomial spread of these bacteria. Equipment and surfaces in the vicinity of patients colonized and infected with VRE have been shown to become contaminated with VRE; VRE could be recovered for at least 7 days from the surfaces of countertops and after 30 minutes from the stethoscopes (37). Furthermore, epidemiologic studies have documented nosocomial VRE transmission by molecular typing techniques (38). Likewise, pulse-field gel electrophoresis demonstrated that 43% of the MRSA isolates causing invasive infections at a transplant unit shared the same pattern, suggesting nosocomial transmission (39). C. difficile is currently the most common cause of infectious diarrhea in transplant recipients. Liver transplantation was identified as the most significant independent risk factor for C. difficile acquisition in one report (40). Although the precise mode of transmission of C. difficile has not been determined, environmental contamination and nosocomial transmission clearly occurs. C. difficile was recovered from 9% to 51% of the environmental cultures; objects contaminated with feces (e.g., bed pan, toilet seats, sinks, and scales were most likely to yield C. difficile) (41). Positive hand cultures were documented in 59% of the hospital personnel caring for the patients with C. Difficile, implicating hands of hospital personnel as a likely mode of transmission (42). Prudent use of antimicrobial agents and measures to curtail nosocomial transmission are key toward effective prevention of infections caused by these pathogens. RISK FACTORS FOR INFECTIONS Surgical factors, intensity of immunosuppression, and variations in local and systemic host response are among the variables that determine not only the type but the site and severity of infections in different types of organ transplant recipients (Table 59.2). TABLE 59.2. RISK FACTORS FOR INFECTION WITH MAJOR NOSOCOMIAL PATHOGENS IN TRANSPLANT RECIPIENTS Pathogen Risk factors Fungi Aspergillus Lung transplantation Single lung transplant, CMV infection Liver transplantation Poor allograft function, renal failure, particularly a requirement for dialysis, OKT3 use, retransplantation Heart transplantation Not determined Pancreatic transplantation Not determined Renal transplantation Augmented immunosuppression and graft failure requiring hemodialysis Candida Liver transplantation Prolonged operation time, high transfusion requirement, high serum creatinine, repeat surgeries Lung transplantation Donor tracheal colonization Pancreatic transplantation Diabetes, exocrine enteric or bladder drainage Pneumocystis carinii Augmented immunosuppression, older recipient age, CMV infection Viruses Cytomegalovirus Donor CMV seropositivity, OKT3 use, allograft rejection, HHV-6 infection Bacteria Vancomycin-resistant Enterococcus faecium Rectal colonization prior to transplant, previous antibiotic use, biliary complications, prolonged hospitalization, surgical reexploration, allograft nonfunction Methicillin-resistant Staphylococcus aureus Nasal S. aureus carriage, prolonged hospitalization, ICU stay Legionella Contaminated hospital potable water system, humidifiers, and ice machines Mycobacterium tuberculosisa (risk factors for early-onset tuberculosis) Nonrenal transplantation, history of prior M. tuberculosis (positive tuberculin test or old active tuberculosis on chest radiographs), OKT3 use Pseudomonas aeruginosa Donor colonization (lung transplants), cystic fibrosis CMV, cytomegalovirus; HHV-6, human herpesvirus-6; ICU, intensive care unit. aEarly-onset tuberculosis implies infection occurring within 12 months of transplantation. Liver Transplantation Liver transplant recipients, by virtue of having hepatic failure and malnutrition before transplantation, represent severely compromised hosts. Many of these patients have concomitant renal failure as a result of hepatorenal syndrome. Renal failure, particularly the requirement for dialysis, was an important predictor of early infections and adversely affected survival after liver transplantation (43, 44) Liver transplant recipients are uniquely susceptible to invasive candidiasis. Most cases originate from endogenous sources; deficient reticuloendothelial function and translocation across the gut mucosa are considered important pathogenetic factors predisposing to invasive candidiasis (45). Vascular and anastomotic P.989 complications are also significant risk factors for infectious morbidity in liver transplant recipients. Duct-to-duct biliary anastomosis compared with Roux-en-Y choledochojejunostomy is associated with a lower incidence of infections, because the latter involves the breach of the bowel integrity and sacrificing the sphincter of Oddi, which may promote reflux of bowel contents into the biliary tree (46). Hepatic artery thrombosis may lead to the development of hepatic infarcts with subsequent gangrene and abscess formation. The clinical presentation is usually acute or fulminant, although hepatic artery occlusion may occasionally be occult and present with a clinical picture of unexplained fever and relapsing subacute bacteremia. Hepatic artery thrombosis may also lead to liver abscesses by compromising the biliary vascular supply. Impaired arterial flow to the hepatic allograft preferentially affects the biliary tree because of the biliary tract's almost total reliance on the hepatic arterial blood supply. Hepatic artery thrombosis may thus lead to biliary tract ischemia and biliary leaks, eventually resulting in intrahepatic abscess formation. The biliary tract may be a source of infection even with an intact vascular supply. Biliary composition is altered during liver transplantation, leading to supersaturation with cholesterol and sludge and stone formation that may predispose to infections (e.g., cholangitis). T-tubes, commonly used to protect duct-to-duct biliary anastomoses, are prone to microbial colonization and form a nidus for the deposition of biliary sludge. Portal vein thrombosis was shown to be the most significant independent predictor of early bacterial infections after liver transplantation (47). Recurrent viral HCV hepatitis has been documented in nearly 50% of the patients undergoing liver transplantation for end-stage liver disease resulting from HCV. HCV is considered an immunosuppressive and an immunomodulatory virus. Patients with HCV recurrence were significantly more likely to develop late-occurring infections, particularly fungal infections after liver transplantation (48). Renal Transplantation Urinary tract and postoperative surgical site infections are two of the most frequent and serious nosocomial infections in renal transplant recipients. Urinary tract infections occur in more than 50% of patients during the first 3 months after transplantation and are the most frequent source of bacteremia during this time. In the absence of antimicrobial prophylaxis, surgical site infections have been reported in up to 20% of patients (49). Organ/space surgical site infections after renal transplantation have been shown to adversely affect graft survival. Surgical site infections in renal transplant recipients are usually due to staphylococci or gram-negative bacilli (49, 50). Staphylococcal infections tended to be associated with incisional surgical site infections and occurred earlier, whereas those due to gram-negative bacilli occurred later; were organ/space surgical site infections; and often led to bacteremia, graft loss, or death. Prolonged urinary catheterization, a surgical site hematoma, a reopened surgical site, and a cadaveric donor graft are risk factors for nosocomial urinary or surgical site infections in renal transplant recipients (51, 52). Renal trauma with nephrectomy and graft contamination during transportation may likely account for a higher risk of infection in cadaveric compared with living allograft recipients. Urinary tract infections occurring in the later posttransplant period (beyond 3 months) are usually benign and rarely symptomatic. Antimicrobial prophylaxis has proven highly effective in reducing the rate of urinary tract and surgical site infections in renal transplant recipients. A single perioperative dose of antibiotics led to a reduction in the incidence of surgical site infections from 25% to 2% (53). Prophylaxis with trimethoprim-sulfamethoxazole has been shown to significantly lower the incidence of urinary tract infections, bacteremias, and infections caused by gram-negative bacilli and S. aureus when compared with placebo (54). Heart and Lung Transplantation Heart and lung transplant recipients are uniquely susceptible to nosocomial bacterial pulmonary infections, particularly in the first month after transplantation. Bacterial pneumonia has been reported in 35% to 48% of the lung and heart-lung transplant recipients (55, 56). Impaired mucociliary clearance, loss of cough reflex, postoperative pain with splinting, and donor tracheal colonization are some factors contributing to a high risk of postoperative pneumonia in lung transplant recipients. Multiply antibiotic-resistant strains of P. aeruginosa and Pseudomonas cepacia are of particular concern in patients undergoing lung transplantation for cystic fibrosis. One-year mortality was shown to be twofold greater in patients who harbored resistant Pseudomonas compared with other patients (57). Indeed, colonization with highly resistant Pseudomonas strains is considered a controversial indication for lung transplantation at many centers. Although the infected lung is removed during transplantation, residual colonization of the airway, nasopharynx, and sinuses remains a potential nidus for subsequent infection. Pretransplant bilateral maxillary sinus drainage followed by monthly irrigation with tobramycin led to an improved outcome in cystic fibrosis patients in one study. Some centers use this approach only if clinically significant sinus infection occurs after transplantation (58). An innovative approach using aerosolized colistin and discontinuation of systemic antibiotics led to the emergence of sensitive Pseudomonas strains in patients who previously harbored resistant Pseudomonas isolates (57). Of 20 cystic fibrosis patients with resistant Pseudomonas who received aerosolized colistin, all became colonized with sensitive isolates of Pseudomonas within a mean of 45 days. In contrast, only 30% (3/10) of the control candidates who received only systemic antipseudomonal antibiotics became colonized with the sensitive isolates. Five of six patients who received colistin and underwent transplantation continued to harbor sensitive microorganisms after transplantation. Circulatory support devices (e.g., intraaortic balloon pump and left ventricular assist devices) are required in many potential heart transplant recipients, and their prolonged placement is a major risk factor for bacterial colonization and subsequent nosocomial infections after transplantation. Sternal surgical site infections occur in 5% to 20% of heart and heart-lung transplant recipients; staphylococci, Enterobacteriaceae, and P. aeruginosa are the most common causative microorganisms. Sternal surgical site infections may directly extend into the mediastinum and P.990 predispose to mediastinitis or mycotic aneurysms at the suture sites. Mediastinitis occurs in 2% to 9% of the heart and heart-lung transplant recipients; S. aureus, P. aeruginosa, and C. albicans have been the most commonly reported microorganisms (59, 60, 61, 62). An unusual cause of mediastinitis in transplant recipients is Mycoplasma hominis (63). Pancreatic Transplantation Surgical site infections, abscesses, or urinary tract infections occur in 7% to 50% of the pancreatic transplant recipients (64, 65, 66, 67). Organ/space surgical site infections are a significant cause of graft loss and mortality in these patients. The postoperative infection rates and the causative pathogens depend primarily on the technique used for the drainage of exocrine secretions of the pancreas. Diversion into the small bowel (enteric drainage), free drainage into the peritoneum, drainage into the bladder, and injection of the duct with synthetic polymer are the approaches used for drainage of exocrine secretions. Infection rates are generally the lowest with duct injection and highest with enteric drainage (which facilitates contamination with gastrointestinal bacteria). Infections occurred in 33% (19/57) of the cases who underwent pancreaticojejunostomy, 33% (5/15) in those with free drainage into the peritoneal cavity, and in 3% (1/39) of the cases in which the duct was injected with a synthetic polymer (66). Duct injection, however, is no longer used because of fibrosis that may lead to loss of endocrine function. Whereas aerobic and anaerobic enteric flora predominate in abscesses associated with enteric drainage, microorganisms in infections in which the viscus has not been opened are usually from the skin flora. Candida, however, is a common pathogen in all types of surgical site infections, including those using bladder drainage. A high incidence of Candida urinary colonization because of diabetes in these patients along with the nonacidic environment in the bladder created by the exocrine pancreatic secretions facilitate Candida colonization. Small Bowel Transplantation Unique features predisposing to infections in small bowel transplant recipients are the fact that the contents of the transplanted organ are nonsterile and that these patients require a higher intensity of immunosuppressive therapy to prevent graft rejection (68, 69, 70, 71, 72). Virtually all small bowel transplant recipients experience at least one episode of infection; the number of infectious episodes per patient may range from 1 to 11 (median 5) (69). Multivisceral transplant recipients and those undergoing colonic segment transplantation with small bowel transplantation are more likely to develop infections (69). It is noteworthy that small bowel transplant recipients remain susceptible to infections, even in the late posttransplant period (i.e., more than 6 months after transplantation) (69). Small bowel transplant recipients, particularly CMV-seronegative recipients of seropositive grafts, are uniquely vulnerable to CMV infection and to recurrent episodes of CMV disease (68). CMV disease in recipient-negative donor-positive patients has been shown to adversely affect outcome in these patients. Consequently, some transplant centers do not use CMVseropositive small bowel grafts for CMV-seronegative recipients (71). Notably, a small bowel graft is involved in 81% to 90% of the patients with CMV disease (68, 71). Bacterial translocation in small bowel transplant recipients predisposes these patients to intraabdominal infections (peritonitis and abscesses). Selective decontamination of the gut after transplantation has been proposed to reduce early postoperative infections in small bowel transplant recipients (70). TIME OF ONSET The relative frequency, types of infection, and the specific pathogens encountered after transplantation generally have a predictable time of onset. Thus, infections in transplant recipients must be evaluated in the context of time elapsed since transplantation. These data also have implications relevant for the institution of prophylaxis and the duration of prophylaxis. Infections During the First 30 Days Most infections occurring within 30 days of transplantation are a consequence primarily of surgical or technical complications related to transplantation, nosocomial acquisition, and rarely reactivation of latent infections (e.g., herpesviruses) in the recipient. Bacterial infections are by far the most frequently occurring infections during this period; vascular catheter-related infections, nosocomial pneumonia, and surgical site infections are the most common types. Fungal infections likely to be encountered in the first month after transplantation include candidiasis and aspergillosis. Nearly 75% of the cases of invasive candidiasis and aspergillosis in liver transplant recipients occur within the first month and virtually all within 2 months of transplantation (45, 73). More recently, however, delayed occurrence of Aspergillus infections has been noted; 55% of the cases of invasive aspergillosis in liver transplant recipients occurred after 90 days of transplantation (74). Liver transplant recipients are uniquely susceptible to invasive candidiasis; disruption of the integrity of the bowel and gastrointestinal translocation are the proposed mechanisms. The only significant viral infection occurring within the first 30 days of transplantation is that due to the HSV. However, there is accumulating evidence to suggest that the novel herpesvirus, HHV-6, may also be a pathogen in the early posttransplant period (75). HHV-6 infection characteristically occurs earlier than CMV and may cause fever of unknown origin and idiopathic cytopenia during this period. Infections Occurring Between 30 and 180 Days Although nosocomial infections may continue to pose a threat in patients requiring prolonged hospitalization, most infections occurring between 30 and 180 days after transplantation are opportunistic infections related to the effects of immunosuppression. The foremost pathogen in transplant recipients during this time period is CMV; however, infections resulting from M. tuberculosis, P. carinii, T. Gondii, and Nocardia are also likely P.991 to be encountered during this interval. Clinically and histopathologically manifest recurrences of HCV hepatitis usually occurs within 6 months of transplantation. In the absence of immunoprophylaxis for HBV, recurrence of HBV infections in the recipient occur a median of 3 months after transplantation. Infections Occurring 6 Months or Later Infectious diseases in the last posttransplant period are typically community-acquired infections similar to those occurring in the general population. However, patients requiring aggressive immunosuppression for recurrent or chronic rejection and those with poorly functioning allografts (e.g., liver transplant recipients with recurrent viral HBV or HCV) continue to be at risk for opportunistic infections. Posttransplant lymphoproliferative disorder, varicella-zoster virus (VZV) infections, cryptococcosis, and infections resulting from dematiaceous fungi typically occur 6 or more months after transplantation. VIRUSES Herpesvirus Infections Cytomegalovirus CMV has been recognized as one of the most significant pathogens in organ transplant recipients. Depending on the pretransplantation CMV serostatus of the recipient, three distinct epidemiologic patterns of CMV infection exist. Primary infection occurs when a seronegative recipient acquires CMV, either from the transplanted allograft or less commonly from blood products. Reactivation infection results from endogenous reactivation of the latent virus. Superinfection implies acquisition of a new strain of CMV in a patient seropositive for CMV before transplantation. Because 50% to 70% of the general population is seropositive for CMV, most infections in transplant recipients represent reactivation infections. However, the clinical impact of CMV is by far greatest in the context of newly acquired or primary infection. Primary CMV acquisition is associated with a higher rate of CMV infection and symptomatic disease, earlier onset of CMV infection posttransplantation, higher incidence of recurrent episodes of CMV, greater risk of dissemination, and higher mortality (76, 77, 78). Symptomatic disease, CMV hepatitis, invasive fungal infections, and death in liver transplantation were more likely to occur when primary infection in the recipient was acquired from the donor organ compared with acquisition from transfusions (31). The time to onset of CMV infection after transplantation is also shorter with donor versus transfusion-associated CMV infection (31). Superinfection, as compared with reactivation infection, is also associated with a higher incidence and severity of symptomatic CMV disease (79). Risk Factors CMV serologic status of the recipient and donor is the most significant factor influencing the rate and severity of CMV infection. Eighty percent to 100% of the seronegative recipients of a seropositive donor allograft acquire CMV infection after transplantation. The risk of CMV infection is lowest (<10%) in seronegative recipients of seronegative organ donors. CMV-seropositive recipients have an intermediate risk (40% to 60%) for developing CMV infection. The intensity and type of immunosuppression are also important determinants of the risk of CMV infection (80). Antilymphocyte preparations (e.g., OKT3) are extremely potent reactivators of CMV. Primary immunosuppressive agents (e.g., cyclosporine and tacrolimus), on the other hand, are not efficient reactivators, but, when CMV reactivation occurs, they interfere with the host's ability to limit viral replication (80). Primary infection with HHV-6, which is considered an immunomodulatory virus, has been proposed to be a risk factor for subsequent CMV invasive disease. Intraoperative hypothermia is a common complication of liver transplant surgery. In a study in liver transplant recipients, intraoperative hypothermia was an independently significant risk factor for early CMV infection and active warming using a convective heating device appeared to curtail this risk (81). Human leukocyte antigen matching and retransplantation have also been shown to be risk factors for CMV infection (77, 82). Pathogenesis CMV-specific major histocompatibility complex (MHC)-restricted cytotoxic T cells are pivotal in host defense against CMV; clinically significant CMV occurs predominantly among patients without an adequate T-lymphocyte response. Humoral immunity, on the other hand, is an ineffective host defense against CMV, although it may modify (or temper) the severity of infection. Tumor necrosis factor-alpha has been shown to be a powerful promoter of CMV (80, 83, 84). Any physiologic stimulus for tumor necrosis factor-alpha release (e.g., OKT3, sepsis, and rejection), therefore, has the potential to activate CMV. CMV is considered an immunosuppressive virus that may facilitate superinfection with opportunistic pathogens (e.g., fungi, gram-negative bacteria, and P. carinii) (77, 85). Other indirect sequelae of CMV infection include a proposed association with allograft rejection in liver transplant recipients, bronchiolitis obliterans in lung transplant recipients, atherogenesis in heart transplant recipients, and glomerulopathy in renal transplant recipients, although this remains controversial. Epidemiology and Clinical Features The overall incidence of CMV infection ranges between 40% and 90% in organ transplant recipients. The highest incidence of CMV infection has been documented in lung or heart-lung transplant recipients (60% to 98%) and the lowest (40% to 50%) in renal transplant recipients. Liver and heart transplant recipients have an intermediate risk of CMV infection (50% to 67%). The frequency of symptomatic disease resulting from CMV ranges from 8% to 15% in renal, 20% to 35% in liver, 27% to 30% in heart, and 55% to 60% in lung transplant recipients. The incidence of CMV infection in small bowel transplant recipients approaches that in lung transplant recipients (68). Small bowel transplant patients also appear to be uniquely susceptible to recurrent episodes of CMV infection (68). Traditionally, most CMV infections have occurred between 4 and 6 weeks. In patients receiving prolonged antiviral prophylaxis, P.992 onset of CMV infection has been noted to be delayed (86, 87, 88). A febrile mononucleosis syndrome characterized by fever, arthralgias, myalgias, leukopenia, and atypical lymphocytosis is the most common symptomatic disease caused by CMV, although localized or disseminated tissue invasive disease may also occur. Predilection to involve the transplanted allograft is a peculiar characteristic of CMV. CMV hepatitis occurs most commonly in liver transplant recipients, CMV pneumonitis occurs most commonly in lung transplant recipients, and CMV enteritis occurs most commonly in small bowel transplant recipients. It is proposed that the transplanted allograft may provide a sequestered site for latently infected cells, because MHC mismatches at these sites may prevent the generation of virus-specific cytotoxic T-cell responses (89). Diagnosis The diagnosis of CMV infection has traditionally been made by viral isolation, which can take up to 4 weeks. The currently available tests not only allow rapid and reliable diagnosis of CMV infection but also may detect viral shedding at an earlier stage. The shell vial assay for viral cultures uses a monoclonal antibody to detect a 72-kDa immediate early CMV antigen and allows detection of CMV within 16 to 24 hours (90). More recently, a CMV antigenemia assay, with monoclonal antibodies directed against the 63-kDa structural late viral protein, has been developed (91, 92). The CMV antigenemia assay is more sensitive and allows earlier detection of CMV than shell vial culture does. Furthermore, results of the antigenemia assay can be quantitated; the number of antigen-positive cells has been shown to correlate with the likelihood of CMV disease and can also be used to monitor response to antiviral therapy. The major drawback of the antigenemia assay is the need for immediate processing of blood samples. Detection of viral DNA by PCR in the leukocyte may be an overly sensitive test for the diagnosis of CMV. Because PCR can detect even minute amounts of viral DNA, it may not differentiate between replicating and latent virus. The precise role of quantitative PCR for the diagnosis of CMV remains to be established. Plasma, as opposed to leukocyte PCR, is specific yet not an excessively sensitive test for the detection of CMV infection (93, 94, 95). Its sensitivity is comparable with CMV antigenemia. Prevention Matching Donors and Recipients by Serologic Status Attempts to decrease the morbidity associated with primary donor-acquired CMV infection have included the use of CMV-seropositive donor organs only for seropositive recipients. Although a decrease in graft loss and mortality attributable to CMV was noted in one report (96), others have not shown a significant impact with such an approach. Widespread adoption of this approach, however, is not feasible given the limited organ donor pool. Prophylaxis Although several prophylactic strategies have been used for CMV, an optimal approach has not been defined. High-dose acyclovir (800 mg orally four times daily) administered to renal transplant patients for 12 weeks decreased the rate of CMV disease (97). Most subsequent studies in solid organ transplant recipients, including renal transplant patients, showed no benefit of acyclovir in the prevention of CMV disease (98). Ganciclovir, administered intravenously for 2 to 4 weeks, was efficacious against CMV, predominantly in the seropositive patients (99, 100). Ganciclovir for 100 days after transplantation was effective both in seropositive patients and in seronegative recipients of a seropositive allograft; however, the feasibility and cost of administration of ganciclovir for 100 days can be prohibitive (101). A valyl ester of ganciclovir, valganciclovir has significantly improved oral bioavailability as compared with ganciclovir administered orally. Systemic levels equivalent to those achievable with intravenous ganciclovir have been documented with valganciclovir. In high-risk organ transplant recipients, valganciclovir 900 mg once daily had comparable efficacy to oral ganciclovir 1,000 mg three times daily for the prevention of CMV disease (88). However, the previously mentioned strategies using universal prophylaxis require administration of ganciclovir to all patients, most of whom will not experience CMV infection. With prolonged ganciclovir use, viral resistance remains a potential concern, and it is increasingly documented in transplant recipients (3, 102, 103). A preferred approach to the prevention of CMV is to target the patients at highest risk for serious disease resulting from CMV. The success of this approach relies on early identification of CMV infection and on the use of prophylaxis preemptively to prevent asymptomatic infection from progressing to CMV disease. The criterion used to identify high-risk patients may either be a laboratory marker (e.g., viral shedding preceding symptomatic disease) or a clinical event (e.g., institution of OKT3 therapy). Surveillance cultures using a sensitive assay (e.g., CMV antigenemia) offer a significant advantage over the previous culture methods (e.g., shell vial assay) for earlier detection of CMV infection (98, 104, 105). The preemptive strategy is particularly appealing and perhaps ideal for seropositive patients in whom CMV disease generally is less severe (98). With diligent monitoring, CMV disease can also be prevented in 80% to 100% of the patients with primary CMV infection (106). Herpes Simplex Virus HSV infections in transplant recipients present as mucocutaneous lesions resulting from reactivation of the latent virus. However, visceral or disseminated HSV infection can be donor transmitted and may have a fulminant presentation with a grave outcome without antiviral therapy. HSV hepatitis is the most frequently documented site of disseminated HSV infection; its incidence (cases per thousand) is reported to be 2.11 in renal, 2.23 in heart, and 4.81 in liver transplant recipients. In a report comprising 12 cases of HSV hepatitis in solid organ transplant recipients, 33% were due to primary HSV infection believed to be acquired from the donor (107). The median time to onset of HSV hepatitis was 18 days, although it occurred as early as 5 days posttransplantation (107). This characteristic time of onset is in contrast with CMV hepatitis, which usually occurs 30 to 40 days after transplantation. Clinical manifestations of HSV hepatitis include fever, leukocytosis, thrombocytopenia, and marked elevation of hepatocellular liver enzymes. Mortality P.993 from primary visceral HSV infection in seronegative recipients was 75%; hypotension, disseminated intravascular coagulation, metabolic acidosis, low platelet count, and high creatinine were significant predictors of mortality (107). The HSV accounted for 41% of all non-CMV isolates from the respiratory tract in lung transplant recipients; 80% of the isolates were deemed clinically significant and were associated with pneumonitis (108). Another clinical presentation of HSV, predominantly reported in intubated lung and cardiac transplant recipients, is HSV tracheobronchitis that manifested as fever, bronchospasm, leukocytosis, and difficulty weaning. Paradoxically, HSV tracheobronchitis had a more severe presentation and worse outcome in immunocompetent compared with immunosuppressed patients (109). It was proposed that this may be due to a more exuberant local immune response in the immunocompetent patients (109). Low-dose acyclovir (200 to 400 mg orally three times daily) generally used for 3 to 4 weeks posttransplant is highly effective as prophylaxis for HSV in transplant recipients. At one institution, HSV hepatitis was documented in 12 of 3,536 solid organ transplant recipients before the routine use of acyclovir prophylaxis and in none of the 1,144 patients since the use of acyclovir prophylaxis (107). Longer duration (up to 3 months) of acyclovir is used for lung transplant recipients in some centers (108). Varicella-Zoster Virus Up to 70% of the pediatric and 5% of the adult transplant recipients have been reported to be seronegative for VZV (110, 111). Exposure to VZV infection may result in primary varicella in these susceptible patients. Median time to onset of varicella was 2 years after transplantation in one report (110) and 2.4 years in another (112). Visceral dissemination, frequently documented in transplant recipients, is the primary cause of mortality in patients with VZV. Hepatitis, pneumonitis, pancreatitis, gastroenteritis, or meningoencephalitis are the most commonly documented sites of visceral dissemination. Varicella may initially present with acute abdominal pain, and in the absence of skin lesions can defy early recognition. It is notable that up to 16% to 18% of the pediatric transplant recipients may have recurrent varicella infections (112, 113). Varicella-zoster immunoglobulin (VZIG) is recommended for susceptible transplant recipients exposed to varicella. VZIG, however, is not entirely protective; in up to one third of the patients with varicella, lesions have occurred despite VZIG prophylaxis. Some centers use high-dose oral acyclovir for the duration of the incubation period of varicella (i.e., 2 to 3 weeks after exposure of susceptible patients to varicella) (102). Varicella vaccine has been found to be safe and effective in transplant recipients. In 704 pediatric renal transplant recipients who received varicella immunization, 62% had VZV antibodies at 1 year and 42% after 10 years (114). The incidence and severity of varicella was significantly less in the immunized patients (114). Human Herpesvirus-6 The newest herpesvirus to be considered a pathogen in transplant recipients is HHV-6. HHV-6 is a large double-stranded DNA virus that is antigenically distinct from other human herpesviruses. Its closest phylogenetic relative is CMV; nucleotide sequencing has revealed 66% DNA homology between CMV and HHV-6. On the basis of genomic DNA sequences, cell tropism, and protein expression, two distinct variants of HHV-6, designated as variant A and variant B, have been described (115, 116). The two variants differ in virulence; the HHV-6A variant is intrinsically more virulent and neurotropic than the HHV-6B variant (117). Most infections in transplant recipients are due to the HHV-6B variant (117). Most HHV-6 infections are believe to result from endogenous reactivation of the recipient's latent virus; however, donor transmission has also been documented. Although the precise incidence and clinical sequelae of HHV-6 infection after transplantation remain to be fully elucidated, HHV-6 infection has been reported in 31% to 55% of solid transplant recipients (117). The usual timing of onset is between 2 and 4 weeks posttransplantation. Bone marrow suppression, interstitial pneumonia, encephalopathy, and fever of unknown origin are the most commonly reported clinical manifestations of HHV-6 (117, 118, 118, 119, 120). Viral culture remains the gold standard for the diagnosis of HHV-6 infection. A shell vial culture assay (analogous to that for CMV), based on the detection of immediate early antigen of HHV-6 in cell culture, can significantly expedite the time required for detection of HHV-6 compared with the conventional cell culture (117). The role of PCR and antibody titer rises to diagnose active HHV-6 infection has not been fully ascertained. The antiviral susceptibilities of HHV-6 resemble those of CMV (121, 122, 123). HHV-6 is sensitive to both ganciclovir and foscarnet and resistant to acyclovir at achievable serum concentrations. The role of prophylaxis for HHV-6 has not yet been fully discerned. Hepatotropic Viruses Hepatitis C Virus End-stage liver disease resulting from HCV has been documented in up to 50% of the patients undergoing liver transplantation in recent years. Up to 95% of the patients with pretransplant HCV infection remain viremic after transplantation, as demonstrated by the presence of HCV RNA in the blood. Clinically and histologically manifest recurrence occurs in 30% to 70% of patients with pretransplant HCV, generally within 1 to 12 months after transplantation (124, 125, 126). Progression to cirrhosis has been observed in 15% to 20% of patients 1 to 3 years after initial transplantation. Despite significant morbidity associated with HCV, survival rates in patients with and without HCV recurrence have not been different (126). The prevalence of HCV positivity in hemodialysis patients ranges between 5% and 54% (127). The number of blood units transfused, the duration of hemodialysis, and the type of dialysis (hemodialysis as opposed to peritoneal dialysis) correlated with a higher incidence of HCV infection in renal transplant candidates (127, 128, 129). After transplantation, chronic liver disease has been reported in 10% to 60% of renal transplant recipients and occurred significantly more frequently in patients with HCV compared with those without HCV (128, 129, 130). Pretransplant HCV P.994 infection, however, does not seem to adversely affect the graft or patient survival after renal transplantation (130, 131). Although most posttransplant HCV infections are due to recurrence of pretransplant HCV, de novo infections resulting from acquisition from the donor organ or transfused blood products have been reported. A 35% rate of acquired HCV-RNA infection was reported in 89 liver transplant recipients, most of whom were transplanted before the routine screening of blood products for anti-HCV (124). Routine screening of blood products has led to a significantly lower acquisition rate of HCV (i.e., 2.5% to 4%) (132). HCV has also been transmitted by organs from anti-HCV–positive donors (see the section on sources of nosocomial infections). HCV infection is also considered a significant risk to the personnel involved in the care of HCV-infected transplant patients. The seroprevalence of HCV (7%) in healthcare workers directly involved in the care of liver transplant patients was significantly higher when compared with those not associated with liver transplantation (0.5%) or in volunteer blood donors (0.3%) (133). None of the transplant personnel had hepatitis or a history of transfusion (133). The risk of acquiring HCV after needle stick injuries may be as high as 10%. Serum immunoglobulin is not protective against HCV and is not recommended (also see Chapter 78). A number of variables are believed to influence the rate and severity of recurrent HCV hepatitis after transplantation, including the level of pretransplant viremia, genotype of the virus, and the intensity of posttransplant immunosuppression (126, 134, 135). HCV genotype 1b has been associated with more severe recurrent HCV infection after liver transplantation (126). Corticosteroids have been shown to result in a severalfold increase in the HCV-RNA level. Finally, allograft rejection and steroid-resistant rejection requiring OKT3 can lead to a higher incidence and earlier onset of recurrent HCV hepatitis after liver transplantation (135, 136). Effective prophylaxis against HCV in transplant recipients is not available. It has been shown that polyclonal immunoglobulin preparations against HBsAg administered to liver transplant patients for HBV were also protective against HCV (137). The incidence of HCV viremia in the patients receiving HBIG was significantly lower than in those who did not receive HBIG (54% vs. 94%, p = .001). This protective effect may have been due to the presence of anti-HCV in HBIG (137). Prophylaxis with interferon-alpha, administered for 6 months posttransplantation in liver transplant recipients, delayed the occurrence of HCV but decreased neither the incidence nor the severity of recurrent HCV hepatitis (138). Hepatitis B Virus The clinical impact of HBV is of greatest importance in the context of liver and renal transplantation. In studies in which long-term immunoprophylaxis was not used, the reinfection rate of the hepatic allograft with HBV was virtually 100% with progression to liver failure and death in as little as 2 to 2.5 years. Anti-HBs immunoprophylaxis has significantly altered the natural history of HBV after liver transplantation. In a large study assessing the outcome in HBsAg-positive patients undergoing liver transplantation, the overall risk of HBV recurrence after 3 years was 67% (139). HBV recurrence was documented in 78% to 90% of patients a mean of 3 months posttransplant in patients who did not receive long-term immunoprophylaxis; recurrence was significantly less (56%) and delayed, occurring a mean of 8 months posttransplant, in patients receiving long-term anti-HBs immunoprophylaxis (139). HBV adversely affects graft and patient survival; survival at 3 years was 54% in patients with HBV recurrence and 83% in those who remained HBsAg negative after transplantation (139). Several factors influence the recurrence of HBV after transplantation. The risk of recurrence is greater for patients with markers for active replication of HBV before transplantation (e.g., those seropositive for HBeAg or HBV DNA). The risk of HBV recurrence was 83% В6% in liver transplant recipients seropositive for HBV DNA compared with 58% В7% in those with neither HBV DNA nor HBeAg at the time of transplantation (139). Fulminant HBV (as opposed to chronic HBV infection) is associated with a lower rate of recurrence. Recurrence was observed in 17% of patients with fulminant HBV compared with 67% in those with chronic HBV cirrhosis (139). Patients with fulminant HBV tend to have lower levels of HBV DNA or replicative HBV. Co-infection with hepatitis delta virus decreases the risk of recurrence in HBV infection after transplantation (139). Delta virus is a naturally occurring inhibitor of HBV replication, and hence HBV DNA levels in delta virus co-infected patients are lower. Strains of HBV that fail to produce HBeAg because of mutations in the precore region of the HBV genome (also known as precore mutants or HBeAg-deficient mutants) have recently been identified (140). Such patients have high levels of viral DNA in the absence of HBeAg. Patients infected with the precore mutants pretransplant, as opposed to the wild-type virus, have a greater risk of hepatic graft loss resulting from early recurrence (140). A unique and particularly aggressive syndrome of recurrent HBV infection observed in 12% to 20% of patients with HBV recurrence is fibrosing cholestatic hepatitis, characterized by marked cholestasis and hypoprothrombinemia but only modest increases in serum transaminases (141). Fibrosing cholestatic hepatitis is more likely to occur in patients with pretransplant HBV replication and results in rapid death in almost all cases. A paucity of inflammatory response in this syndrome suggests that the virus may be directly cytopathic. HBV infection also follows an aggressive clinical course after renal transplantation. Progression of liver disease to cirrhosis and death, however, occurs considerably later than in liver transplantation (i.e., 6 to 8 years after transplantation). Chronic active or persistent hepatitis occurred in 76% of HBsAg-positive patients undergoing renal transplantation compared with 31% in HBsAg-negative patients (142). The most effective approach to prevent recurrent HBV is the use of combination therapy with lamivudine and HBIG. Prophylaxis with the use of lamivudine monotherapy is not recommended because of the reappearance of HBsAg after liver transplantation in 32% to 50% of the patients. Combination therapy with HBIG and lamivudine prevents HBV recurrence in more than 90% of the patients undergoing liver transplantation for HBV (143, 144). P.995 Adefovir dipivoxil and entecavir have been shown to suppress the replication of lamivudine-resistant HBV. In a compassionate-use protocol in 40 patients with lamivudine-resistant HBV awaiting liver transplantation and 127 liver transplant recipients with recurrent HBV resulting from lamivudine-resistant HBV, a two to three log reduction in HBV DNA levels and biochemical improvement was noted (145, 146). Hepatitis G Virus Hepatitis G virus is a newly identified RNA virus belonging to the family Flaviviridae with an estimated prevalence of 1% to 2% in the volunteer blood donor population in the United States. Sequence analysis suggests that hepatitis G virus and hepatitis GB virus represent independent isolates of the same virus. Hepatitis G virus also bears 26% sequence homology with HCV. Among patients undergoing liver transplantation for end-stage liver disease resulting from HCV, up to 25% may be co-infected with hepatitis G virus. However, hepatitis G virus appears to have minimal clinical impact and influences neither the graft nor patient outcome in liver or renal transplant recipients (147, 148, 149). Up to 30% of the patients undergoing liver transplantation for cryptogenic cirrhosis have demonstrated histologic evidence of hepatitis in the absence of any known viruses, including hepatitis G virus (150). It is suspected that as yet unidentified viruses may be the cause of posttransplant hepatitis occurring in the absence of hepatitis A to G viruses (150). Other Viruses BK Virus Within the last decade, BK virus (BKV) has emerged as a significant pathogen in renal transplant recipients. BKV is a polyomavirus that is acquired during childhood. Renal and uroepithelial cells are the main site of latency. Seroprevalence rates in the general population range from 70% to 90%. Nephropathy resulting from BKV has been reported in 1% to 5% of the renal transplant recipients with allograft loss occurring in nearly half of those patients (151). This entity was encountered only rarely before the mid-1990s. Although precise reasons for the recent emergence of BKV as a significant pathogen is unclear, use of novel, more potent immunosuppressive agents (e.g., tacrolimus, mycophenolate mofetil, and sirolimus) is considered to play a role. Donor transmission has been reported; however, most cases of BKV nephropathy occur as a result of reactivation of the latent virus. The usual time to onset of BKV nephropathy is 28 to 40 weeks posttransplantation. Typical manifestations include a modest rise in creatinine that fails to respond to antirejection therapy. The hallmark of BKV replication are decoy cells, which are urinary epithelial cells bearing ground-glass intranuclear inclusions. Decoy cells, however, lack specificity for the diagnosis. Quantitative BKV viremia, however, has been shown to correlate with BKV nephropathy (152) and may also be used to monitor response to therapy. Specific antiviral therapy for BKV is not available currently. Judicious reduction of immunosuppression has been attempted as a management strategy but may not always be successful. The role of cidofovir for the treatment of BKV nephropathy or that of retransplantation has not been fully defined. Adenovirus Adenoviral infections have been documented in up to 10% of the pediatric and 1% to 15% of adult transplant recipients (153, 154, 155, 156). Symptomatic disease is more common and generally more severe in pediatric compared with adult patients after transplantation; 60% of the children and 27% of the adult transplant recipients with adenoviral shedding have been shown to have disease resulting from adenovirus (153, 154, 155). The precise mode of transmission of adenoviral infections has not been determined, although both donor transmission and nosocomial transmission have been proposed to occur (153, 154, 155). In pediatric liver transplant recipients, most severe disease occurred in seronegative children (154), and donor serology was positive in five of six patients evaluated, suggesting that donor transmission is a likely source of infection. Nosocomial acquisition is also a consideration, because several patients with the similar adenovirus strains were found temporally clustered in one report (153). Hepatitis and pneumonitis are the most common invasive forms of adenoviral disease. Hepatitis in liver transplant recipients, pneumonitis in lung transplant recipients, and hemorrhagic cystitis in renal transplant recipients are the most frequently involved sites of disease. Serotypes 5 and 11 were the most frequent serotypes causing hepatitis and hemorrhagic cystitis, respectively, in transplant recipients. Diagnosis is suggested by the detection of microabscesses with smudgy intranuclear targeted inclusions in histopathologic specimens (157). Immunohistochemistry and culture can be used to confirm adenoviral infection. Although ribavirin has been anecdotally used as therapy (158), an effective prophylaxis for adenoviral infections is currently not available. Respiratory Viral Infections The impact of respiratory viral infections [e.g., respiratory syncytial virus (RSV), influenza, and parainfluenza viruses] has not been fully characterized in solid organ transplant recipients. Over a 6-year period, 3.4% of pediatric liver transplant recipients were documented to have RSV infections (159). The median time to diagnosis was 24 days posttransplant. Seventy-six percent of the infections were nosocomially acquired. Early-onset infection and preexisting lung disease portended a more severe disease. Late infections, on the other hand, occurring in the absence of rejection, were usually without untoward sequelae. Of 19 cases of paramyxovirus infection in lung transplant recipients, 9 were due to RSV and 10 due to parainfluenza virus (160). Three cases of parainfluenza type 3 infections in that report occurred within a 3-week period; two of these patients had contact with each other and with hospital personnel during a 1-month period of the infection (160). Influenza can be a serious viral infection in pediatric organ transplant P.996 recipients. Of 12 such cases, 5 were exposed to influenza B while hospitalized (161). Annual influenza immunization of pediatric organ transplant recipients, their household contacts, and healthcare workers is recommended (161). BACTERIAL INFECTIONS Staphylococci Staphylococci, particularly S. aureus, are increasingly recognized as pathogens in transplant recipients and have emerged as the leading cause of bacterial infections in liver, heart, kidney, and pancreatic transplant recipients at many centers (162, 163, 164, 165). This increase largely parallels the more widespread rise in gram-positive infections in the nosocomial setting in recent years. Forty-nine percent of the bacteremias in liver transplant recipients in one report were due to S. aureus (163). Although intravascular cannulas, accounting for 54% of all MRSA bacteremias, were the predominant source, wound infections, nosocomial pneumonia, intraabdominal abscess, and peritonitis were also documented as sources of MRSA bacteremia (163). Over one half of the S. aureus infections occur in the intensive care unit setting (163). Requirement of invasive procedures, mechanical ventilation, continuous need for intravenous access, and overall debilitated condition of the patients in the intensive care unit provide conditions conducive to the development of nosocomial S. aureus infections. S. aureus infections generally occur very early after transplantation. In a study in liver transplant recipients, nearly one third of such infections occurred within the first week of transplantation; the median time to onset was 16 days (39). S. aureus is also the most frequent cause of endocarditis in organ transplant recipients (166). Notably, 74% of the cases of endocarditis were associated with previous hospital-acquired infection, especially venous access device and wound infections (166). S. aureus colonization of the anterior nares has recently been shown to be a significant predictor of infections resulting from S. aureus in liver transplant patients (39). Overall, nasal carriage was documented in 67% of the patients; infected patients were significantly more likely to be nasal carriers of S. aureus compared with the noninfected patients. Pulse-field gel electrophoresis documented that the isolates causing infections matched the isolates from the anterior nares in all cases (39). Furthermore, 43% of infected patients shared the isolates with the same restriction pattern, indicating cross-transmission in the nosocomial setting (39). Eradication of nasal carriage by mupirocin, however, has not been shown to prevent S. aureus infections in liver transplant recipients (167). Although 87% of the colonized patients were successfully decolonized, recolonization occurred in 37% (167). Nosocomial transmission leading to exogenous colonization and colonization at non-nasal sites that may be unaffected by nasal administration of mupirocin likely accounted for the failure of mupirocin to decrease S. aureus infections (167). Enterococci Enterococci, which are normal inhabitants of the gastrointestinal tract, are of greatest relevance in liver transplant recipients. Most enterococcal bacteremias in these patients result from complications related to the biliary tree. Roux-en-Y choledochojejunostomy (which facilitates reflux of enteric bacteria into the biliary tree) and biliary strictures have been shown to be independent risk factors for enterococcal bacteremia after liver transplantation (168). Vancomycin-resistant Enterococcus faecium (VREF) have emerged as nosocomial pathogens of grave concern, particularly after liver transplantation. VREF infections were documented in 10.5% of the liver transplant recipients in one recent study and 16% in another (169, 170). Notably, VREF was the most frequently isolated pathogen in infected liver transplant recipients in the latter study. Infections were documented a median of 39 and 42 days after transplantation in two studies (169, 170) but considerably earlier (a median of 10 days) in another report (171). Intraabdominal infections were the most frequent site of infection resulting from VREF. VREF fecal carriage before transplantation; previous antibiotic use (including vancomycin); biliary complications; prolonged hospitalization and intensive care unit stay; surgical reexploration; surgical complications during transplantation, including hypotension; and primary nonfunction of the allograft have been identified as significant risk factors for VREF infections (169, 170, 172). Mortality in the infected patients range between 23% and 50%. Intensive care unit stay before transplantation, hemodialysis, liver failure, and shock have been shown to be independent predictors of mortality in patients with VREF infections (170, 172). Not surprisingly, antibiotics have not been effective in reducing mortality; outcome did not differ significantly among patients who received the drugs with activity against the isolate in vitro compared with those who received no therapy (169). VREF colonization once established is often a persistent event; spontaneous conversion to VREF-negative carriage is uncommon. These patients, therefore, remain at risk for invasive infections and a threat for nosocomial transmission. A variety of gut decontamination regimens, including oral bacitracin (173), have been tried; however, none have been shown to be consistently effective. Consequently, infection control practices to prevent nosocomial acquisition and cross-transmission and judicious use of antimicrobial agents, particularly vancomycin, are critically important in curtailing VREF infections. Mycobacterium tuberculosis The incidence of tuberculosis in solid organ transplant patients ranges from 0.35% to 5% in the United States and Europe (24, 174). However, in highly endemic areas (e.g., India and Pakistan), tuberculosis may develop in 5% to 15% of the transplant patients. The median time to onset after transplantation is 9 months and ranges from 0.5 to 144 months (24, 174). Tuberculosis occurs significantly later after transplantation in renal compared with nonrenal transplant recipients. Disseminated disease occurs in nearly one third of the transplant recipients with tuberculosis (24). The gastrointestinal tract is the most frequent extrapulmonary site of tuberculosis in transplant recipients. Other reported extrapulmonary sites of involvement include the skin and osteoarticular tissue, central nervous system, kidneys, and urogenital tract. Tuberculin reactivity has been documented in 20% of the transplant recipients with tuberculosis, and chest radiographs P.997 with evidence of old active tuberculosis were documented in 12% of the patients (24). These patients are more likely to develop tuberculosis earlier after transplantation than those without a history of tuberculin reactivity or abnormal chest radiograph before transplantation. Most tuberculosis infections in transplant recipients represent reactivation of old dormant disease. However, nosocomial acquisition or donor transmission are also well-documented modes of transmission. Tuberculosis has been shown to be transmitted both by living and cadaveric organ donors. Tuberculosis, involving the renal allograft, was documented 35 and 39 days after renal transplantation in two recipients of the same donor who died of hypoglycorrhachic lymphocytic meningitis of unknown etiology; the donor's cerebrospinal fluid culture was positive for M. tuberculosis 3 weeks after death (23). One renal allograft recipient of this donor died of disseminated tuberculosis, whereas the second recovered, although rejection secondary to antituberculosis therapy necessitated allograft nephrectomy (23). Two recipients of a single lung transplant from a common donor had the same M. tuberculosis isolate as demonstrated by restriction fragment length polymorphism (175). Tuberculosis involving a hepatic allograft was documented in a pediatric transplant recipient who received a living related lateral segment hepatic allograft from the mother (166). Pulmonary tuberculosis was detected concomitantly in the mother who was apparently asymptomatic at the time of donation of the hepatic segmental graft (176). A nosocomial outbreak involving ten renal transplant patients was documented from one institution; eight of these cases were clustered within a 5-month period (7). The source case was a renal transplant recipient who was exposed to tuberculosis at another hospital. Tuberculosis was not suspected in the source case on admission, thus delaying the isolation precautions. Restriction fragment length polymorphism documented transmission of M. tuberculosis from the index case to five renal transplant recipients. The median incubation period for tuberculosis in this outbreak was only 7.5 weeks, and death occurred in five of ten patients a median of 8 weeks after diagnosis (7). It is noteworthy that the exposed transplant recipients were more likely to contract tuberculosis compared with the nontransplant contacts of the source case (7). Overall mortality in organ transplant recipients with tuberculosis is approximately 30% (24). Disseminated compared with localized tuberculosis, prior rejection, and OKT3 receipt were significant predictors of mortality in transplant recipients with tuberculosis. All transplant recipients should have a tuberculin skin test administered before transplantation. Isoniazid prophylaxis should be considered for the transplant recipients with the characteristics outlined in Table 59.3 regardless of the tuberculin skin test reactivity. A recent study has documented that isoniazid chemoprophylaxis initiated during liver transplant candidacy was safe and effective (177). Such an approach minimizes the exposure to the new allograft to isoniazid and may mitigate potential drug interactions of isoniazid with the immunosuppressive agents. Tuberculin skin test reactivity per se is a controversial indication for prophylaxis in transplant recipients. The rate of tuberculosis among skin-test positive liver transplant candidates and recipients who receive no chemoprophylaxis has been estimated to be 1,585.3 cases per 1,000,000 person-years (178). Tuberculosis has been documented in up to 2% of the tuberculin skin-test–negative liver transplant recipients (179). It has been proposed that clinical or radiographic evidence of previous tuberculosis may more reliably identify high-risk patients as compared with the tuberculin skin test result. Optimal management of tuberculin skin test positive or anergic patients, however, remains to be determined. TABLE 59.3. INDICATIONS FOR CHEMOPROPHYLAXIS WITH ISONIAZID IN ORGAN TRANSPLANT RECIPIENTS Tuberculin skin reactivity ≥5 mm before transplantation Patients with the following characteristics, regardless of tuberculin skin test reactivity: Radiographic evidence of old active tuberculosis and no prior prophylaxis Prior history of inadequately treated tuberculosis Close contact with an infectious case Receipt of an allograft from a donor with a history of tuberculosis or tuberculin reactivity Newly infected persons (recent tuberculin skin test converters) Legionella Legionellosis has been reported in 2% to 9% of solid organ transplant recipients with pneumonia; however, at certain institutions, 25% to 38% of the bacterial pneumonias have been due to Legionella (5). Legionella pneumophila and Legionella micdadei are the most common species implicated; however, Legionella bozemanii, Legionella birminghamensis, Legionella dumoffii, and Legionella cincinnatiensis have also caused infections in transplant recipients. Inhalation of aerosols containing Legionella has been proposed as the mode of transmission for this microorganism. However, aspiration is considered the most likely mode of transmission and Legionella-contaminated potable water distribution systems as the predominant source of legionellosis (5). Molecular fingerprinting methods have linked L. pneumophila infection in transplant recipients to hospital drinking water (180). Ice machines (34) and ultrasonic humidifiers (181) have also been shown to be the sources of Legionella infection after transplantation. Pneumonia is the predominant clinical manifestation of legionellosis, although pericarditis, necrotizing cellulitis, peritonitis, hepatic allograft infection, and hemodialysis fistula infections have also been reported after transplantation (5). Nodular pulmonary densities and cavitation (reported in 50% to 70% of the pulmonary infections in some reports) are characteristic radiographic features but may not be invariably present. Legionella are fastidious microorganisms that do not grow on standard bacteriologic media. Selective media containing dyes and antimicrobial agents are needed for optimal growth. Urinary antigen is both sensitive and specific for the detection of Legionella and may also be diagnostically useful for detecting Legionella in body fluids (e.g., pleural fluids). It is recommended that hospitals performing large numbers of transplants should routinely culture the hospital water supply for Legionella, perhaps once a year (182). If such cultures are positive, specialized Legionella laboratory tests, especially culture P.998 on selective media and urinary antigen tests, should be made routinely available in the clinical microbiology laboratory. Two disinfection methods for the water supply have emerged as cost effective: superheating the water to 70ВC and flushing the distal outlets or the installation of copper-silver ionization units. Hyperchlorination is no longer recommended because of the expense, erratic efficacy, corrosive damage to the piping, and the carcinogenic potential of ingested chlorine. Electric showers have been used when hyperchlorination and superheating and flushing proved inadequate. Unfortunately, this method is unlikely to be effective as a long-term mode of prevention because showering is not the primary mode of transmission. Nocardiosis Infections resulting from Nocardia species may occur in 2% to 4% of organ transplant recipients; the median time to the onset of nocardiosis after transplantation ranges between 2 and 8 months. Central nervous system involvement occurs in 17% to 38% of these patients. Brain abscesses are usually multiple; meningitis is rare and usually associated with an abscess. An important clue to central nervous system nocardiosis is concomitant skin or subcutaneous lesions from which Nocardia species can be readily isolated. Nocardia is a soil microorganism whose primary portal of entry is the lung. There is evidence to suggest that nosocomial transmission of nocardiosis may occur (183, 184, 185, 186). Cases of Nocardia infection clustered in time have been reported in renal transplant units. An epidemic strain of Nocardia common to the infected patients and environmental dust samples from the unit housing the patients but distinct from environmental isolates elsewhere in the hospital was documented to cause seven infections in a renal transplant and dialysis unit (184). Respiratory isolation of the cases of nocardiosis during outbreaks has been recommended by some (184, 185). Trimethoprim-sulfamethoxazole used as prophylaxis for PCP is also effective against nocardiosis. FUNGAL INFECTIONS Aspergillus Invasive aspergillosis remains a devastating fungal infection in all types of transplant recipients. It has, however, unique clinical characteristics and risk factors in different types of solid organ transplant recipients. Epidemiology Lung Transplantation Lung transplant recipients are more likely than other solid organ recipients to develop infection with Aspergillus (Table 59.4). Up to 8% of lung transplant recipients develop invasive aspergillosis with an additional 10% demonstrating Aspergillus colonization (187, 188, 189, 190). Risk factors for Aspergillus infection after lung transplantation include CMV infection and single lung transplantation (189, 191). Cystic fibrosis is not a risk factor for Aspergillus infection posttransplant; isolation of Aspergillus species from respiratory secretions of patients with cystic fibrosis pretransplant has not been shown to predict subsequent development of invasive disease (190, 192). Most cases of invasive aspergillosis in lung transplant recipients occur within the first 9 months posttransplantation. A unique form of invasive aspergillosis occurring in lung transplant recipients is ulcerative tracheobronchitis. TABLE 59.4. INCIDENCE OF INVASIVE ASPERGILLOSIS AND MORTALITY IN DIFFERENT TYPES OF SOLIDORGAN TRANSPLANT RECIPIENTS Type of transplant Cumulative incidence of invasive disease (%) Cumulative incidence of colonization (%) Crude mortality in patients with invasive aspergillosis (%) Kidney 0.7 1.7 75 Pancreas 1.3 NA 100 Liver 1.7 0.5 87 Small bowel 3.5 NA 100 Heart 6.2 NA 78 Lung 8.4 10.4 74 NA, data not available. Liver Transplantation The incidence of invasive aspergillosis in liver transplant recipients ranges from 1% to 4% (193, 194, 195, 196, 197). The infection is most often diagnosed between 2 and 4 weeks after transplantation. A poorly functioning hepatic allograft and renal insufficiency, particularly the requirement for hemodialysis, are considered important risk factors for invasive aspergillosis in liver transplant recipients. Although OKT3 use was shown in early studies to be a risk factor for invasive aspergillosis, a recent large study showed that only 8% of liver transplant recipients with invasive aspergillosis had received OKT3 (195). Approximately 25% of the cases of invasive aspergillosis in liver transplant recipients occur after retransplantation (195). Rarely, the Aspergillus infection is confined to the surgical site (producing necrotizing fasciitis) or intraabdominal sites in liver transplant recipients. Heart Transplantation Invasive aspergillosis occurs in 1% to 6% of heart transplant recipients (198, 199, 200, 201). The median time to development of invasive aspergillosis in these patients is 1 to 2 months. Most infections originate in the lungs, and 20% to 35% disseminate to other organs. Renal Transplantation Invasive aspergillosis has been reported in 0.7% to 1% of the patients undergoing renal transplantation (198, 202, 203). Cases of invasive aspergillosis in renal transplant recipients have usually been pulmonary infections and occasionally disseminated disease. Augmented immunosuppression and graft failure requiring hemodialysis are risk factors for invasive aspergillosis in renal transplant recipients (203). P.999 Diagnosis Early diagnosis is critically important in reducing the mortality from invasive aspergillosis. Aspergillus can be cultured from sputum in only 8% to 34% and from bronchoalveolar lavage fluid in 45% to 62% of patients with invasive pulmonary aspergillosis (204). Respiratory cultures, therefore, may not detect aspergillosis before significant vascular invasion has occurred. Surveillance serologic tests, however, may be potentially more useful. ELISA has been shown to have a sensitivity of 50% to 90% and specificity of 81% to 93% for the diagnosis of invasive aspergillosis (204, 205, 206). Furthermore, the antigen detection tests may be positive as long as 28 days before clinical and radiographic signs of invasive aspergillosis become apparent (205). The efficacy of Aspergillus antigen detection tests, however, has not been extensively evaluated in solid organ transplant recipients. High-resolution thoracic computed tomography (CT) may be able to raise the index of suspicion for invasive pulmonary aspergillosis soon after the development of symptoms and before culture results are available. Such imaging in neutropenic patients whose fever persisted for more than 2 days despite empiric antibiotic treatment showed findings highly suggestive of invasive pulmonary aspergillosis 5 days earlier than the use of chest roentgenograms (207). Prevention and Prophylaxis Outbreaks of aspergillosis have been linked to construction activity within or near a transplant unit and to contaminated or poorly maintained ventilating ducts, grids, and air filters. Outbreaks associated with construction activity in bone marrow transplant recipients have been curtailed by use of laminar air flow units with high-efficiency particulate air filtration. Effective prophylaxis against invasive aspergillosis is currently not available. Although itraconazole is highly active in vitro against Aspergillus, its efficacy has not yet been demonstrated in clinical trials. Furthermore, the absorption of itraconazole in capsule form can be erratic in transplant recipients. Itraconazole solubilized in cyclodextrin is better absorbed; its efficacy as prophylaxis, however, remains unproven. Low-dose intravenous amphotericin B (0.1 mg/kg/day) or liposomal intravenous amphotericin B (1 mg/kg/day) have also not been shown to be consistently efficacious. On the contrary, it has been proposed that low-dose amphotericin B may promote the emergence of Aspergillus infection (208). Aerosolized amphotericin B prophylaxis administered during the posttransplant hospital stay was shown to reduce the incidence of fungal infections, including aspergillosis in lung, heart-lung, and heart transplant recipients in one report (209). Others have not found it to be effective. Furthermore, the formulation, optimal dosage, and precise mode of delivery of aerosolized amphotericin B preparations has not been determined. Given the lack of availability of effective prophylaxis, preemptive therapy or empiric therapy in selected high-risk patients may be considered as an alternative to prophylaxis. Candidiasis With the exception of heart transplant recipients, invasive candidiasis is the most frequently occurring fungal infection in solid organ transplant recipients (45). The incidence of Candida infections is highest in liver transplant recipients. Virtually all Candida infections are nosocomially acquired, although the source may vary depending on the type of organ transplant recipients. Whereas in liver transplant recipients candidiasis results from endogenous (generally gut) colonization, donor organs are the potential source in heart-lung and lung transplant recipients. Karyotypic analysis has demonstrated Candida infection originating in the donor lung as a cause of disseminated disease in a lung transplant recipient (28). In the earlier studies in liver transplant recipients, 15% to 20% of the patients were documented to have invasive candidiasis (196, 210). Intraabdominal infections, with or without subsequent dissemination, are the usual clinical manifestations. Prolonged operation time, retransplantation, greater transfusion requirements, high serum creatinine, and CMV infection were the proposed risk factors for Candida infections (Table 59.2). More recently, however, many transplant centers have documented a decline in the incidence of invasive candidiasis, even in the absence of specific antifungal prophylaxis (73, 164, 211). More conservative immunosuppression but, more importantly, improvement in surgical technique likely accounts for this decline. After pancreatic transplantation, Candida infections occur in 15% to 30% of the patients and manifest predominantly as surgical site or bloodstream infections. In heart-lung or lung transplant recipients, the clinical pattern of Candida infections may range from tracheobronchitis to systemic invasive disease. Invasive candidiasis in these patients may also result in anastomotic dehiscence, mediastinitis, and mycotic aneurysm. The anastomotic site is particularly vulnerable because of poor blood supply and the presence of suture material. Invasive bronchial infection can then result with breakdown of the anastomosis. The precise patient population to be targeted, optimal regimen, and duration of antifungal prophylaxis for Candida remains controversial. Nystatin-containing selective bowel decontamination regimens have been proposed to lead to a lower incidence of invasive candidiasis in liver transplant recipients (73, 211). Its efficacy, however, has never been assessed in a controlled trial. Currently, fluconazole is used for prophylaxis at many liver and pancreatic transplant centers. A randomized trial from Europe compared fluconazole (100 mg once a day) with nystatin (106 units every 6 hours) for 28 days after liver transplantation (212). Although the incidence of Candida colonization and superficial fungal infections (thrush and cystitis) were decreased, differences in the incidence of invasive candidiasis were not observed in the two groups (212). Universal use of azole prophylaxis must be undertaken with caution. A major concern with routine azole antifungal prophylaxis is the emergence of azole-resistant Candida, a scenario already documented in liver transplant recipients (213). Azole-resistant invasive Candida glabrata infection occurred in 4% (4/101) of the liver transplant recipients receiving fluconazole prophylaxis and was the direct cause of death in one patient (213). The routine use of fluconazole is expected to have the potential of selecting fungi innately resistant to fluconazole (e.g., Aspergillus). Although this association has not been proved, an unusually high incidence of invasive aspergillosis (8%; 8/101) in liver transplant recipients who routinely received fluconazole P.1000 prophylaxis prospectively is of concern (213). Finally, the need for fluconazole prophylaxis should be assessed on the basis of the institutional trends in the incidence of invasive candidiasis. Given the low incidence of invasive candidiasis at certain transplant centers, the widespread use of fluconazole prophylaxis for all patients may not be warranted to prevent the infection in 5% of the patients. A more appealing approach might be prophylaxis targeted toward high-risk patients only. Pneumocystis Carinii P. carinii has long been classified as a protozoan microorganism on the basis of morphologic features and lack of growth on fungal media. However, gene sequencing of P. carinii suggests that the microorganism is indeed a fungus. Epidemiology The prevailing assumption has long been that P. carinii infection arises from reactivation of endogenous infections acquired in childhood. However, clusters of P. carinii infection occurring in transplant recipients have been reported. In one report, all renal transplant recipients affected had attended the same outpatient facility as did patients with advanced HIV infection (214, 215, 216, 217). Compared with matched control subjects, PCP cases had more outpatient clinic visits coinciding with visits of HIV-infected patients with P. carinii infection (216). Furthermore, P. carinii DNA has been demonstrated in more than 50% of the air samples from the hospital rooms of P. carinii-infected patients (216). Detection of P. carinii DNA was by filtration of air. As molecular epidemiologic techniques become more sophisticated, additional information on such outbreaks may provide more convincing evidence of patient-to-patient transmission of P. carinii. Lung transplant recipients are at greatest risk of developing pulmonary infection with P. carinii; in the absence of prophylaxis, PCP may develop in up to 80% of the patients (218). Not all of these patients, however, are symptomatic. Up to 40% of the lung transplant recipients have been shown to have normal chest radiographs, are asymptomatic, and have the microorganism detected on routine posttransplant bronchoscopy. Several factors may account for the high incidence of P. carinii infection in lung transplant recipients. First, local defense mechanisms are impaired as a result of lung denervation. Second, it has been hypothesized that an incompatibility exists between the immune effector cells and parenchymal cells in the allograft lung. Infiltrating lymphocytes and mononuclear phagocytes recruited to the infected allograft are derived from the recipient and, therefore, express different MHC antigens than do the donor-derived parenchymal cells (55). Finally, surveillance bronchoscopy may increase the chance of early detection of occult infection. P. carinii infection in lung transplant recipients usually occurs in the fourth month after transplantation. Up to 25% of the cases may occur more than a year after transplant; these patients had usually received more intense immunosuppressive therapy. The incidence of infection with P. carinii in patients not receiving PCP prophylaxis is 2% in renal transplant recipients, 5% in heart, and 9% in liver transplant recipients. Most infections occur between 3 and 6 months after transplantation. Between 10% and 20% of the cases occur greater than 6 months posttransplantation, usually in those receiving augmented immunosuppression for rejection (219). Other risk factors for PCP include need for OKT3, CMV infection, and older recipient age. Interestingly, the new immunosuppressive agent mycophenolate mofetil has been shown to have in vitro activity against P. carinii. In four randomized controlled trials of mycophenolate mofetil use in renal transplant recipients, none of 1,068 patients given mycophenolate mofetil developed P. carinii infections compared with 10 of 563 (1.8%) of those randomized not to receive mycophenolate mofetil (p = .00006) (220). Unlike HIV-infected patients, PCP in transplant recipients is rarely diagnosed by examination of an induced sputum sample. Virtually all cases require bronchoalveolar lavage for diagnosis. Co-infection with CMV or bacteria (especially in lung transplant recipients) is common. Prophylaxis Prophylaxis with oral trimethoprim-sulfamethoxazole has proven highly efficacious and is recommended for all transplant recipients. Adverse effects of trimethoprim-sulfamethoxazole are relatively uncommon in solid organ transplant recipients. Rash occurs in only 1%. Leukopenia is somewhat more common, occurring in 3% to 20% of patients. A more controversial issue in prophylaxis of P. carinii infection in transplant recipients is the duration of prophylaxis. Many transplant centers (including ours) offer life-long trimethoprim-sulfamethoxazole. Others use it for the first 6 months posttransplantation for heart, renal, and liver transplant recipients and for the first 12 months in lung transplant recipients. The rationale for this approach is that the risk for P. carinii infection in stable patients declines substantially 6 months after transplantation. However, cases of PCP have been described several months after discontinuing prophylaxis. At one center, 36% of cases of PCP occurred more than 1 year after transplant and 18% occurred greater than 2 years after transplantation (221). In patients receiving augmented immunosuppression for late-occurring rejection, PCP prophylaxis should be resumed or continued. An additional advantage of trimethoprim-sulfamethoxazole prophylaxis is that it is also effective against other microorganisms such as Nocardia, Listeria, Toxoplasma, and Legionella in transplant recipients. Monthly nebulized pentamidine is an alternative prophylaxis in patients intolerant of trimethoprim-sulfamethoxazole. Breakthrough infections, however, have been reported in such patients. Orally administered atovaquone or dapsone would be third-line options. Neither drug has significant interactions with cyclosporine or tacrolimus. Cryptococcus neoformans The incidence of Cryptococcus in organ transplant recipients ranges from 0.6% to 2.6%. However, there are significant regional variations, with some centers reporting cryptococcosis in up to 5% of the transplant recipients (222). Most cases occur more than 6 months after transplantation; some cases are detected P.1001 4 or more years after transplantation. The precise pathogenesis of cryptococcosis is unclear; whether it is a newly acquired or reactivation infection in transplant settings remains unresolved. Given the low incidence of cryptococcosis in transplant recipients and the delayed and often unpredictable time of onset after transplantation, fluconazole prophylaxis is not usually considered necessary in transplant recipients. Endemic Mycosis (Histoplasmosis, Coccidioidomycosis, and Blastomycosis) Histoplasmosis has been infrequently reported in transplant recipients. Some 0.3% of renal transplant recipients followed for several years in a nonendemic area of the United States developed histoplasmosis (223). In endemic regions, about 0.5% of renal transplant recipients may develop histoplasmosis (224), although during an outbreak associated with construction activity near a hospital in Indianapolis, the prevalence of disseminated infection rose to 2.1% (224). The median time to onset is 6 to 15 months after transplantation (223, 224). In endemic areas, primary infection is thought to be the usual presentation. In contrast, in nonendemic regions, reactivation of latent infection with subsequent hematogenous spread is more likely. Transmission through a cadaveric renal allograft from an infected donor has been described (225). Disseminated disease occurs in more than 75% of the transplant recipients developing histoplasmosis. Culture of H. capsulatum remains the gold standard of diagnosis but is often delayed. Serologic tests have proven useful in providing a more rapid diagnosis of histoplasmosis in transplant recipients (224). Coccidioidomycosis in transplant recipients has been described predominantly from the centers in Arizona and southern California. The risk of overt infection in solid organ transplant recipients in Arizona is about 3% per year, with an overall prevalence of 4.5% for heart transplant recipients (226) and 6.9% for renal transplant recipients (227). In liver transplant recipients in Los Angeles, 0.6% of patients developed overt coccidioidomycosis. The usual time to onset is 2 to 6 months posttransplant (226, 227, 228). However, cases have also been reported in the first 4 weeks posttransplantation. Sometimes, this early infection may manifest as fever, a sepsis-like syndrome, and an aggressive pneumonia (228). A subacute presentation occurring several to many months after transplantation, however, is a more common finding. Some patients have disseminated disease with arthritis, meningitis, or skin lesions (229). Transplant candidates who reside in endemic areas should be screened for coccidioidomycosis before transplantation. Patients at high risk of developing coccidioidomycosis posttransplant include those with detectable titers of coccidioidal antibodies on complement fixation tests, those with radiographic evidence of prior pulmonary infection, and those with a history of active coccidioidomycosis. Prophylactic fluconazole (400 mg/day) should be considered for these patients posttransplant. Blastomycosis is rarely reported in transplant recipients. A renal transplant recipient has been reported who sustained a needle puncture while working as a veterinarian's assistant and subsequently developed local skin infection and then disseminated disease (230). In this case it was presumed that the infection was transmitted from an infected dog. Zygomycosis The incidence of zygomycosis complicating organ transplantation ranges between 0.3% and 5% (231). The usual time to onset is 2 months after transplantation (range, 5 days to 8 years). Most cases are due to Rhizophus species, although Mucor, Absidia, and Cunninghamella have also been reported. Rhinocerebral disease has been observed in 57% of the cases and is the most common clinical presentation. Zygomycosis can be acquired nosocomially; the usual portal of entry is believed to be pulmonary. Adhesive bandages have been incriminated as a source of surgical site infections in transplant recipients. PROTOZOA Toxoplasma Gondii Heart transplantation poses the highest risk for transmission of T. gondii because of the parasite's predilection to invade muscular tissue. Fifty percent to 75% of seronegative recipients of seropositive cardiac allografts may develop primary T. gondii infection (232). T. gondii infections are distinctly unusual in noncardiac transplant recipients. Most symptomatic infections occur within 6 months of transplantation. Meningoencephalitis, brain abscess, myocarditis, and pneumonitis are the usual clinical manifestations. Demonstration of tachyzoites in tissue sections establishes the diagnosis of acute infection. An immunoperoxidase stain is both sensitive and specific. Tissue cell cultures (e.g., those used for the isolation of viruses) can also detect T. gondii in culture within 3 to 6 days (233). Significant changes in antibody titer may not occur in transplant recipients with acute toxoplasmosis. Conversely, a rise in IgM and IgG titers is frequent after heart transplantation without evidence of clinical disease. Prophylaxis with pyrimethamine plus folinic acid for 6 weeks is indicated in seronegative cardiac transplant patients who receive seropositive allografts. APPROACH TO MAJOR NOSOCOMIAL INFECTIONS IN TRANSPLANT RECIPIENTS Pulmonary Infiltrates Pulmonary infiltrates, including those resulting from pneumonitis, remain a serious and frequently encountered complication in transplant recipients. In heart transplant recipients, pulmonary infections have been documented in 28% to 40% of patients (2, 162). Fifty-one percent (36/71) of all pulmonary infections in heart transplant recipients in one report were nosocomial in origin; 32% (23/71) were opportunistic and only 17% (12/71) were community-acquired. The latter usually occur in the late posttransplant period (>1 year) (2). Nearly one half of all pneumonias in these patients are bacterial in origin. Prolonged intubation, reintubation, and high-dose corticosteroids were significant risk factors for pneumonia in heart transplant recipients. P.1002 The differential diagnosis of pulmonary infiltrates in lung and heart-lung transplant recipients, among other entities, includes acute rejection. Acute rejection may develop in up to 60% of these patients, and most of these episodes occur within the first 3 months. Obliterative bronchitis is a significant late-occurring complication in lung transplant recipients. Rejection and/or infection, particularly resulting from CMV, are proposed to be the leading risk factors. After liver transplantation, pneumonia occurs in 13% to 34% of patients. Forty-four percent of the liver transplant recipients requiring intensive care unit admission developed pulmonary infiltrates in one study; pulmonary edema (40%), pneumonia (38%), atelectasis (10%), and adult respiratory distress syndrome (8%) were the documented causes (234). Adult respiratory distress syndrome has been reported in 5% to 17% of liver transplant recipients. Large-volume transfusions, liver failure, retransplantation, and sepsis are considered factors predisposing to adult respiratory distress syndrome in these patients (234). Bacteria account for 40% to 67% of all pneumonias in liver transplant recipients. In the earlier reports, viral pneumonitis (CMV or HSV) was documented in 15% to 20% of the pulmonary infections. This incidence, however, has declined to less than 5% in the more recent reports (234). Pneumonias are less common after renal transplantation (occurring in 8% to 16% of the patients) but are nevertheless a significant complication in these patients. Nosocomial bacterial pneumonitis is the predominant cause of pneumonia in all types of solid organ transplantation; P. aeruginosa, Enterobacteriaceae, and S. aureus are the bacteria usually implicated. Legionellosis has been reported in 2% to 9% of solid organ transplant recipients with pneumonia (5). Fungal pneumonias in transplant recipients are predominantly due to invasive aspergillosis. Although isolation of Candida species from respiratory cultures is common, Candida pneumonitis has been documented usually in heart-lung or lung transplant recipients (28). Other less frequent causes of pulmonary infection include cryptococcosis, zygomycosis, coccidioidomycosis, histoplasmosis, and dematiaceous fungi. CMV and to a lesser degree HSV account for most cases of viral pneumonitis. Whereas 5% to 20% of renal, liver, and heart transplant patients develop CMV pneumonitis, the incidence in heart-lung and lung transplant recipients is higher and ranges from 10% to 50%. It is believed that the host immune response is more important for the development of CMV pneumonitis than viral replication is (17). Virus-coded CMV proteins in the lung recognized by host T cells lead to uncontrolled recruitment and accumulation of such cells in the lungs. Pneumonia resulting from RSV generally occurs in pediatric patients and may occur 3 weeks to 2 years after transplantation. Up to 20% to 50% of the cases of RSV are nosocomially acquired. Whereas mortality resulting from RSV in marrow transplant recipients approaches 50%, death resulting from RSV is rare in solid organ transplant recipients. Early identification and isolation of cases, however, is crucial to prevent nosocomial spread. Given the diversity of likely causes, early and aggressive pursuit of the etiology of pulmonary infection is warranted in all transplant recipients. Although the radiographic appearance of the lesion is never diagnostic, a number of entities may have distinctive or suggestive radiographic characteristics. Nodular pulmonary infiltrates of infectious etiology in liver transplant recipients were most frequently due to Aspergillus or Cryptococcus in one study (235). However, S. aureus, Nocardia, tuberculosis, PCP, CMV, zygomycosis, Bartonella henselae, and coccidioidomycosis may present similarly in transplant recipients. Non-infectious causes of pulmonary nodules include metastatic hepatocellular carcinoma, pulmonary calcification, and lymphoproliferative disorders in liver transplant recipients. Pulmonary infarcts, rounded atelectasis, and pulmonary varix in cardiac transplant recipients and acute or chronic rejection in lung transplant patients may have a nodular appearance. Cavitary pneumonia in transplant recipients may be due to Aspergillus, Cryptococcus, Nocardia, Legionella, M. tuberculosis, Rhodococcus equi, or other fungi. CT offers a number of advantages over conventional radiographs, including detection of additional lesions, precise morphology of the lesion, and delineation of mediastinal lymphadenopathy. CTs in patients suspected of having aspergillosis have often revealed lesions that appeared nonspecific or were not visualized on routine x-ray examination. By CT, aspergillosis frequently manifests as nodular lesions, halo signs (nodular opacities with a hazy margin), or an air crescent sign (crescent-shaped area of hyperlucency within a cavitary mass). Isolation or detection of Legionella, Nocardia, Cryptococcus, M. tuberculosis, or P. carinii in the sputum or respiratory secretions is diagnostic of pulmonary infection resulting from these pathogens. However, smears and cultures of sputum or respiratory secretions may be diagnostic in fewer than 50% of patients. In a patient with focal air space disease and nondiagnostic noninvasive tests, the choice lies between empiric antibacterial therapy or a diagnostic procedure; we recommend early bronchoscopy with bronchoalveolar lavage. In patients with focal nodular infiltrates, percutaneous needle aspiration is superior to bronchoalveolar lavage with a diagnostic accuracy of 70% to 90%. In patients with diffuse pulmonary infiltrates, bronchoalveolar lavage with or without transbronchial biopsy is the preferred approach. Transbronchial biopsy is particularly valuable for the diagnosis of rejection in lung transplant recipients and for the differentiation of allograft rejection and CMV pneumonitis in these patients. Open lung biopsy should be reserved only for patients with progressive disease refractory to antimicrobial agents in whom bronchoalveolar lavage or percutaneous needle aspiration is nondiagnostic. Nosocomial Bacteremias Although the incidence of bacteremia may vary, identifiable portals of entry and defined pathogens exist for most solid organ transplant recipients with bacteremia. The frequency of bacteremia varies from 6% to 10% in kidney, 11% in heart, 11% in pancreatic, and 20% to 25% in liver transplant recipients (165, 236). Ninety-four percent (75/80) of all bacteremias in liver transplant recipients, 56% (15/27) in renal transplant recipients, and 78% (13/18) in heart transplant recipients were nosocomially acquired (236). Pneumonia in heart and heart-lung transplant recipients, urinary tract infections in renal transplant recipients, abdominal and biliary infections in liver transplant P.1003 recipients, and surgical site and urinary tract infections in pancreatic transplant recipients have been shown to be the most common identifiable sources of bacteremia (165, 236). Aerobic gram-negative bacilli constituted 48% of the bacteremic isolates in renal transplant recipients, 49% in liver transplant recipients, and 39% in heart transplant recipients (236). E. coli, Klebsiella, and other Enterobacteriaceae were the predominant gram-negative bacteria in renal transplant recipients; P. aeruginosa and Enterobacter species were predominant in liver transplant recipients; and P. aeruginosa and E. coli were predominant in heart transplant recipients. The sources and pathogens causing bacteremias in transplant recipients appear, however, to have undergone a striking evolution in the recent years. Many transplant centers have documented the emergence of gram-positive cocci (enterococci and staphylococci) as foremost pathogens in transplant recipients. In a recent study in liver transplant recipients, intravascular cannulas were by far the most common source, accounting for 39% of all bacteremias (163). Fifty-eight percent of all catheter-associated infections were due to MRSA (163). Within the hospital, intensive care units are the most common site of acquisition of nosocomial bacteremias. Ninety-three percent of the bacteremias in one report were nosocomially acquired, and 52% occurred in the intensive care unit setting (163). Indeed, intensive care unit stay has been shown to be an independent predictor of bacteremic compared with nonbacteremic infections in liver transplant recipients (44). Patients developing bacteremia in the intensive care unit were also more likely to die than those not in the intensive care unit (44); this likely reflected the greater severity of illness of the patients hospitalized in the intensive care unit. In a study in small bowel transplant recipients, 72% of patients had at least one episode of bloodstream infection (69). Intravascular cannulas accounted for 43% of these infections and were the most frequently identifiable portal of entry. Sixty-two percent of all bloodstream infections were due to gram-positive bacteria (69). In another report, 45% of all bacterial infections and 72% of all bacteremias after small bowel transplantation were due to staphylococci (72). Most VRE infections originate from an abdominal or biliary source; 38% to 68% of these infections have been associated with bacteremia (170, 172). Vancomycin resistance was an independent predicator of mortality in liver transplant recipients with enterococcal bacteremia (170). A noteworthy observation is the predilection of VRE to cause endovascular complications, including mycotic aneurysms and endocarditis (170). Delayed metastatic complications, including endocarditis and osteomyelitis, have also been documented in transplant recipients with MRSA bacteremia. Prophylaxis and infection control measures pertinent to these infections are discussed in the sections on staphylococci and enterococci. Intraabdominal Infections Intraabdominal abscesses, peritonitis, and biliary infections are a significant complication after liver transplantation. Intrahepatic abscesses usually occur within 30 days of transplantation; technical problems involving the implanted allograft (e.g., hepatic artery thrombosis, biliary leak, and, rarely, tear of the donor liver) are the primary risk factors. Nearly one half of patients with intrahepatic abscesses may be bacteremic (46). Peritonitis after liver transplantation is typically related to biliary anastomotic leaks or, less frequently, bowel perforation (46). Aerobic enteric gram-negative bacteria, enterococci, anaerobes, and, rarely, Candida species are the causal pathogens in most intraabdominal abscesses and peritonitis. An unusual cause of abdominal abscesses in liver transplant recipients is M. hominis. Cholangitis has been documented in 4% to 15% of patients after liver transplantation, with most episodes occurring within 30 days of transplantation. Biliary strictures and biliary leaks are significant predisposing risk factors. Enterococci are characteristically the most common pathogen, whereas anaerobes are encountered rarely. Biliary strictures (e.g., in patients with primary biliary cirrhosis and in those with previous bile duct surgeries) often necessitate Roux-en-Y anastomosis, which is associated with a high rate of intrahepatic and biliary complications. Sterile intraabdominal fluid collections are common after liver transplant surgery. Diagnosis of peritonitis or abscess, thus, requires percutaneous or open drainage and culture. Cultures from abdominal drains often reflect colonization and are not reliable in diagnosing the infection or its etiology. Central Nervous System Lesions Although mild neurologic complications may occur in 10% to 47% of organ transplant recipients, 1% to 8% of such patients have major neurologic sequelae or central nervous system lesions. Brain abscesses are among the most serious central nervous system lesions in these patients (237). The frequency of brain abscesses was 0.36% in kidney, 0.63% in liver, and 1.17% in heart and heart-lung transplant recipients (238). Most brain abscesses in solid organ transplant recipients are fungal and represent a nosocomial complication (237, 238). Although Aspergillus is the most common cause of brain abscesses in organ transplant recipients, less frequently encountered fungal pathogens include Mucorales, Candida species, and dematiaceous fungi. Central nervous system lesions resulting from T. gondii and Nocardia have been reported mainly in heart transplant recipients. Selby et al. (238) showed that two distinct groups of solid organ transplant recipients existed with regard to timing and susceptibility to brain abscesses. One group, comprised predominantly of liver and renal transplant recipients, developed brain abscesses a median of 24 days posttransplant. Ninety-five percent of these patients were in the intensive care unit and were ventilator dependent; brain abscesses in this setting were exclusively fungal. In the second group, abscesses developed a median of 264 days after transplantation, occurred almost exclusively in heart transplant recipients, and were due to T. gondii and Nocardia. Most patients (up to 75%) with fungal brain abscesses have been shown to concurrently have pulmonary lesions resulting from the same fungus (237). A brain biopsy may not be required in such cases. In the absence of an extraneural focus, brain biopsy should be considered, given the diversity of causal fungal pathogens in such lesions (237). P.1004 CONCLUSIONS Improvement in surgical techniques, the advent of modern immunosuppressive drugs, the availability of rapid and reliable diagnostic modalities, and effective prophylaxis against a number of pathogens have undoubtedly led to a decrease in infectious morbidity and improved outcome in transplant recipients in recent years. Increasing documentation of the emergence of antimicrobial resistance in several key pathogens in transplant recipients, however, is worrisome. Strategies for antimicrobial prophylaxis must comprise approaches that are not only efficacious but minimize the emergence of resistance. Finally, for therapy and prevention to be effective, classic opportunistic infections typically encountered in these patients must be considered and the emerging trends in new infectious agents and the changing epidemiology of these complicating infections must be understood. REFERENCES 1. Chang FY, Singh N, Gayowski T, et al. Fever in liver transplant recipients: changing spectrum of etiologic agents. Clin Infect Dis 1998;26:59–65. 2. Waser M, Maggiorini M, Luthy A, et al. Infectious complications in 100 consecutive heart transplant recipients. Eur J Clin Microbiol Infect Dis 1994;13:12–18. 3. Singh N, Yu VL. Oral ganciclovir usage for cytomegalovirus prophylaxis in organ transplant recipients. Is emergence of resistance imminent? Dig Dis Sci 1998;43:1190–1192. 4. Sawyer RG, Crabtree TD, Gleason TG, et al. Impact of solid organ transplantation and immunosuppression on fever, leukocytosis, and physiologic response during bacterial and fungal infections. Clin Transplantation 1999;13:260–265. 5. Chow JW, Yu VL. Legionella a major opportunistic pathogen in transplant recipients. Semin Respir Infect 1998;13:132–139. 6. Kirby BD, Snyder K, Meyer R, et al. Legionnaires' disease: a report of 65 nosocomial cases and a review of the literature. Medicine 1980;59:188–205. 7. Jereb JA, Burwen DR, Dooley SW, et al. Nosocomial outbreak of tuberculosis in a renal transplant unit: application of a new technique for restriction fragment length polymorphism analysis of Mycobacterium tuberculosis isolates. J Infect Dis 1993;168:1219–1224. 8. Paterson DL, Muder RR, Squier C, et al. Outbreak of infection with E. coli harboring extended spectrum beta-lactamase (ESBL) in a liver transplant unit: could norfloxacin administration to carriers contribute to control? Clin Infect Dis 2001;33:126–128. 9. DelMonico FL, Snydman DR. Organ donor screening for infectious diseases. Transplantation 1998;65:603–610. 10. Douglas D, Rakela R, Marnish D, et al. Transmission of hepatitis B virus infection from orthotopic donor livers. Hepatology 1992;16:49A. 11. Wachs M, Amend W, Ascher N, et al. The risk of transmission of hepatitis B from HBsAg(-), HBcAb(+), HBIgM(-) organ donors. Transplantation 1995;59:230. 12. Munoz SJ. Use of hepatitis B core antibody-positive donors for liver transplantation. Liver Transplantation 2002;10:S82–S87. 13. Vargas HE, Laskus T, Wang LF, et al. Outcome of liver transplantation in hepatitis C virus-infected patients who received hepatitis C virus-infected grafts. Gastroenterology 1999;117:149–153. 14. Torres M, Weppler D, Reddy KR, et al. Use of hepatitis C virus-infected donors for hepatitis C positive OHCT recipients. Gastroenterology 1999;117:1253. 15. Ho M, Suwansirikul S, Dowling JN, et al. The transplanted kidney as a source of cytomegalovirus infection. N Engl J Med 1975;293:1109–1112. 16. Betts RF, Freeman RB, Douglas RG, et al. Transmission of cytomegalovirus infection with renal allografts. Kidney Int 1975;8:385–392. 17. Grundy JE, Shanley JD, Griffiths PD. Is cytomegalovirus interstitial pneumonitis in transplant recipients an immunopathological condition? Lancet 1987;2:996–999. 18. Dummer JS, Armstrong J, Somers J, et al. Transmission of infection with herpes simplex virus by renal transplantation. J Infect Dis 1987;155:202–206. 19. Cen H, Breinig MD, Atchison RW, et al. Epstein-Barr virus transmission via the donor organs in solid organ transplantation: polymerase chain reaction and restriction fragment length polymorphism analysis of IR2, IR3, and IR4. J Virol 1991;65:976–980. 20. Ward KN, Gray JJ, Efstathiou S. Primary human herpesvirus 6 infection in a patient following liver transplantation from a seropositive donor. J Med Virol 1989;28:69–72. 21. Singh N, Dummer JS, Kusne S, et al. Infections with cytomegalovirus and the herpesviruses in 121 liver transplant recipients: transmission by donated organ and the effect of OKT3 antibodies. J Infect Dis 1988;158:124–131. 22. Caballero F, Domingo P, Rabella N, et al. Successful transplantation of organs retrieved from a donor with syphilis. Transplantation 1998;65:598–599. 23. Peters TG, Reiter CG, Boswell RL. Transmission of tuberculosis by kidney transplantation. Transplantation 1984;38:514–516. 24. Singh N, Paterson DL. M. tuberculosis infection in solid organ transplant recipients: impact and implications for management. Clin Infect Dis 1998;27:1266–1277. 25. Lumbreras C, Samz F, Gonazlez A, et al. Clinical significance of donor-unrecognized bacteremia in the outcome of solid-organ transplant recipients. Clin Infect Dis 2001;33:722–726. 26. Freeman RB, Giatras I, Falagas ME, et al. Outcome of transplantation of organs procured from bacteremic donors. Transplantation 1999;68:1107–1111. 27. Lopez-Navidad A, Domingo P, Caballero F, et al. Successful transplantation of organs retrieved from donors with bacterial meningitis. Transplantation 1997;64:365–368. 28. Kanj SS, Welty-Wolf K, Madden J, et al. Fungal infections in lung and heart-lung transplant recipients: report of 9 cases and review of the literature. Medicine 1996;75:142–156. 29. Gottesdiener KM. Transplanted infections: donor-to-host transmission with the allograft. Ann Intern Med 1989;110:1001–1016. 30. Preiksaitis JK, Straufman TM. The risk of transfusion-acquired cytomegalovirus (CMV) infection in seronegative solid organ transplant recipients receiving unscreened blood products (1984–1996) (abstract no. 450, p. 191). Seventeenth annual meeting of the American Society of Transplant Physicians, Chicago, 1998. 31. Falagas ME, Snydman DR, Ruthazer R, et al. Primary cytomegalovirus infection in liver transplant recipients: comparison of infections transmitted via donor organs and via transfusions. Clin Infect Dis 1996;23:292–297. 32. Ernst A, Gordon FD, Hayek J, et al. Lung abscess complicating Legionella micdadei pneumonia in an adult liver transplant recipient. Transplantation 1998;65:130–134. 33. Prodinger WM, Bonatti H, Allerberger F, et al. Legionella pneumonia in transplant recipients: a cluster of cases of eight years duration. J Hosp Infect 1994;26:191–202. 34. Bangsborg JM, Uldum S, Jensen JS, et al. Nosocomial legionellosis in three heart-lung transplant patients: case reports and environmental observations. Eur J Clin Microbiol Infect Dis 1995;14:99–104. 35. Wald A, Leisenring W, vanBurik J, et al. Epidemiology of Aspergillus infections in a large cohort of patients undergoing bone marrow transplantation. J Infect Dis 1997;175:1459–1466. 36. Anaissie E. Invasive aspergillosis: a new look at an old disease. In Focus on fungal infections. Orlando, FL, 1998. 37. Noskin GA, Stosor V, Cooper I, et al. Recovery of vancomycin-resistant enterococci on fingertips and environmental surfaces. Infect Control Hosp Epidemiol 1995;16:577–581. 38. Morris JG Jr, Shay DK, Hebden JN, et al. Enterococci resistant to multiple antimicrobial agents, including vancomycin: establishment of endeminicity in a university medical center. Ann Intern Med 1995;123:250–259. P.1005 39. Chang FY, Singh N, Gayowski T, et al. Staphylococcus aureus nasal colonization and association with infections in liver transplant recipients. Transplantation 1998;65:1169–1172. 40. Samore MH, DeGirolami PC, Tlucko A, et al. Clostridium difficile colonization and diarrhea at a tertiary care hospital. Clin Infect Dis 1994;18:181–187. 41. Kaatz GW, Gitlin SD, Schaberg DR, et al. Acquisition of Clostridium difficile from the hospital environment. Am J Epidemiol 1998;127:1289–1294. 42. McFarland LV, Mulligan ME, Kwok RYY, et al. Nosocomial acquisition of Clostridium difficile infection. N Engl J Med 1998;320:204–210. 43. Gayowsksi T, Marino IR, Singh N, et al. Orthotopic liver transplantation in high risk patients: risk factors associated with mortality and infectious morbidity. Transplantation 1997;65:499–504. 44. Singh N, Gayowski T, Wagener MM, et al. Predictors and outcome of early versus late-onset major bacterial infections in liver transplant recipients receiving tacrolimus (FK506) as primary immunosuppression. Eur J Clin Micro Infect Dis 1997;16:821–826. 45. Paya CV. Fungal infections in solid-organ transplantation. Clin Infect Dis 1993;16:677–688. 46. Kusne S, Dummer JS, Singh N, et al. Infections after liver transplantation, an analysis of 101 consecutive cases. Medicine 1988;67:132–143. 47. Gayowski T, Marino IR, Doyle HR, et al. A high incidence of native portal vein thrombosis in veterans undergoing liver transplantation. J Surg Res 1996;60:333–338. 48. Singh N, Gayowski T, Wagener MM, et al. Increased infections in liver transplant recipients with recurrent hepatitis C virus hepatitis. Transplantation 1996;61:402–406. 49. Hoy WE, May AG, Freeman RB. Primary renal transplant wound infections. N Y State J Med 1981;81:1469–1473. 50. Lobo PI, Rudolf LE, Kreiger JN. Wound infections in renal transplant recipients—a complication of urinary tract infections during allograft malfunction. Surgery 1982;92:491–496. 51. Lapchik MS, Filho AC, Restana JO, et al. Risk factors for nosocomial urinary tract and postoperative wound infections in renal transplant patients: a matched pair case control study. J Urology 1992;147:994–998. 52. Muakkassa WF, Goldman MH, Mendez-Picon G, et al. Wound infections in renal transplant patients. J Urology 1983;130:17–19. 53. Tilney NL, Strom TB, Vineyard GC, et al. Factors contributing to the declining mortality rate in renal transplantation. N Engl J Med 1978;299:1321–1325. 54. Fox BD, Sollinger HW, Belzer FO, et al. A prospective, randomized double-blind study of trimethoprim-sulfamethoxazole for prophylaxis of infection in renal transplantation: clinical efficacy, absorption, effects on microflora and the cost-benefit of prophylaxis. Am J Med 1990;89:255–274. 55. Dauber JH, Paradis IL, Dummer JS. Infectious complications in pulmonary allograft recipients. Clin Chest Med 1990;11:291–308. 56. Kramer MR, Marshall SE, Starnes VA, et al. Infectious complications in heart-lung transplantation, analysis of 200 episodes. Arch Intern Med 1993;153:2010–2016. 57. Bauldoff GS, Nunley DR, Manzetti JD, et al. Use of aerosolized colistin sodium in cystic fibrosis patients awaiting lung transplantation. Transplantation 1997;64:748–752. 58. Lewiston N, King V, Umetsu D, et al. Cystic fibrosis patients who have undergone heart-lung transplantation benefit from maxillary sinus antrostomy and repeated sinus lavage. Transplant 1991;23:1207–1208. 59. Baldwin RT, Radovancevic B, Sweeney MS, et al. Bacterial mediastinitis after heart transplantation. J Heart Lung Transplant 1992;11:545–549. 60. Karwande SV, Renlund DG, Olsen SL, et al. Mediastinitis in heart transplantation. Ann Thorac Surg 1992;54:1039–1045. 61. Trento A, Dummer JS, Hardesty RL, et al. Mediastinitis following heart transplant: incidence, treatment and results. Heart Transplant 1984;3:336–340. 62. Pearl SN, Weiner MA, Dibbell DG. Sternal infection after cardiac transplantation. Successful salvage utilizing a variety of techniques. Thorac Cardiovasc Surg 1982;83:632–634. 63. Steffenson DO, Dummer JS, Granick MS, et al. Sternotomy infections with Mycoplasma hominis. Ann Intern Med 1987;106:204–208. 64. Sollinger HW, Ploeg RJ, Eckhoff DE, et al. Two hundred consecutive simultaneous pancreas-kidney transplants with bladder drainage. Surgery 1993;114:736–744. 65. Everett JE, Wahoff DC, Statz GKJ, et al. Characterization and impact of wound infection after pancreas transplantation. Arch Surg 1994;129:1310–1317. 66. Hesse UJ, Sutherland DER, Simmons RL, et al. Intra-abdominal infections in pancreas transplant recipients. Clin Transplantation 1990;4:287–291. 67. Tesi RJ, Henry ML, Elkhammas EA, et al. Decreased wound complications of combined kidney/pancreas transplants using intraabdominal pancreas graft placement. Clin Transplantation 199;4:287–291. 68. Manez R, Kusne S, Green M, et al. Incidence and risk factors associated with the development of cytomegalovirus disease after intestinal transplantation. Transplantation 1995;59:1010–1014. 69. Kusne S, Furukawa H, Abu-Elmagd K, et al. Infectious complications after small bowel transplantation in adults: an update. Transplant Proc 1996;5:2761–2762. 70. Woodward JE, Mayer D. The unique challenge of small intestinal transplantation. Br J Hosp Med 1996;xx:285–290. 71. Reyes J, Bueno J, Kocoshis S, et al. Current status of intestinal transplantation in children. J Pediatr Surg 1998;33:243–254. 72. McAlister V, Grant D, Wall W, et al. Immunosuppressive requirements for small bowel/liver transplantation. Transplant Proc 1993;25:1204–1205. 73. Singh N, Gayowski T, Wagener MM, et al. Invasive fungal infections in liver transplant recipients receiving tacrolimus as primary immunosuppressive agent. Clin Infect Dis 1997;24:179–184. 74. Singh N, Avery RK, Munoz P, et al. Trends in risk profiles for and mortality associated with invasive aspergillosis among liver transplant recipients. Clin Infect Dis 2003;36:46–52. 75. Singh N, Carrigan DR. Human herpesvirus-6 in transplantation: an emerging pathogen. Ann Intern Med 1996;124:1065–1071. 76. Hibberd PL, Basgoz N, Fishman J, et al. Primary cytomegalovirus disease is a risk factor for recurrent CMV disease in transplant recipients (abstract H28). Thirty-first Interscience Conference of Antimicrobial Agents and Chemotherapy, 1995, Orlando, FL. 77. Stratta RJ, Shaeffer MS, Markin RS, et al. Cytomegalovirus infection and disease after liver transplantation, an overview. Diagn Dis Sci 1992;37:673–688. 78. Hibberd PL, Tolkoff-Rubin NE, Conti D, et al. Preemptive ganciclovir therapy to prevent cytomegalovirus disease in cytomegalovirus antibody-positive renal transplant recipients. A randomized controlled trial. Ann Intern Med 1995;123:18–26. 79. Grundy JE, Lui SF, Super M, et al. Symptomatic cytomegalovirus infection in seropositive kidney recipients: reinfection with donor virus rather than reactivation of recipient virus. Lancet 1988;2:132–135. 80. Fietze E, Prosch S, Reinke P, et al. Cytomegalovirus infection in transplant recipients. The role of tumor necrosis factor. Transplantation 1994;58:675–680. 81. Paterson DL, Singh N, Stapelfeldt W, et al. Intraoperative hypothermia is an independent risk factor for cytomegalovirus (CMV) infection in liver transplant recipients. Transplantation 1999;67:1151–1155. 82. Manez R, White LT, Linden P, et al. The influence of HLA matching on cytomegalovirus hepatitis and chronic rejection after liver transplantation. Transplantation 1993;55:1067–1071. 83. Docke WD, Prosch S, Fietze E, et al. Cytomegalovirus reactivation and tumour necrosis factor. Lancet 1994;343:268–269. 84. Haagmans BL, Stals FS, van der Meide PH, et al. Tumor necrosis factor alpha promotes replication and pathogenicity of rat cytomegalovirus. J Virol 1994;68:2297–2304. 85. Chatterjee SN, Fiala N, Weiner J. Primary cytomegalovirus and opportunistic infections: incidence in renal transplant recipients. JAMA 1978;240:2446–2449. P.1006 86. Limaye AP, Corey L, Koelle DM, et al. Emergence of ganciclovir-resistant cytomegalovirus disease among solid organ transplant recipients. Lancet 2000;356:645–649. 87. Singh N. Delayed occurrence of cytomegalovirus disease in organ transplant recipients receiving antiviral prophylaxis: are we winning the battle only to lose the war? Eur J Clin Microbiol Infect 2002. 88. Paya CV. A randomized, double-blind, double-dummy, active-comparator controlled multi-center study of the efficacy and safety of valganciclovir vs oral ganciclovir for prevention of CMV disease in 372 high-risk (D+/R-) heart, liver and kidney recipients. Forty-second Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, September 27–30, 2002. 89. Ho M. Infection and organ transplantation. In: Gelman S, editor. Anesthesia and organ transplantation. Philadelphia: WB Saunders, 1987:49–60. 90. Gleaves CA, Smith TF, Shuster EA, et al. Comparison of standard tube and shell vial cell culture techniques for the detection of cytomegalovirus in clinical specimens. J Clin Microbiol 1985;21:217–221. 91. The TH, Vander BIJW, VandenBerg AP, et al. Cytomegalovirus antigenemia. Rev Infect Dis 1990;12(Suppl):737–744. 92. Gerna G, Zavattoni M, Percivalle E, et al. Rising levels of human cytomegalovirus (HCMV) antigenemia during initial antiviral treatment of solid-organ transplant recipients with primary HCMV infection. J Clin Microbiol 1998;36:1113–1116. 93. Patel R, Smith TF, Espy M, et al. A prospective comparison of molecular diagnostic techniques for the early detection of cytomegalovirus in liver transplant recipients. J Infect Dis 1995;171:1010–1014. 94. Cunningham R, Harris A, Frankton A, et al. Detection of cytomegalovirus using PCR in serum from renal transplant recipients. J Clin Pathol 1995;48:575–577. 95. Schmidt CA, Oettle H, Peng R, et al. Comparison of polymerase chain reaction from plasma and buffy coat with antigen detection and occurrence of immunoglobulin M for the demonstration of cytomegalovirus infection after liver transplantation. Transplantation 1995;59:1133–1138. 96. Ackermann JR, LeFor WM, Weinstein S, et al. Four-year experience with exclusive use of cytomegalovirus antibody (CMV-Ab)-negative donors for CMV-Ab-negative kidney recipients. Transplant Proc 1988;20(Suppl 1):469–471. 97. Balfour HH, Chace BA, Stepleton JT, et al. A randomized, placebo controlled trial of oral acyclovir for the prevention of cytomegalovirus disease in recipients of renal allografts. N Engl J Med 1989;320:1381–1387. 98. Singh N, Yu VL, Mieles L, et al. High-dose acyclovir compared with short-course preemptive ganciclovir therapy to prevent cytomegalovirus disease in liver transplant recipients, a randomized trial. Ann Intern Med 1994;120:375–381. 99. Martin MM, Manez R, Linden P, et al. A prospective randomized trial comparing sequential ganciclovir-high dose acyclovir to high dose acyclovir for prevention of cytomegalovirus disease in adult liver transplant recipients. Transplantation 1994;58:779–785. 100. Merigan TC, Renlund DG, Keay S, et al. A controlled trial of ganciclovir to prevent cytomegalovirus disease after heart transplantation. N Engl J Med 1993;326:1182–1186. 101. Winston DJ, Wirin D, Shaked A, et al. Randomized comparison of ganciclovir and high-dose acyclovir for long-term cytomegalovirus prophylaxis in liver transplant recipients. Lancet 1995;346:69–74. 102. Alain S, Honderlick P, Grenet D, et al. Failure of ganciclovir treatment associated with selection of a ganciclovir resistant cytomegalovirus strain in a lung transplant recipient. Transplantation 1997;63:1533–1536. 103. Rosen HR, Brenner KG, Flora KD, et al. Development of ganciclovir resistance during treatment of primary cytomegalovirus infection after liver transplantation. Transplantation 1997;63:476–478. 104. Iberer F, Tscheliessnigg K, Halwache G, et al. Definitions of cytomegalovirus disease after heart transplantation: antigenemia as a marker for antiviral therapy. Transplant Int 1996;9:236–242. 105. Verdonck LF, Dekker AW, Rosenberg-Arska M, et al. A risk-adapted approach with a short-course of ganciclovir to prevent cytomegalovirus pneumonia in CMV-seropositive recipients of allogeneic bone marrow transplants. Clin Infect Dis 1997;24:901–907. 106. Grossi P, Kusne S, Rinaldo C, et al. Guidance of ganciclovir therapy with pp65 antigenemia in cytomegalovirus-free recipients of livers from seropositive donors. Transplantation 1996;61:1659–1660. 107. Kusne S, Schwartz M, Breinig MK, et al. Herpes simplex virus hepatitis after solid organ transplantation in adults. J Infect Dis 1991;163:1001–1007. 108. Holt ND, Gould FKI, Taylor CE, et al. Incidence and significance of noncytomegalovirus viral respiratory infection after adult lung transplantation. J Heart Lung Transplant 1997;16:416–419. 109. Schuller D, Spessert C, Fraser VJ, et al. Herpes simplex virus from respiratory tract secretions: epidemiology, clinical characteristics, and outcome in immunocompromised and nonimmunocompromised hosts. Am J Med 1993;94:29–34. 110. McGregor RS, Zitelli BJ, Urbach AH, et al. Varicella in pediatric orthotopic liver transplant recipients. Pediatrics 1989;83:256–261. 111. Kusne S, Pappo O, Manez R, et al. Varicella zoster virus hepatitis and a suggested management plan for prevention of VZV infection in adult transplant recipients. Transplantation 1995;xx:619–621. 112. Feldhoff C, Balfour H, Simmons SR, et al. Varicella in children with renal transplants. J Pediatr 1981;98:25–31. 113. Kashtan CE, Cook M, Chavers BM, et al. Outcome of chicken pox in 66 pediatric renal transplant recipients. J Pediatr 1997;131:874–877. 114. Broyer M, Tete MJ, Guest G, et al. Varicella and zoster in children after kidney transplantation: long-term results of vaccination. Pediatrics 1997;99:35–39. 115. Ablashi D, Agut H, Berneman Z, et al. Human herpesvirus-6 strain groups: a nomenclature. Arch Virol 1993;129:363–366. 116. Schirmer EC, Wyatt LS, Yamanishi K, et al. Differentiation between two distinct classes of viruses now classified as human herpesvirus-6. Proc Natl Acad Sci U S A 1991;88:5922–5926. 117. Singh N, Gayowski T, Wagener M, et al. Bloodstream infections in liver transplant recipients; changing patterns of microbial origin (abstract). Thirty-sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, New Orleans, 1996. 118. Yoshikawa T, Suga S, Asano Y, et al. A prospective study of human herpesvirus-6 infection in renal transplantation. Transplantation 1992;54:879–883. 119. Randhawa PS, Jenkins FJ, Nalesnik MA, et al. Herpesvirus 6 variant A infection after heart transplantation with giant cell transformation in bile ductular and gastroduodenal epithelium. Am J Surg Pathol 1997;21:847–853. 120. DesJardin JA, Gibbons L, Cho E, et al. Human herpesvirus 6 reactivation is associated with cytomegalovirus infection and syndromes in kidney transplant recipients at risk for primary cytomegalovirus infection. J Infect Dis 1998;178:1783–1786. 121. Agut H, Collandre H, Aubin JT, et al. In vitro sensitivity of human herpesvirus 6 to antiviral drugs. Res Virol 1989;140:219–228. 122. Russler SK, Tapper MA, Carrigan DR. Susceptibility of human herpesvirus 6 to acyclovir and ganciclovir. Lancet 1989;2:382. 123. Streicher HZ, Hung CL, Ablashi DV, et al. In vitro inhibition of human herpesvirus-6 by phosphonoformate. J Virol Method 1988;21:301–304. 124. Wright TL, Donegan E, Hsu HH, et al. Recurrent and acquired hepatitis C viral infection in liver transplant recipients. Gastroenterology 1992;103:317–322. 125. Muller H, Otto G, Gosser T, et al. Recurrence of hepatitis C viral infection in liver transplant recipients. Transplantation 1992;54:743–745. 126. Gane EJ, Portman BC, Naumov NV, et al. Long-term outcome of hepatitis C infection after liver transplantation. N Engl J Med 1996;334:815–820. 127. Druwe PM, Michielsen PP, Ramon AM, et al. Hepatitis C and nephrology. Nephrol Dial Transplant 1994;9:223–237. 128. Morales JM, Munoz MA, Castellano G, et al. Impact of hepatitis C in long-functioning renal transplants: a clinicopathological followup. Transplant Proc 1993;25:1450–1453. 129. Klauser R, Franz M, Traindl O, et al. Hepatitis C antibody in renal transplant patients. Transplant Proc 1992;24:286–288. P.1007 130. Roth D, Ferrandez JA, Burke GW, et al. Detection of antibody to hepatitis C virus in renal transplant recipients. Transplantation 1991;51:396–400. 131. Stempel CA, Lake J, Kuo G, et al. Hepatitis C—its prevalence in end-stage renal failure patients and clinical course after kidney transplantation. Transplantation 1993;55:273–276. 132. Feray C, Gigou M, Samuel D, et al. Natural history of hepatitis C virus (HCV) infection after liver transplantation. Hepatology 1992;16:47A. 133. Goetz A, Ndimbie OK, Wagener MM, et al. Prevalence of hepatitis C infection in health care workers affiliated with a liver transplant center. Transplantation 1995;59:990–994. 134. Gayowski T, Singh N, Marino IR, et al. Hepatitis C virus genotypes in liver transplant recipients: impact on posttransplant recurrence, infections, response to interferon-alpha therapy and outcome. Transplantation 1997;64:422–426. 135. Sheiner CP, Schwartz ME, Mor E, et al. Severe or multiple rejection episodes are associated with early recurrence of hepatitis C after orthotopic liver transplantation. Hepatology 1995;21:30–34. 136. Singh N, Gayowski T, Ndimbie OK, et al. Recurrent hepatitis C virus hepatitis in liver transplant recipients receiving tacrolimus: association with rejection and increased immunosuppression after transplantation. Surgery 1996;119:452–456. 137. Feray C, Gigou M, Samuel D, et al. Incidence of hepatitis C in patients receiving different preparations of hepatitis B immunoglobulins after liver transplantation. Ann Intern Med 1998;128:810–816. 138. Singh N, Gayowski T, Wannstedt CF, et al. Interferon-alpha as prophylaxis for recurrent viral hepatitis C after liver transplantation. Transplantation 1998;75:82–85. 139. Samuel D, Muller R, Alexander G, et al. Liver transplantation in European patients with hepatitis B surface antigen. N Engl J Med 1993;329:1842–1847. 140. Angus PW, Locarini SA, McGurghan GW, et al. Hepatitis B virus precore mutant infection is associated with severe recurrent disease after liver transplantation. Hepatology 1995;21:14–18. 141. Davies SE, Portmann B, O'Grady JG, et al. Hepatic histological findings after transplantation for chronic hepatitis B virus infection, including a unique pattern of fibrosing cholestatic hepatitis. Hepatology 1991;13:150–157. 142. Rao KV, Kasiske BL, Anderson WR. Variability in the morphological spectrum and clinical outcome of chronic liver disease in hepatitis B-positive and B-negative renal transplant recipients. Transplantation 1991;51:391–396. 143. Markowitz JS, Martin P, Conrad AJ, et al. Prophylaxis against hepatitis B recurrence following liver transplantation using combination lamivudine and hepatitis B immune globulin. Hepatology 1998;28:585–589. 144. Han S-H, Ofman J, Holt C, et al. An efficacy and cost-effectiveness analysis of combination hepatitis B immune globulin and lamivudine to prevent recurrent hepatitis B after orthotopic liver transplantation compared with hepatitis B immune globulin monotherapy. Liver Transplant 2000;6:741–748. 145. Schiff E, Neuhaus P, Tillmann HL, et al. Safety and efficacy of adefovir dipivoxil for the treatment of lamivudine resistant HBV in patients post-liver transplantation. Hepatology 2001;34:446A(abst). 146. Shakil AO, Lilly L, Angus P, et al. Entecavir reduces viral load in liver transplant patients who have failed prophylaxis or treatment for hepatitis B. Hepatology 2001;34:619A(abst). 147. Cotler SJ, Gretch D, Bronner MP. Hepatitis G virus co-infection does not alter the course of recurrent hepatitis C virus infection in liver transplantation recipients. Hepatology 1997;26:432–436. 148. Vargas HE, Laskus T, Radkowski M, et al. Hepatitis G virus coinfection in hepatitis C virus-infected liver transplant recipients. Transplantation 1997;64:786–788. 149. Dickson RC, Qian K, Lau JYN. High prevalence of GB virus-C/hepatitis G virus infection in liver transplant recipients. Transplantation 1997;63:1695–1697. 150. Pessoa MG, Terrault NA, Ferrell LD, et al. Hepatitis G virus in patients with cryptogenic liver disease undergoing liver transplantation. Hepatology 1997;25:1266–1270. 151. Randhawa PS, Demetris AJ. Nephropathy due to polyomavirus type BK. N Engl J Med 2000;342:1361–1362. 152. Hirsch HH, Knowles W, Dickenmann M, et al. Prospective study of polyomavirus type BK replication and nephropathy in renal-transplant recipients. N Engl J Med 2002;347:488–496. 153. Michaels MG, Green M, Wald ER, et al. Adenovirus infection in pediatric liver transplant recipients. J Infect Dis 1992;165:170–174. 154. Koneru B, Atchison R, Jaffe R, et al. Serologic studies of adenoviral hepatitis following pediatric liver transplantation. Transplant Proc 1990;22:1547–1548. 155. Ohori NP, Michaels MG, Jaffe R, et al. Adenovirus pneumonia in lung transplant recipients. Hum Pathol 1995;26:1073–1079. 156. McGrath D, Falagas ME, Freeman R, et al. Adenovirus infection in adult orthotopic liver transplant recipients: incidence and clinical significance. J Infect Dis 1998;177:459–462. 157. Saad RS, Demetris AJ, Lee RG, et al. Adenovirus hepatitis in the adult allograft liver. Transplantation 1997;64:1483–1485. 158. Liles WC, Cushing H, Holt CS, et al. Severe adenoviral nephritis following bone marrow transplantation: successful treatment with intravenous ribavirin. Transplantation 1993;12:409–412. 159. Pohl C, Green M, Wald ER, et al. Respiratory syncytial virus infections in pediatric liver transplant recipients. J Infect Dis 1992;165:166–169. 160. Wendt CH, Fox JNK, Hertz MI. Paramyxovirus infection in lung transplant recipients. J Heart Lung Transplant 1995;14:479–485. 161. Mauch TJ, Crouch NA, Freese DK, et al. Antibody response of pediatric solid organ transplant recipients to immunization against influenza virus. J Pediatr 1995;127:957–960. 162. Miller R, Naftel DC, Bourg RC, et al. Infection after heart transplantation: a multi-institutional study. J Heart Lung Transplant 1994;13:381–393. 163. Singh N, Gayowski T, Wagener MM, et al. Bloodstream infections in liver transplant recipients receiving tacrolimus (FK506): changing pattern of microbial origin. Clin Transplantation 1997;11:275–281. 164. Wade JJ, Rolando N, Hallar K, et al. Bacterial and fungal infections after liver transplantation: an analysis of 284 patients. Hepatology 1995;21:1328–1366. 165. Lumbreras C, Fernandez I, Velosa J, et al. Infectious complications following pancreatic transplantation: incidence, microbiological, clinical characteristics and outcome. Clin Infect Dis 1995;20:514–520. 166. Paterson DL, Dominguez E, Chang FY, et al. Infective endocarditis in solid organ transplant recipients. Clin Infect Dis 1998;26:689–694. 167. Paterson DL, Rihs JD, Squier C, et al. Lack of efficacy of mupirocin in the prevention of with Staphylococcus aureus in liver liver transplant recipients and candidates. Transplantation 2003;75:194–198. 168. Patel R, Badley AD, Larson-Keller J, et al. Relevance and risk factors of enterococcal bacteremia following liver transplantation. Transplantation 1996;61:1192–1197. 169. Newell KA, Millis JM, Arnow PM, et al. Incidence and outcome of infection by vancomycin-resistant Enterococcus following orthotopic liver transplantation. Transplantation 1998;65:439–442. 170. Linden PK, Pasculle AW, Manez R, et al. Differences in outcome for patients with bacteremia due to vancomycin-resistant Enterococcus faecium or vancomycin susceptible E. faecium. Clin Infect Dis 1996;22:663–670. 171. Kusne S, Molmenti E, Krystofiak S, et al. Risk factors associated with vancomycin resistant Enterococcus faecium (VREF) bacteremia in liver transplant recipients (Abstract J-37). Thirty-seventh Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 1997. 172. Papanicolaou GA, Myers BR, Meyers J, et al. Nosocomial infections with vancomycin-resistant Enterococcus faecium in liver transplant recipients: risk factors for acquisition and mortality. Clin Infect Dis 1996;23:760–766. 173. Chia JKS, Nakata M, Park SS, et al. Use of bacitracin therapy for infection due to vancomycin-resistant Enterococcus faecium. Clin Infect Dis 1995;21:1520. 174. Aguado JM, Herrero JA, Gavalda J, et al. Clinical presentation and outcome of tuberculosis in kidney, liver, and heart transplant recipients in Spain. Transplantation 1997;63:1278–1286. P.1008 175. Ridgeway AL, Warner GS, Phillips P, et al. Transmission of mycobacterium tuberculosis to recipients of single lung transplants from the same donor. Am J Respir Crit Care Med 1996;153:1166–1168. 176. Kiuchi T, Tanaka K, Inomata Y, et al. Experience of tacrolimus-based immunosuppression in living-related liver transplantation complicated with graft tuberculosis: interaction with rifampicin and side effects. Transplant Proc 1996;28:3171–3172. 177. Singh N, Wagener MM, Gayowski T. Safety and efficacy of isoniazid chemoprophylaxis administered during liver transplant candidacy for the prevention of posttransplant tuberculosis. Transplantation 2002;74:892–895. 178. Chaparro SV, Montoya JG, Keeffe EB, et al. Risk of tuberculosis in tuberculin skin test-positive liver transplant patients. Clin Infect Dis 1999;29:207–208. 179. Benito N, Sued O, Moreno A, et al. Diagnosis and treatment of latent tuberculosis infection in liver transplant recipients in an endemic area. Transplantation 2002;74:1381–1386. 180. Marrie TJ, Johnson WM, Tyler SD, et al. Genomic stability of Legionella pneumophila isolates recovered from two cardiac transplant patients with nosocomial Legionnaires' disease. J Clin Microbiol 1994;3085–3087. 181. Phillips SJ, Zeff RH, Gervick D. Legionnaires' disease. Ann Thorac Surg 1987;44:564. 182. Singh N, Muder RR, Yu VL, et al. Legionella infection in liver transplant recipients: implications for management. Transplantation 1993;15:1549–1551. 183. Sahathevan M, Harvey FAH, Forbes G, et al. Epidemiology, bacteriology and control of an outbreak of Nocardia asteroides infection on a liver transplant unit. J Hosp Infect 1991;18:473–480. 184. Stevens DA, Pier AC, Beaman BL, et al. Laboratory evaluation of an outbreak of nocardiosis in immunocompromised hosts. Am J Med 1981;71:928–934. 185. Hellyar AG. Experience with Nocardia asteroides in renal transplant recipients. J Hosp Infect 1988;12:13–18. 186. Baddour LM, Baselski VS, Herr MJ, et al. Nocardiosis in recipients of renal transplants: evidence for nosocomial acquisition. Am J Infect Control 1986;14:214–219. 187. Horvath J, Dummer S, Lloyd J, et al. Infection in the transplanted and native lung after single lung transplantation. Chest 1993;104:681–685. 188. Yeldandi V, McCabe MA, Larson R, et al. Aspergillus and lung transplantation. J Heart Lung Transplant 1995;14:883–890. 189. Westney GE, Kesten S, De Hoyos A, et al. Aspergillus infection in single and double lung transplant recipients. Transplantation 1996;61:915–919. 190. Paradowski LJ. Saprophytic fungal infections and lung transplantation—revisited. J Heart Lung Transplant 1997;16:524–531. 191. Husni R, Gordon S, Quereshi M, et al. Risk factors for invasive aspergillosis in lung transplant recipients (abstract 33). In program and abstracts of 34th annual meeting of the Infectious Diseases Society of American. New Orleans, 1996. 192. Kanj SS, Tapson V, Davis D, et al. Infections in patients with cystic fibrosis following lung transplantation. Chest 1997;112:924–930. 193. Castaldo P, Stratta RJ, Wood RP, et al. Clinical spectrum of fungal infections after orthotopic liver transplantation. Arch Surg 1991;126:149–156. 194. Briegel J, Forst H, Spill B, et al. Risk factors for systemic fungal infections in liver transplant recipients. Eur J Clin Microbiol Infect Dis 1995;14:375–382. 195. Singh N, Arnow PM, Bonham A, et al. Invasive aspergillosis in liver transplant recipients in the 1990s. Transplantation 1997;64:716–720. 196. Collins LA, Samore MH, Roberts MS, et al. Risk factors for invasive fungal infections complicating orthotopic liver transplantation. J Infect Dis 1994;170:644–652. 197. Kusne S, Torre-Cisneros J, Manez R, et al. Factors associated with invasive lung aspergillosis and significance of positive cultures after liver transplantation. J Infect Dis 1992;166:1379. 198. Munoz P, Torre J, Bouza E, et al. Invasive aspergillosis in transplant recipients. A large multicentric study. In: Program and abstracts of the 36th Interscience Conference on Antimicrobial Agents and Chemotherapy (New Orleans). Washington DC: American Society for Microbiology, 1996. 199. Faggian G, Livi U, Bortolotti U, et al. Itraconazole therapy for acute invasive pulmonary aspergillosis in heart transplantation. Transplant Proc 1989;21:2506–2507. 200. Rabito FJ, Pankey GA. Infections in orthotopic heart transplant patients at the Ochsner Medical Institutions. Med Clin North Am 1992;76:1125–1134. 201. Podzimkova M, Gebauerova M, Malek I, et al. Infectious complications in patients after heart transplantation. Cor et Vasa 1993;35:263–266. 202. Brown RS Jr, Lake JR, Katzman BA, et al. Incidence and significance of Aspergillus cultures following liver and kidney transplantation. Transplantation 1996;61:666–669. 203. Cofan F, Inigo P, Ricart MJ, et al. Invasive aspergillosis following kidney and kidney-pancreas transplantation. Nefrologia 1996;16:253–260. 204. Paterson DL. Invasive aspergillosis. In: Yu VL, Merrigan T, eds. Antimicrobial therapy and vaccines. Baltimore: Williams & Wilkins, 1998. 205. Verweij PE, Dompeling EC, Donnelly JP, et al. Serial monitoring of Aspergillus antigen in the early diagnosis of invasive aspergillosis. Preliminary investigations with two examples. Infection 1997;25:86–89. 206. Swanink CMA, Meis JFGM, Rijs AJMM, et al. Specificity of a sandwich enzyme-linked immunosorbent assay for detecting Aspergillus galactomannan. J Clin Microbiol 1997;35:257–260. 207. Heussel CP, Kauczoe HU, Heussel G, et al. Early detection of pneumonia in febrile neutropenic patients: use of thin-section CT. Radiology 1997;169:1347–1353. 208. Singh N, Mieles L, Yu VL, et al. Invasive aspergillosis in liver transplant recipients: association with candidemia, consumption coagulopathy, and failure of prophylaxis with low-dose amphotericin B. Clin Infect Dis 1993;17:906–908. 209. Reichenspurner H, Gamberg P, Nitschke M, et al. Significant reduction in the number of fungal infections after lung, heart-lung, and heart transplantation using aerosolized amphotericin B prophylaxis. Transplant Proc 1997;29:627–628. 210. Tollemar J, Ericzon BG, Barkhold L, et al. Risk factors for deep candida infections in liver transplant recipients. Transplant Proc 1990;22:1826–1827. 211. Patel R, Portelar D, Bradley AD, et al. Candida and non-Candida fungal infections after liver transplantation. Transplantation 1996;62:926–934. 212. Lumbreras C, Cuervas-Mons V, Jara P, et al. Randomized trial of fluconazole versus nystatin for the prophylaxis of Candida infection after liver transplantation. J Infect Dis 1996;174:583–588. 213. Fortun J, Lopez-San Roman A, Velasco JJ, et al. Susceptibility of Candida glabrata strains with reduced susceptibility of azoles in four liver transplant patients with invasive candidiasis. Eur J Clin Microbiol Infect Dis 1997;16:314–318. 214. Bensousan T, Garo B, Islam S, et al. Possible transfer of Pneumocystis carinii between kidney transplant recipients. Lancet 1990;336:1066–1067. 215. Branten AJ, Beckers PJ, Tiggeler RG, et al. Pneumocystis carinii pneumonia in renal transplant recipients. Nephrol Dial Transplant 1995;10:1194–1197. 216. Chave JP, David S, Wauters JP, et al. Transmission of Pneumocystis carinii from AIDS patients to other immunosuppressed patients: a cluster of Pneumocystis carinii pneumonia in renal transplant recipients. AIDS 1991;5:927–932. 217. Hennequin C, Page B, Roux P, et al. Outbreak of Pneumocystis carinii pneumonia in a renal transplant unit. Eur J Clin Microbiol Infect Dis 1995;14:122–126. 218. Gryzan S, Paradis IL, Zeevi A, et al. Unexpectedly high incidence of Pneumocystis carinii infection after lung-heart transplantation. Implications for lung defense and allograft survival. Am Rev Respir Dis 1988;137:1268–1274. P.1009 219. Arend SM, Westendorp RGJ, Kroon FP, et al. Rejection treatment and cytomegalovirus infection as risk factors for Pneumocystis carinii pneumonia in renal transplant recipients. Clin Infect Dis 1996;22:920–925. 220. Paterson D, Singh N. Pulmonary mycoses in solid-organ transplant recipients. In: Sarosi GA, Davies SF, eds. Fungal infections of the lung. Baltimore: Lippincott Williams & Wilkins, 2000:239–269. 221. LaRosa S, Gordon S, Kalmadi S, et al. Should prophylaxis for Pneumocystis carinii pneumonia (PCP) in solid organ transplant recipients ever be discontinued? (abstract 230) Thirty-fourth annual meeting of the Infectious Diseases Society of America, New Orleans, 1996. 222. Singh N, Gayowski T, Wagener MM, et al. Clinical spectrum of invasive cryptococcosis in liver transplant recipients receiving tacrolimus. Clin Transplant 1997;11:66–70. 223. Davies SF, Sarosi GA, Peterson PK. Disseminated histoplasmosis in renal transplant recipients. Am J Surg 1979;137:686–691. 224. Wheat LJ, Smith EJ, Sathapatayavongs B, et al. Histoplasmosis in renal allograft recipients. Two large urban outbreaks. Arch Intern Med 1983;143:703–707. 225. Wong SY, Allen DM. Transmission of disseminated histoplasmosis via cadaveric renal transplantation: case report. Clin Infect Dis 1992;14:232–234. 226. Hall KA, Sethi GK, Rosado LJ, et al. Coccidioidomycosis and heart transplantation. J Heart Lung Transplant 1993;188:491–497. 227. Cohen IM, Galgiani JN, Potter D, et al. Coccidioidomycosis in renal replacement therapy. Arch Intern Med 1982;142:489–494. 228. Holt CD, Winston DJ, Kubak B, et al. Coccidioidomycosis in liver transplant patients. J Infect Dis 1997;24:216–221. 229. Vartivarian SE, Coudron PE, Markowitz SM. Disseminated coccidioidomycosis. Unusual manifestations in a cardiac transplantation patient. Am J Med 1987;83:949–952. 230. Butka BJ, Bennett SR, Johnson AC. Disseminated inoculation blastomycosis in a renal transplant recipient. Am Rev Respir Dis 1984;130:1180–1183. 231. Singh N, Gayowski T, Yu VL. Invasive gastrointestinal zygomycosis in a liver transplant recipient: case report and review of zygomycosis in solid-organ transplant recipients. Clin Infect Dis 1995;20:617–620. 232. Luft BJ, Noat Y, Araujo GG, et al. Primary and reactivated Toxoplasma infection in patients with cardiac transplants. Ann Intern Med 1983;99:27–33. 233. Shepp DH, Hackman RC, Conley FK, et al. Toxoplasma gondii reactivation identified by detection of parasitemia in tissue culture. Ann Intern Med 1985;103:218–221. 234. Singh N, Gayowski T, Wagener MM, et al. Pulmonary infiltrates in liver transplant recipients in the intensive care unit. Transplantation 1999;67:1138–1144. 235. Paterson DL, Singh N, Gayowski T, et al. Pulmonary nodules in liver transplant recipients. Medicine 1998;77:50–58. 236. Wagener MM, Yu VL. Bacteremia in transplant recipients: a prospective study of demographics, etiologic agents, risk factors, and outcomes. Am J Infect Control 1992;20:239–247. 237. Bonham CA, Dominguez EA, Fukui MB, et al. Central nervous system lesions in liver transplant recipients: prospective assessment of indications for biopsy and implications for management. Transplantation 1998;66:1596–1604. 238. Selby R, Ramirez CB, Singh R, et al. Brain abscess in solid organ transplant recipients receiving cyclosporine-based immunosuppression. Arch Surg 1997;132:304–310.
Воскресенье, 06 Ноябрь 2011 | 1282 Просмотры | Печать | PDF
Vascular catheters have become an indispensable part of modern medicine, allowing administration of medications, fluids, electrolytes, blood products, and nutritional therapy. Specialized functions include hemodynamic monitoring, hemodialysis, and plasmapheresis. Vascular catheters are inserted into more than half of all patients admitted to hospitals in both the United States and Europe and are increasingly being used in the outpatient setting as well (1). In the United States alone, more than 150 million catheters are sold each year (2). About 3 million central venous catheters (CVCs) are inserted into patients in the United States each year (3). Although these catheters provide lifesaving therapy, they can also provide a route for microorganisms to bypass normal host defenses and cause serious infection. Infections related to short-term vascular catheters are the focus of this chapter. DEFINITIONS Colonization of an indwelling vascular catheter refers to significant growth of microbes on either the endoluminal surface or the external catheter surface beneath the skin. Colonization is confirmed by the results of semiquantitative culture of catheter segments growing at least 15 colony-forming units (CFUs) (4) or quantitative cultures growing at least 100 CFUs (5). For this purpose a catheter segment has usually consisted of either the distal 5 cm of the catheter (tip) or the 5-cm segment just beneath the skin (subcutaneous segment). Local catheter-related infection is usually manifested clinically by local inflammation, which may include erythema, tenderness, warmth, and/or a purulent discharge from the catheter tract. If a culture is performed, the catheter usually is significantly colonized at the time of a local catheter-related infection. Exit-site infection is a type of local catheter-related infection of a tunneled catheter where it exits from the skin. This has been defined clinically by the occurrence of erythema and tenderness within 2 cm of the exit site of such a catheter or by the discharge of pus at the exit site (6). A catheter segment culture would usually show colonization if performed. Tunnel infection has been defined as the presence of erythema, tenderness, and induration extending along the tunnel of a tunneled catheter more than 2 cm from the exit site (6). Local signs of inflammation over the tunneled portion of the catheter may or may not be accompanied by signs of inflammation and/or purulent discharge at the exit site. Catheter-related bloodstream infection has been defined as the isolation of the same microbe from blood cultures that is shown to be significantly colonizing the catheter of a patient with clinical features of bloodstream infection (most often manifested by fever alone but also possibly including hypothermia, rigors, hypotension, tachypnea, tachycardia, and confusion) in the absence of any other local infection caused by the same microbe that could have given rise to bloodstream infection (4). Primary bloodstream infection is one that arises without apparent local infection elsewhere resulting from the same microbe. Vascular catheter-related bloodstream infection is considered to be one type of primary bloodstream infection. PATHOGENESIS Infection of short-term catheters has most often been due to microbes from the skin moving along the catheter surface where the catheter enters the skin (789). Progression of these invading microbes distally along the catheter tract can result in bloodstream infection with or without obvious local catheter-related infection. Experiments involving catheter placement in a guinea pig skin model have suggested that microbes can move rapidly along the length of a catheter placed under the skin, perhaps by capillary action (10). By contrast, infection related to long-term catheters (i.e., those in place for more than 3 weeks) is more often due to microorganisms gaining access to the catheter through the catheter hub and then moving down the endoluminal surface of the catheter to the bloodstream (11, 12, 13). Scanning electron microscopy of the external and endoluminal P.232 surfaces of catheters after varying durations of placement has supported this hypothesis by showing more microbes on the external than endoluminal surfaces of catheters that had been in place for less than 10 days and the opposite for catheters that had been in place for more than 30 days (13) (see also Chapter 18). CLINICAL MANIFESTATIONS Local catheter-related infection is manifested as described previously. Clinical manifestations of bloodstream infections may include fever, rigors, hypothermia, tachypnea, hypotension, septic shock, and/or confusion. In the neonate, apnea and/or bradycardia may be prominent symptoms. The most frequent presentation for patients with catheter-related bloodstream infections is simply fever developing in a patient with a catheter in place in the absence of symptoms suggesting infection at another body site. Such a presentation has high positive predictive value for catheter-related bloodstream infections in relatively healthy outpatients with an indwelling catheter. By contrast, among patients in the intensive care unit who develop fever in the presence of an indwelling catheter, 75% to 88% of patients do not have a catheter-related infection (14, 15, 16, 17). In 70% of bloodstream infections related to CVCs, there are no local signs of inflammation around the catheter (18). By contrast, with bloodstream infection resulting from peripheral venous catheters, inflammation or purulent discharge is usually evident at the site of the catheter (19). Most inflammation at the site of peripheral intravenous catheters, however, is due to bland physicochemical thrombophlebitis not infection (20). This was found to remain true among patients with Staphylococcus aureus nasal colonization (21). Such phlebitis occurs in about 30% of patients after only 2 to 3 days of infusion therapy. Principal causes of this noninfectious phlebitis include the catheter material, inexperience of the person inserting the catheter, duration of catheter placement, the rate of infusion, and the pH and chemical composition of the infusate. A high infusion rate results in a higher risk of phlebitis, whereas a heparin lock infusing nothing is associated with a very low risk for phlebitis. Hypertonic solutions; those with low pH; and infusions containing potassium chloride or antibiotics such as vancomycin, metronidazole, or amphotericin B are more likely to cause phlebitis (20). Catheter-related septic phlebitis is associated with local inflammation and induration over the involved vein in most but not all cases involving a peripheral intravenous catheter (22, 23, 24, 25). In cases involving a CVC, septic phlebitis is not usually associated with signs of inflammation or of venous obstruction; these are found in only a small proportion of radiographically confirmed cases of septic thrombophlebitis. The diagnosis of septic thrombophlebitis should be suspected in patients remaining febrile and/or bacteremic despite appropriate antimicrobial therapy with drugs to which the causative microbes are susceptible. Infective arteritis resulting from an arterial catheter is usually manifested by fever and inflammation at the site of the catheter. As the lesion develops, a pulsatile mass may become palpable. In some patients, Osler nodes, petechiae, purpura, and/or septic arthritis may occur distal to the infected aneurysm (26). Femoral neuropathy and arterial insufficiency to the leg have been reported with infection resulting from a femoral artery catheter (22, 23, 24, 25, 26). Endocarditis is usually suspected clinically when a febrile patient is noted to have a new or changing murmur, splenomegaly, or embolic lesions, but such classic manifestations of endocarditis are often absent (27). Several studies suggested that nosocomial bloodstream infections and those associated with a primary focus of infection such as a vascular catheter were less likely to develop endocarditis (27, 28, 29, 30). New criteria for diagnosing endocarditis using echocardiography (31) were used in a recent study, which performed transthoracic and transesophageal echocardiography on all cases and found that 16 (23%) of 69 patients with catheter-related S. aureus bloodstream infection had developed endocarditis (32). The major criteria for diagnosing endocarditis included (a) blood culture results suggestive of endocarditis because of the species involved or the continuousness of the bacteremia and (b) evidence of endocardial involvement by echocardiography (31). The sensitivity of transthoracic echocardiography for diagnosing these cases was only 27% (32). LABORATORY DIAGNOSIS Sixteen different methods and 17 different variations of these methods have been proposed for the laboratory diagnosis of catheter-related bloodstream infection, but relatively few studies have examined the relative accuracy and cost-effectiveness of these different methods. A recent study found that there were sufficient publications regarding six of the different methods to allow pooling of data for a meta-analysis (33). These six methods included three methods for catheter segment culture and three methods for culture of blood drawn through the catheter. The catheter segment methods involve (a) qualitative culture of the catheter segment (i.e., dropping the segment into a tube of broth, incubating for 2 to 3 days, and identifying any microbe that grows), (b) semiquantitative culture of the segment (rolling the segment across the top of a sheep blood agar plate four times, incubating the plate for 2 to 3 days, enumerating any colonies, and identifying the microbes) (34), and (c) quantitative culture of the segment (dropping the segment into a tube containing a milliliter of culture broth and sonicating or vortexing to remove microbes from the catheter surfaces, performing serial dilutions of the resulting suspension, plating aliquots of these dilutions, incubating for 2 to 3 days, enumerating any colonies, and identifying the microbes). The first of the three catheter blood culture methods involved qualitative culture of blood drawn through a suspect catheter (i.e., colonies were not counted). The other two methods both involved quantitative cultures of blood (i.e., serial dilutions of the original specimen were made and plated allowing precise enumeration of the concentration of microbes circulating in the blood at the time the specimen was taken). One of these two methods simply enumerates the number of CFUs per milliliter in blood drawn through the catheter and attributes the infection to an infected catheter if the concentration is higher than a certain threshold level; most of the studies used 100 CFU/mL as their threshold. The remaining method compared the result of a quantitative blood culture drawn through the catheter with one drawn percutaneously from a peripheral vein, reasoning that P.233 a higher concentration in blood from the catheter implies a catheter origin for the infection. Each of the five studies used a different threshold for diagnosing catheter involvement (one a 30-CFU absolute difference, the second a 3:1 ratio, the third a 4:1 ratio, and the fifth a 5:1 ratio). A more recent study of the paired quantitative blood culture method has advocated using a threshold of 8:1 (14). Quantitative catheter segment culture was the only method of the six evaluated in the meta-analysis that was associated with a pooled sensitivity greater than 90% and a pooled specificity greater than 90% (94% and 92%, respectively) (33). By comparison, the test most often used by hospital laboratories in the United States, the semiquantitative catheter segment culture method, had a pooled sensitivity of 85% and a pooled specificity of 85%. Receiver operating characteristic (ROC) curve analysis found a significant increase in accuracy with increasing quantitation of the method for catheter segment culture (p = .03). Analysis of costs showed that the semiquantitative culture was the cheapest of the catheter segment culture methods for the laboratory to perform, at $38.63 per test. It also offered a marginally lower cost per accurate result than did the quantitative catheter segment culture ($401.38 compared with $415.62) (33). Catheter blood culture methods showed a trend toward increasing accuracy with increasing quantitation of the method, but this did not reach statistical significance (33). This could be due to the fact that fewer and smaller studies had been performed with these methods, resulting in lower statistical power for detecting a significant difference among methods despite pooling of available data. Quantitative catheter blood culture appeared to be less sensitive albeit more specific than the catheter segment culture methods. Sensitivity of the quantitative catheter blood culture appeared to be lower in studies focusing on short-term catheters. Nevertheless, the unpaired quantitative catheter blood culture, which had a pooled sensitivity of 78%, offered the lowest cost per accurate test result ($198.18) (33). Thus, it would be preferred when evaluating for bloodstream infection in a febrile patient with a long-dwelling tunneled catheter. A new approach to diagnosing catheter-related bloodstream infection has compared time to positivity for blood drawn from a catheter with that drawn percutaneously from a peripheral vein. When blood drawn from the catheter turns culture positive more than 2 hours earlier than blood drawn from a peripheral vein, this has been shown to suggest that the catheter is the source of infection. To date, the published studies have shown this to work with long-dwelling catheters, and it remains unclear that it will work as well with short-term catheters (34, 35). INCIDENCE Studies of the risk of catheter-related bloodstream infection with peripheral intravenous catheters have documented rates less than 1 per 1,000 catheter insertions and less than 1 per 1,000 catheter-days (36, 37, 38, 39). Early data had found arterial catheter-related bloodstream infection after 4% of arterial catheter insertions (40), but more recent studies have found a much lower incidence. Three studies reported no episodes of arterial catheter-related bloodstream infection after 639 arterial catheter insertions (41, 42, 43). By contrast, the risk of bloodstream infection with short-term CVCs ranges from 1% to 10% of catheter insertions (2). The incidence of primary bloodstream infections during the 1980s was 0.28 per 100 discharges among patients in hospitals reporting data to the National Nosocomial Infections Surveillance (NNIS) system (44). During that decade, the rate of primary bloodstream infections increased by 70% in large teaching hospitals and by 279% in small nonteaching hospitals. Although data regarding the frequency of use of CVCs within these hospitals was not available, it is believed that increased use of such devices contributed to this large increase in catheter-related infection. A 7-month study of the frequency of use and complications of CVC at a university hospital during 1995 documented the use of 2,806 catheters in 1,393 hospitalized adult patients (10% of all adult patients in the hospital during the study period) (45). These catheters were in place for 22,369 catheter-days, which involved 31% of all adult patient-days, including 69% of all patient days in intensive care units (ICUs) and 22% of patient-days on other adult wards. Short-term CVCs accounted for 71% of all central catheters used and 13,709 catheter-days (61% of catheter-days) (45). The overall rates of bloodstream infection during the 7-month study were 1.4 per 100 catheters inserted and 1.7 per 1,000 CVC-days. The rates were similar for patients in ICUs and hospital wards [1.7 vs. 1.6 per 1,000 CVC days, relative risk (RR) = 1.07, p = .98]. The rate of infection observed in this study was also similar to the rate observed during a randomized trial of catheter management in ICUs in this hospital several years before (1.5 per 100 catheter insertions) (46). OUTCOMES The case fatality rate associated with catheter-related bloodstream infection was found to be 14% in a meta-analysis that evaluated publications reporting 2,573 cases of catheter-related bloodstream infection and information regarding outcome (47). The authors of the original publications reported that 11.3% of patients with bloodstream infections died because of their underlying illness, accounting for 81% of the deaths in patients with catheter-related bloodstream infections. The authors of the original publications believed that 2.7% [95% confidence interval (CI): 2.0% to 3.4%] of patients with catheter-related bloodstream infection died because of the infection itself, accounting for 19% of deaths following the infections (47). Adverse prognostic factors identified in the meta-analysis included neutropenia, bloodstream infection on a surgical service, and bloodstream infection in an ICU. Infections resulting from coagulase-negative staphylococci and enterococci were associated with lower rates of attributed mortality than for other pathogens (0.7% and 0%, respectively) (47). Candida, by contrast, was associated with a higher attributed mortality rate as compared with other pathogens (9%, p = .001) as was S. aureus (8.2%, p < .001). Another review of the outcome of catheter-related S. aureus bacteremias included data from 25 studies and reported 59 deaths among 177 patients for an overall case fatality rate of P.234 33.3% (95% CI: 26.4% to 40.2%) (48). The proportion of cases in which death was attributed to the infection itself by the authors of the original publications was14.8% (95% CI: 10.8% to 18.8%), accounting for 44% of all of the deaths following catheter-related S. aureus bacteremia (48). A recent study found that Candida catheter-related bloodstream infections were associated with a sevenfold higher death rate than for coagulase-negative staphylococci (49). Patients surviving a nosocomial bloodstream infection have remained in the hospital for an extra 7 to 24 days (50, 51). This longer stay and the additional costs of antibiotics and diagnostic tests add significantly to hospital costs. The Centers for Disease Control and Prevention (CDC) estimated from data collected in the Study on the Efficacy of Nosocomial Infection Control (SENIC) conducted in the 1970s and updated in the 1980s that a nosocomial bloodstream infection added $3,517 to the cost of hospitalization (50). In 1988 Maki et al. (52) estimated that this excess cost had risen to $6,000. Pittet et al. (51) measured the costs incurred by patients suffering nosocomial bloodstream infection and control patients (matched for ICU admission, major diagnosis, number of diagnoses, and length of stay until the time of the infection in the case patient) in a surgical ICU between 1988 and 1990. Hospital costs averaged $33,268 more for case patients than for controls and $40,890 more for surviving case patients than for their matched controls. RISK FACTORS The most important risk factor for a catheter-related bloodstream infection in most patients has been the type of catheter used. Although peripheral intravenous catheters are used for most infusion therapy, they account for only a tiny proportion of all catheter-related bloodstream infections. CVCs account for about 2% of all catheters inserted (1, 2) and 97% of all published cases of catheter-related bloodstream infections (47). Several observational studies have suggested a higher rate of infection with multilumen as compared with single-lumen CVCs (53, 54, 55, 56, 57, 58), but randomized trials have produced conflicting results (59, 60). One study involving multilumen catheters used in an ICU reported that only 1.5% of catheterizations resulted in bloodstream infections (46). Total parenteral nutrition (TPN) has been considered a risk factor for catheter-related bloodstream infection, (61) but three studies found no association, suggesting that when managed properly it need not be a risk factor (7, 46, 62). One study found that the use of such catheters for other purposes was associated with a higher risk for infection (63). Another study found an association with infection when TPN catheters were used with a needleless infusion system (64). Infection rates associated with TPN catheters have been lower with the use of a dedicated team for the insertion and maintenance of the catheters (65, 66, 67, 68). A prolonged outbreak of bloodstream infections in an ICU was linked to infection of catheters being used to infuse TPN when the nurse-to-patient ratio fell below a critical level, suggesting that overworked staff may be unable to manage such high-risk devices in an optimal manner (69). A follow-up study confirmed this association between understaffing and risk for infection (70). Hyperglycemia has been correlated with a higher rate of postoperative infections among diabetics including bloodstream infections (71), and hyperglycemia was documented more often in patients randomized to TPN than to enteral nutrition in a randomized trial (72). Candida and staphylococci have been the microbes most often associated with TPN catheter-related infections (17, 63). Lipid infusions used with TPN have been specifically associated with risk of catheter-related bloodstream infections resulting from coagulase negative staphylococci (73, 74) and Malassezia furfur (75). Three recent pediatric studies have also found TPN to be a risk factor for catheter-related bloodstream infection (49, 76, 77), and two found lipid infusions to be risk factors for fungal catheter-related bloodstream infection, one for candidemia (49) and the other for Pichia anomola bloodstream infection during an ICU outbreak (76). Birth weight and use of a Broviac catheter were also important predictors in one of these studies(77). Duration of catheterization is a risk factor for infection, with the cumulative risk of infection increasing linearly as a function of catheter duration (78). The incidence density of infection does not appear to increase after the first few days of catheterization with CVCs used in the ICU (79) or for hemodialysis (80). These data were supported by studies using an experimental catheter infection model in rabbits, which showed that the highest risk of infection occurred during the first 2 days after insertion, with lower daily rates thereafter (81). Acquired immunodeficiency syndrome (AIDS) has been a risk factor for catheter-related bloodstream infection, (82, 83) but this risk appears to be somewhat lower with use of highly active antiretroviral therapy (84). Site selection may be an important risk factor for infection because several observational studies have suggested higher infection rates for jugular than for subclavian catheters (85, 86, 87, 88, 89) and for femoral than subclavian catheters or internal jugular (IJ) catheters (89, 90, 91). Not all observational studies have found the same association, however (92, 93, 94). Two recent randomized controlled trials have shown higher complication rates for femoral catheters than for those in the subclavian or jugular vein (95, 96). These complications included significantly higher rates of deep venous thrombosis in both studies and of catheter-related infection in one (96). Peripherally inserted central catheters (PICCs) have been associated with a very low incidence of bloodstream infection but have mostly been used in outpatients until recently (97, 98). Limited data are available with hospitalized patients (99, 100, 101). Lower infection rates with this approach than with subclavian or internal jugular catheters could be due to lower concentrations of resident flora on the arm than the chest or neck (102, 103). More randomized trials of site selection are needed. Inexperience of the individual inserting the catheter was shown to be significantly associated with risk for infection in a study that showed an inverse correlation between the total number of catheters inserted by the physician inserting the catheter and risk for significant colonization of the catheter (104). Another observational study found a higher rate of infection when only a small sterile towel and sterile gloves were used for insertion P.235 of the catheter as compared with insertion using large sterile drapes while wearing a mask, cap, and sterile gloves and gown (85). These two studies suggest that inadvertent contamination during insertion may be an important risk factor for infection. Several other studies suggest that management of the catheter after insertion may also significantly alter the risk for infection. In a randomized trial that preceded the onset of the use of universal or standard precautions, Klein et al. (105) documented a significantly lower overall rate of nosocomial infections and a trend toward a lower rate of primary bloodstream infections (0.3% vs. 1.3%, p = .08) in a pediatric ICU when gowns and gloves were routinely used for caring for patients than when they were not used. Use of a TPN catheter for other purposes than infusing TPN was found to be a risk factor for catheter-related bloodstream infection in another study referred to previously (63). The association of a higher infection rate of TPN catheters with understaffing referenced previously also suggests risk from faulty manipulation of catheters (69). Guidewire exchanges have been found to be a risk factor in several studies of catheter infection (46, 53, 106, 107, 108). Failure to properly process and disinfect pressure transducers has been shown to correlate with infection in multiple outbreaks in ICUs (109). Transparent dressings have resulted in a significantly higher rate of colonization of catheters than have cotton gauze dressings covered with tape (110, 111). One randomized trial documented a significantly higher rate of bloodstream infection when catheters were covered with transparent dressings (112). Two recent outbreaks of catheter-related bloodstream infection were noted to be associated with cost-cutting measures and reliance on weekly transparent dressing changes (113, 114). Triple antibiotic ointment [polymyxin-neomycin-bacitracin (PNB)] has been associated with a halving of the rate of bacterial colonization of catheters but a fivefold increase in the rate of fungal colonization (115). One study evaluating the use of PNB ointment under a transparent dressing reported that 4 (14%) of 29 patients in a surgical ICU developed candidemia (115). MICROBIAL ETIOLOGY Gram-positive cocci have been the most frequent causes of primary bloodstream infection in hospitals reporting data to the NNIS system. In data collected between October 1986 and December 1990 (116), coagulase-negative staphylococci accounted for 28.2% of cases, followed by S. aureus (16.1%) and enterococci (12.0%) (116). These were followed by Candida spp. (10.2%) and Enterobacter spp. (5.3%). Because NNIS system data include cases diagnosed as bacteremia by physicians even if only one of many blood cultures were positive for a coagulase-negative Staphylococcus, these data probably overestimate the true frequency of bacteremia resulting from these microorganisms. A different approach requiring more convincing proof that coagulase-negative staphylococci actually caused catheter-related bloodstream infection provides a somewhat different relative frequency distribution. When the results of several studies were compiled using stricter definitions of catheter-related bloodstream infection, coagulase-negative staphylococci were still the most frequent cause accounting for 27 of 100 cases, but S. aureus was almost as frequent accounting for 26 of 100 cases (117). These were followed in order by yeast (17 of 100 cases), Enterobacter spp. (7 of 100 cases), Serratia spp. (5 of 100 cases), Enterococcus spp. (5 of 100 cases), Klebsiella spp. (4 of 100 cases), viridans species of Streptococcus (3 of 100 cases), Pseudomonas spp. (3 of 100 cases), Proteus spp. (2 of 100 cases), and Yersinia spp. (1 of 100 cases) (117). Another study found that 32% of colonized catheters were associated with bloodstream infection resulting from the same species (118). The study found that the probability of positive blood cultures associated with a colonized catheter varied depending on the species colonizing the catheter. For example, catheters colonized by Candida albicans were associated with bloodstream infection resulting from the same species in 68.4% of cases, followed by S. aureus in 60%, Enterobacter cloacae in 42.9%, Staphylococcus epidermidis in 32.1%, Pseudomonas aeruginosa in 27.7%, and Enterococcus faecalis in 23.3%. The frequencies of bloodstream infection with C. albicans and with S. aureus significantly exceeded those for catheters colonized by the other species (118). This was confirmed by another study (119) (see also Chapter 19). PREVENTION Prevention of any disease involves avoidance of known risk factors for the disease. For catheter-related infection, this approach might include the following, based on the data discussed in the section on risk factors: (a) selecting a subclavian, basilic, or cephalic vein site rather than an internal jugular or femoral vein site; (b) avoiding the use of TPN catheters for purposes other than infusion of TPN; (c) using a special team for the insertion and maintenance of TPN catheters; (d) avoiding the use of triple antibiotic ointment (PNB) on CVCs (115); (e) using maximal aseptic technique for insertion of the catheter; (f) avoiding high glucose levels in diabetic patients with a CVC by careful regulation of serum glucose concentrations, especially for those receiving TPN; (g) using cotton gauze dressings; (h) avoiding understaffing in the management of patients with CVCs (69, 70, 120); and (i) having an experienced physician insert the catheter. Data from randomized controlled trials provide additional support for some of these approaches. For example, a randomized trial comparing maximal aseptic technique for insertion (large sterile drapes, sterile gowns, and sterile gloves, with mask and cap being worn by the inserting physician as compared with an older approach using a mask, small drape, and sterile gloves but none of the additional barriers) confirmed the importance of using maximal sterile barrier precautions for avoiding inadvertent contamination at the time of insertion (121). Povidone-iodine ointment placed at the catheter site with dressing changes was associated with a significantly lower risk of catheter-related bloodstream infection in a randomized trial involving hemodialysis patients (122). An older study found no benefit with the ointment, however, and further studies are needed (123). Mupirocin ointment has also been found effective in preventing bacterial colonization of catheters (124) but has been associated with P.236 development of mupirocin resistance among skin flora, and recent CDC guidelines do not recommend using mupirocin at the catheter site (125). Scheduled replacement of CVCs was used for decades for prevention of infection based on a theory that inserting a new catheter every few days should lower the patient's risk for infection. Unfortunately the strategy did not work when tested in randomized trials (46, 80, 107, 126, 127, 128). When new site puncture was used, the infection rate was unchanged and the risk of major mechanical complications was significantly increased by scheduled replacement (46). When guidewire exchange was used for scheduled replacement, the rate of infection was paradoxically higher than if catheters were changed only as clinically necessary (46, 107). The current CDC guideline for prevention of vascular device-related infections recommends against scheduled replacement of CVCs (125). The antiseptic used for prepping the site before catheter insertion and with each dressing change was studied in multiple recent trials with all but one of them finding that chlorhexidine gluconate was associated with lower colonization and bloodstream infection rates than were povidone-iodine or alcohol (129, 130, 131). It is possible that tincture of iodine would work well for preventing catheter-related infection, but most healthcare providers have avoided using it for this purpose because of reports of increased skin rash with its use. Two recent studies suggested that it was better than povidone-iodine for antisepsis of skin before drawing percutaneous blood cultures (132, 133), but one of these studies also found that tincture of iodine offered no advantage over 70% isopropyl alcohol, which was less expensive and has been associated with fewer side effects (133). Tunneling of CVCs has been used primarily for long-dwelling catheters, but some have advocated use of tunneling even for short-term catheters in the ICU. One study with nontunneled catheters in cancer patients reported infection rates that were somewhat lower than usually reported with tunneled catheters (97), and two randomized trials found no benefit from tunneling, including one in ICU patients (134, 135). Three other randomized trials showed lower infection rates with tunneled catheters (136, 137, 138). The cost-effectiveness of this approach requires further scrutiny, however. A silver-impregnated collagen cuff placed on the catheter at the time of insertion was found to be effective in preventing bloodstream infection in two randomized trials (52, 115). Catheters with an antiseptic coating have been shown to prevent catheter-related bloodstream infection (89, 139, 140, 141). Antibiotic coating has been shown to work (142, 143, 144, 145, 146), but concern has been raised about promoting the development of antibiotic resistance from exposing the cutaneous flora of large numbers of patients to such surfaces (52, 141, 147). Copper and silver-copper coatings have been shown to decrease adherence of S. aureus to catheters made of silicon rubber, polyvinyl chloride, and Teflon (148). A negatively charged direct current of electricity with 10 ВA reduced bacterial adherence to catheters, but a positively charged current had no effect (149). Current flowing through a silver wire wrapped in a helical fashion around a catheter prevented adherence in an experimental animal model of infection (150). Antibiotic lock therapy has been used to salvage a catheter for continued use after a catheter infection (151, 152, 153, 154, 155, 156, 157, 158, 159). This approach has been modified by using a periodic antimicrobial flush with vancomycin (160), taurolidine (161), or minocycline-ethylenediaminetetraacetic acid (EDTA) (162, 163). Concern has been raised about the use of clinical antibiotics such as vancomycin for this purpose because of the risk of potentiating development of resistance (164). Several recent studies evaluated novel approaches to the prevention of catheter-related infection and found no benefit from use of prophylactic intravenous immunoglobulin infusions for ICU patients in general (165), granulocyte colony stimulating factor in neutropenic ICU patients (166), or hypocaloric TPN for patients requiring TPN (167). REFERENCES 1. Widmer AF. Central venous catheters. In: Seifert H, Jansen B, Farr BM, editors. Catheter-related infections. New York: Marcel Dekker, 1997:183–216. 2. Maki DG. Infections due to infusion therapy. In: Bennet JV, editor. Hospital infections. Boston: Little, Brown and Company, 1992:849. 3. Elliott TSJ, Faroqui MH, Tebbs SE, et al. An audit programme for central venous catheter-associated infections. J Hosp Infect 1995;30:181–191. 4. Maki DG, Weise CE, Sarafini HW. A semiquantitative culture method for identifying intravenous catheter related infection. N Engl J Med 1977;296:1305–1309. 5. Raad I, Sabbagh MF, Rand KH, et al. Quantitative tip culture methods and the diagnosis of central venous catheter-related infections. Diagn Microbiol Infect Dis 1992;15:13–20. 6. Press OW, Ramsey PG, Larson EB, et al. Hickman catheter infections in patients with malignancies. Medicine 1984;63:189–200. 7. Maki DG, Stolz S. The epidemiology of central-venous catheter-related bloodstream infection (BSI)[abstract 67]. In: Programs and abstracts of the 34th Interscience Conference of Antimicrobial Agents and Chemotherapy, Orlando, FL, 1994. 8. Bjornson HS, Colley R, Bower RH, et al. Association between microorganism growth at the catheter insertion site and colonization of the catheter in patients receiving total parenteral nutrition. Surgery 1982;92:720–727. 9. Cheesborough JS, Finch RG, Burden RP. A prospective study of the mechanisms of infection associated with hemodialysis catheters. J Infect Dis 1986;154:579–589. 10. Cooper CL, Schiller AL, Hopkins CC. Possible role of capillary action in pathogenesis of experimental catheter-associated dermal tunnel infections. J Clin Micro 1988;26:8–12. 11. Linares J, Sitges-Serra A, Garau J, et al. Pathogenesis of catheter sepsis: a prospective study with quantitative and semiquantitative cultures of catheter hub and segments. J Clin Micro 1985;21:357–360. 12. Moro ML, Vigano EF, Lepri AC. Risk factors for central venous catheter related infections in surgical and intensive care units. Infect Control Hosp Epidemiol 1994;15:253–264. 13. Raad I, Costerton W, Sabharwal U, et al. Ultrastructural analysis of indwelling vascular catheters: a quantitative relationship between luminal colonization and duration of placement. J Infect Dis 1993;168:400–407. 14. Quilici N, Audibert G, Conroy MC, et al. Differential quantitative blood cultures in the diagnosis of catheter-related sepsis in intensive care units. Clin Infect Dis 1997;25:1066–1070. 15. Maki DG. Nosocomial bacteremia: a epidemiologic overview. Am J Med 1981;70:719–732. 16. Bozzetti F, Terno G, Camerini E, et al. Pathogenesis and predictability of central venous catheter sepsis. Surgery 1982;91:383–389. 17. Ryan JAJ, Abel RM, Abbott WM, et al. Catheter complications in total parenteral nutrition: a prospective study of 200 consecutive patients. N Engl J Med 1974;290:757–761. P.237 18. Pittet D, Chuard C, Rae AC, et al. Clinical diagnosis of central venous catheter line infections: a difficult job [abstract 174]. In: Programs and abstracts of the 31st Interscience Conference of Antimicrobial Agents and Chemotherapy, Chicago, 1991. 19. Watanakunakorn C, Baird IM. Staphylococcus aureus bacteremia and endocarditis associated with a removable infected intravenous device. Am J Med 1977;63:523–526. 20. Maki DG, Ringer M. Risk factors for infusion-related phlebitis with small peripheral venous catheters a randomized controlled trial. Ann Intern Med 1991;114:845–854. 21. Lipsky BA, Peugeot RL, Boyko EJ, et al. A prospective study of Staphylococcus aureus nasal colonization and intravenous therapy-related phlebitis. Arch Intern Med 1992;152:2109–2112. 22. Farr BM. Nonendocardial vascular infections. In: Hoeprich PD, Jordanb MC, Roanld AR, eds. Infectious diseases. Philadelphia: JB Lippincott, 1994:1248–1258. 23. Berkowitz FE, Argent AC, Faise T. Suppurative thrombophlebitis: a serious nosocomial infection. Pediatr Infect Dis J 1987;6:64–67. 24. Johnson RA, Zajac RA, Evan ME. Suppurative thrombophlebitis: correlation between pathogen and underlying disease. Infect Control 1986;7:582–585. 25. Kaufman J, Demas C, Stark K, et al. Catheter-related septic central venous thrombosis: current therapeutic options. West J Med 1986;145:200–203. 26. Cohen A, Reyes R, Kirk M, et al. Osler's nodes, pseudoaneurysm formation and sepsis complicating percutaneous radial artery cannulation. Crit Care Med 1984;12:1078–1079. 27. Bayer AS, Scheld WM. Endocarditis and intravascular infections. In: Mandell G, Bennet JE, Dolin R, ed. Principles and practice of infectious diseases, 5th ed. Philadelphia: Churchill Livingstone, 2000:857–902. 28. Nolan CM, Beaty HN. Staphylococcus aureus bacteremia. Current clinical patterns. Am J Med 1976;60:495–500. 29. Cooper R, Platt R. Staphylococcus aureus bacteremia in diabetic patients. Am J Med 1982;73:658–662. 30. Fowler VG, Olsen M, Corey GR, et al. Predictors of complications in patients with Staphylococcus aureus bacteremia [abstract 415]. In: Abstracts of the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, September and December, 2001. 31. Durack DT, Lukes AS, Bright DK, et al. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994;96:200–209. 32. Fowler VG, LI J, Corey GR, et al. Role of echocardiography in evaluation of patients with Staphylococcus aureus bacteremia: experience in 103 patients. J Am Coll Cardiol 1997;30:1072–1078. 33. Siegman-Igra Y, Anglim AM, Shapiro DE, et al. Diagnosis of vascular catheter-related bloodstream infection: a meta-analysis. J Clin Microbiol 1997;35:928–936. 34. Blot F, Schmidt E, Nitenberg G. Earlier positivity of central venous versus peripheral blood cultures is highly predictive of catheter related sepsis. J Clin Microbiol 1998;36:105–109. 35. Blot F, Nitenberg G, Chachaty E, et al. Diagnosis of catheter-related bacteraemia: a prospective comparison of the time to positivity of central vs. peripheral blood cultures. Lancet 1999;354:1071–1077. 36. Garland JS, Nelson DB, Cheah TE, et al. Infectious complications during peripheral intravenous therapy with Teflon catheters: a prospective study. Pediatr Infect Dis J 1998;6:918–921. 37. Tully JL, Griedland GH, Baldini LM, et al. Complications of intravenous therapy with steel needles and Teflon catheters a comparative study. Am J Med 1981;70:702–706. 38. Tager IB, Ginsberg MB, Ellis SE. An epidemiologic study of the risks associated with peripheral intravenous catheters. Am J Epidemiol 1983;118:839–851. 39. Maki DG, Botticelli JT, LeRoy ML, et al. Prospective study of replacing administration sets for intravenous therapy at 48 vs 72-hour intervals: 72 hours is safe and cost effective. JAMA 1987;258:1777–1781. 40. Band JD, Maki DG. Infections caused by arterial catheters used for hemodynamic monitoring. Am J Med 1979;67:735–741. 41. Norwood SH, Cormier B, McMahon NG, et al. Prospective study of catheter-related infection during prolonged arterial catheterization. Crit Care Med 1988;16:836–839. 42. Leroy O, Billiau V, Beuscart C, et al. Nosocomial infections associated with long-term radial artery cannulation. Intens Care Med 1989;15:241–246. 43. Furfaro S, Gauthier M, Lacroix J, et al. Arterial catheter-related infections in children. Am J Dis Child 1991;145:1037–1042. 44. Bannerjee SN, Emori TG, Culver DH. Secular trends in nosocomial primary bloodstream infections in the United States., 1980–1989. Am J Med 1991;91(Suppl 3B):87S–89S. 45. Byers KE, Anneski CJ, Anglim AM, et al. Central venous catheter utilization and complications at a university hospital [abstract]. Presented at the seventh annual meeting of the Society of Hospital Epidemiology of America, St. Louis, 1997. 46. Cobb DK, High KP, Sawyer RG, et al. A controlled trial of scheduled replacement of central venous and pulmonary-artery catheters. N Engl J Med 1992;327:1062–1068. 47. Byers KE, Adal KA, Anglim AM, et al. Case fatality rate for catheter-related bloodstream infections (CRBSI): a meta-analysis. Infect Control Hosp Epidemiol 1995;16(Part 2, Suppl):23. 48. Jernigan JA, Farr BM. Short-course therapy of catheter-related Staphylococcus aureus bacteremia: a meta-analysis. Ann Intern Med 1993;119:304–311. 49. Saiman L, Ludington E, Pfaller M. Risk factors for candidemia in neonatal intensive care unit patients. The National Epidemiology of Mycosis Survey study group. Pediatr Infect Dis J 2000;19:319–324. 50. Centers for Disease Control and Prevention. Public health focus: surveillance, prevention, and control of nosocomial infections. MMWR 1992;41:783–787. 51. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, and attributable mortality. JAMA 1994;271:1598–1601. 52. Maki DG, Cobb L, Garman JK, et al. An attachable silver-impregnated cuff for prevention of infection with central venous catheters: a prospective randomized multicenter trial. Am J Med 1988;85:307–314. 53. Hilton E, Haslett TM, Borenstein MT, et al. Central catheter infections: single-versus triple-lumen catheters: influence of guide wires on infection rates when used for replacement of catheters. Am J Med 1988;84:667–672. 54. Yeung C, May J, Hughes R. Infection rate for single lumen v triple lumen subclavian catheters. Infect Control Hosp Epidemiol 1988; 9:154–158. 55. Gil RT, Kruse JA, Thill-Beharozian MC, et al. Triple-vs. single-lumen central venous catheters: a prospective study in a critically ill population. Arch Intern Med 1989;149:1139–1143. 56. Miller JJ, Venus B, Mathru M. Comparison of the sterility of long-term central venous catheterization using single lumen, triple lumen, and pulmonary catheters. Crit Care Med 1984;12:634–637. 57. Lee RB, Buckner M, Sharp KW. Do multi-lumen catheters increase central venous catheter sepsis compared to single-lumen catheters? J Trauma 1988;28:1472–1475. 58. Pemberton LB, Lyman B, Lander V, et al. Sepsis from triple- vs. single lumen catheters during total parenteral nutrition in surgical or critically ill patients. Arch Surg 1986;121:591–594. 59. Clark-Christoff N, Watters VA, Sparks W, et al. Use of triple-lumen subclavian catheters for administration of total parenteral nutrition. J Parenteral Enteral Nutr 1992;16:403. 60. Farkas JC, Liu N, Bleriot JP, et al. Single-versus triple-lumen central catheter related sepsis: a prospective randomized study in a critically ill population. Am J Med 1992;93:277. 61. Safdar N, Kluger D, Maki DG. A review of risk factors for catheter-related bloodstream infection caused by percutaneously inserted, noncuffed central venous catheters: implications for preventive strategies. Medicine 2002;81:466–479. 62. Howell PB, Walters PE, Donowitz GR, et al. Risk factors for infection of adult patients with cancer who have tunnelled central venous catheters. Cancer 1995;75:1367–1375. 63. Snydman DR, Murray SA, Kornfeld SJ, et al. Total parenteral nutrition related infections. Am J Med 1982;73:695–699. P.238 64. Danzig L, Short L, Collins K, et al. Bloodstream infections (BSIs) associated with a needleless intravenous (IV) infusion system used in patients on home infusion therapy [abstract 195]. In: Programs and abstracts of the 34th Interscience Conference of Antimicrobial Agents and Chemotherapy, Orlando, FL, 1994. 65. Faubion WC, Wesley JR, Khalidi N, et al. Total parenteral nutrition catheter sepsis: impact of the team approach. J Parenteral Enteral Nutr 1986;10:642–645. 66. Freeman JB, Lemire A, Maclean LD. Intravenous alimentation and septicemia. Surg Gynecol Obstet 1972;135:708–712. 67. Nehme AE. Nutritional support of the hospitalized patient. The team concept. JAMA 1980;243:1906–1908. 68. Nelson DB, Kien CL, Mohr B, et al. Dressing changes by specialized personnel reduce infection rates in patients receiving central venous parenteral nutrition. J Parenteral Enteral Nutr 1986;10:220–222. 69. Fridkin SK, Pear SM, Williamson TH, et al. The role of understaffing in central venous catheter-associated bloodstream infections. Infect Control Hosp Epidemiol 1996;17:150–158. 70. Robert J, Fridkin SK, Blumberg HM. The influence of the composition of the nursing staff on primary bloodstream infection rates in a surgical intensive care unit. Infect Control Hosp Epidemiol 2000;21:12–17. 71. Baxter JK, Babineau TJ, Apovian CM, et al. Perioperative glucose control predicts increased nosocomial infection in diabetics. Crit Care Med 1990;18:S207. 72. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. Perioperative total parenteral nutrition in surgical patients. N Engl J Med 1991;325:525. 73. Shiro H, Muller E, Takeda S, et al. Potentiation of Staphylococcus epidermidis catheter-related bacteremia by lipid infusions. J Infect Dis 1995;171:220–224. 74. Avila-Figueroa C, Goldmann DA, Richardson DK, et al. Intravenous lipid emulsions are the major determinant of coagulase-negative staphylococcal bacteremia in very low birth weight newborns. Pediatr Infect Dis 1998;17:10–17. 75. Barber GR, Brown AE, Kiehn TE, et al. Catheter-related Malassezia furfur fungemia in immunocompromised patients. Am J Med 1993;95:370. 76. Aragao PA, Oshiro IC, Manrique EI. Pichia anomala outbreak in a nursery: exogenous source? Pediatr Infect Dis 2001;20:843–848. 77. Brodie SB, Sands KE, Gray JE. Occurrence of nosocomial bloodstream infections in six neonatal intensive care units. Pediatr Infect Dis 2000;19:56–65. 78. Widmer AF. Intravenous-related infections. In: Wenzel RP, ed. Prevention and control of nosocomial infections. New York: Williams & Wilkins, 1997:771–805. 79. Stenzel JP, Green TP, Fuhrman BP, et al. Percutaneous central venous catheterization in a pediatric intensive care unit: a survival analysis of complications. Crit Care Med 1989;17:984–988. 80. Uldall PR, Merchant N, Woods F, et al. Changing subclavian haemodialysis cannulas to reduce infection. Lancet 1981;1:1373. 81. Sherertz RJ, Carruth WA, Hu Q, et al. Factors modifying the risk of Staphylococcus aureus infection associated with silicone catheters in a rabbit model [abstract]. Presented at the 30th Interscience Conference of Antimicrobial Agents and Chemotherapy, 1990. 82. Jacobson MA, Gellermann H, Chambers H. Staphylococcus aureus bacteremia and recurrent staphylococcal infection in patients with acquired immunodeficiency syndrome and AlDS related complex. Am J Med 1988;85:172–176. 83. Skoutelis AT, Murphy RL, MacDonell KB, et al. Indwelling central venous catheter infections in patients with acquired immunodeficiency syndrome. J Acquir Immune Defic Syndr 1990;3:335–342. 84. Tumbarello M, Tacconelli E, Donati KG, et al. HIV-associated bacteremia: how it has changed in the highly active antiretroviral therapy (HAART) era. J Acquir Immune Defic Syndr 2000;23:145–151. 85. Mermel LA, McCormick RD, Springman SR, et al. The pathogenesis and epidemiology of catheter related infection with pulmonary artery Swan-Ganz catheters: a prospective study utilizing molecular subtyping. Am J Med 1991;91(Suppl 3B):197S–205S. 86. Richet H, Hubert B, Nitemberg G, et al. Prospective multicenter study of vascular catheter related complications and risk factors for positive central-catheter cultures in intensive care unit patients. J Clin Microbiol 1990;28:2520–2525. 87. Santre CH, Georges H, Jacquier JM, et al. A comparison between semi-quantitative catheter tip cultures and non quantitative blood cultures in the diagnosis of catheter-related infections in critically ill patients [abstract 68]. In: Programs and abstracts of the 34th Interscience Conference of Antimicrobial Agents and Chemotherapy, Orlando, FL, 1994. 88. Hagley MT, Martin BM, Gast P, et al. Infectious and mechanical complications of central venous catheters placed by percutaneous venipuncture and over guidewires. Crit Care Med 1992;20:1426. 89. Norwood SH, Wilkins HE 3rd, Vallins VL, et al. The safety of prolonging the use of central venous catheters: a prospective analysis of the effects of using antiseptic-bonded catheters with daily site care. Crit Care Med 2000;28:1376–1382. 90. Goetz AM, Muder RR, Wagener MM, et al. Risk of infection due to femoral placement of central venous catheters [abstract 258]. In: Programs and abstracts of the 35th Interscience Conference of Antimicrobial Agents and Chemotherapy, San Francisco, 1995. 91. Oliver MJ, Callery SM, Thorpe KE, et al. Risk of bacteremia from temporary hemodialysis catheters by site of insertion and duration of use: a prospective study. Kidney Int 2000;58:2543–2545. 92. Montagnac R, Bernard C, Guillaumie J, et al. Indwelling silicone femoral catheters: experience of three haemodialysis centres. Nephrol Dial Transplant 1997;12:772–775. 93. Williams JF, Seneff MG, Friedman BC, et al. Use of femoral venous catheters in critically ill adults: prospective study. Crit Care Med 1991;19:550–553. 94. Yurtkuran M. Catheterization of the femoral vein for chronic hemodialysis. Angiology 1987;38:847–850. 95. Trottier SJ, Veremakis C, O'Brien J, et al. Femoral deep vein thrombosis associated with central venous catheterization: results from a prospective, randomized trial. Crit Care Med 1995;23:52–59. 96. Merrer J, De Jonghe B, Bernard MD, et al. Complications of femoral and subclavian venous catheterization in critically ill patients: a randomized controlled trial. JAMA 2001;286:700–707. 97. Raad I, Davis S, Becker M, et.al. Low infection rate and long durability of nontunnelled silastic catheters. Arch Intern Med 1993;153:1791–1796. 98. Tice AD, Bonstell RP, Marsh PK, et al. Peripherally inserted central venous catheters for outpatient intravenous antibiotic therapy. Infect Dis Clin Pract 1993;2:186–190. 99. Linblad B, Wolff T. Infectious complications of percutaneously inserted central venous catheters. Acta Anaesthesiol Scand 1985;29:587–589. 100. Lam S, Scannell R, Roessler D, et al. Peripherally inserted central catheters in an acute-care hospital. Arch Intern Med 1994;154:1833–1837. 101. Bottino J, MeCredie K, Grosehel DHM, et al. Long-term intravenous therapy with peripherally inserted silicone elastomer central venous catheters in patients with malignant diseases. Cancer 1979;43:1937–1943. 102. Maki DG. Marked differences in skin colonization of insertion sites for central venous, arterial and peripheral IV catheters. The major reason for differing risks of catheter-related infection? [abstract 205] In: Programs and abstracts of the 30th Interscience Conference of Antimicrobial Agents and Chemotherapy, Atlanta, GA, 1990. 103. Noble WC. Dispersal of skin microorganisms. Br J Dermatol 1975;93:477–485. 104. Armstrong CW, Mayhall CG, Miller KB, et.al. Prospective study of catheter replacement and other risk factors of infection of hyperalimentation catheters. J Infect Dis 1986;154:808–816. 105. Klein BS, Perloff WH, Maki DG. Reduction of nosocomial infection during pediatric intensive care by protective isolation. N Engl J Med 1989;320:1714–1721. 106. Maki DG, Ringer M. Prospective study of arterial catheter-related infection: incidence, sources of infection, and risk factors [abstract 284]. In: Programs and abstracts of the 29th Interscience Conference of Antimicrobial Agents and Chemotherapy, Houston, TX, 1989. P.239 107. Snyder RH, Archer FJ, Endy T, et al. Catheter infection: a comparison of two catheter maintenance techniques. Ann Surg 1988;208:651–653. 108. Cook D, Randolph A, Kernerman P, et al. Central venous catheter replacement strategies: a systematic review of the literature. Crit Care Med 1997;25:1417–1424. 109. Beck-Sague CM, Jarvis WR. Epidemic bloodstream infections associated with pressure transducers: a persistent problem. Infect Control Hosp Epidemiol 1989;10:54–59. 110. Hoffman KK, Weber DJ, Samsa GP, et al. Transparent polyurethane film as an intravenous catheter dressing: a meta-analysis of the infection risks. JAMA 1992;267:2072–2076. 111. Maki DG, Mermel L, Martin M, et al. A highly-semipermeable polyurethane dressing does not increase the risk of CVC-related BSI: a prospective, multicenter, investigator-blinded trial [abstract 230]. In: Programs and abstracts of the 36th Interscience Conference of Antimicrobial Agents and Chemotherapy, New Orleans, LA, 1996. 112. Conly JM, Grieves K, Peters B. A prospective, randomized study comparing transparent and dry gauze dressings for central venous catheters. J Infect Dis 1989;159:310–319. 113. Curchoe RM, Powers J, El-Daher N. Weekly transparent dressing change linked to increased bacteremia rates. Infect Control Hosp Epidemiol 2002;23:730–732. 114. Price CS, Hacek D, Noskin GA, et al. An outbreak of bloodstream infections in an outpatient hemodialysis center. Infect Control Hosp Epidemiol 2002;23:725–729. 115. Flowers RH, Schwenzer RJ, Kopel RJ, et al. Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection: a randomized controlled trial. JAMA 1989;261:878–883. 116. Jarvis WR, Martone WJ. Predominant pathogens in hospital infections. J Antimicrob Chemo 1992; 29(Suppl A):19–24. 117. Hampton AA, Sherertz RJ. Vascular-access infections in hospitalized patients. Surg Clin North Am 1988;68:57–71. 118. Sherertz RJ, Raad I, Belani A. Three year experience with sonicated vascular catheter cultures in a clinical microbiology laboratory. J Clin Microbiol 1990;28:76–82. 119. Peacock SJ, Eddleston M, Emptage A, et al. Positive intravenous line tip cultures as predictors of bacteremia. J Hosp Infect 1998;40:35–38. 120. Farr BM. Understaffing: a risk factor for infection in the era of downsizing? Infect Control Hosp Epidemiol 1996;17:147–149. 121. Raad I, Hohn DC, Gilbreath BJ, et al. Prevention of central venous catheter related infections by using maximal sterile barrier precautions during insertion. Infect Control Hosp Epidemiol 1994;15(4 Pt 1):231–238. 122. Levin A, Mason AJ, Jindal KK, et al. Prevention of hemodialysis subclavian vein catheter infections by topical povidone iodine. Kidney Int 1991;40:934–938. 123. Prager RL, Silva J. Colonization of central venous catheters. South Med J 1984;77:458–461. 124. Hill RLR, Fisher AP, Ware RJ, et al. Mupirocin for the reduction of colonization of internal jugular cannulae-randomized controlled trial. J Hosp Infect 1990;15:321. 125. O'Grady N, Alexander M, Dellinger E, et al. Guidelines for the prevention of intravascular catheter-related infections. Infect Control Hosp Epidemiol 2002;23:759–769. 126. Bock SN, Lee RE, Fisher B, et.al. A prospective randomized trial evaluating prophylactic antibiotics to prevent triple-lumen catheter-related sepsis in patients treated with immunotherapy. J Clin Oncol 1990;8:161–169. 127. Powell C, Kudsk KA, Kulich PA, et al. Effect of frequent guidewire changes on triple-lumen catheter sepsis. J Parenteral Enteral Nutr 1988;12:462–464. 128. Eyer S, Brummitt C, Crossley K, et al. Catheter-related sepsis: a prospective, randomized study of three methods of long-term catheter maintenance. Crit Care Med 1990;18:1079. 129. Maki DG, Ringer M, Alvarado CJ. Prospective randomized trial of povidone-iodine, alcohol, and chlorhexidine for prevention of infection associated with central venous and arterial catheters. Lancet 1991;338:339–343. 130. Mimoz O, Pieroni L, Lawrence C, et al. Prospective, randomized trial of two antiseptic solutions for prevention of central venous or arterial catheter colonization and infection in intensive care unit patients. Crit Care Med 1996;24:1818–1823. 131. Sheehan G, Leicht K, O, Biren M, et al. Chlorhexidine versus povidone-iodine as cutaneous antisepsis for prevention of vascular-catheter infection [abstract 414]. In: Programs and abstracts of the 33rd Interscience Conference of Antimicrobial Agents and Chemotherapy, New Orleans, LA, 1993. 132. Strand CL, Wajsbort RR, Sturmann K. Effect of iodophor vs iodine tincture skin preparation on blood culture contamination rate. JAMA 1993;269:1004–1006. 133. Calfee DP, Farr BM. Comparison of four antiseptic preparations for skin in the prevention of contamination of percutaneously-drawn blood cultures: a randomized trial. J Clin Microbiol 2002;40:1660–1665. 134. Andrivet P, Bacquer A, Vu Ngoc C, et al. Lack of clinical benefit from subcutaneous tunnel insertion of central venous catheters in immunocompromised patients. Clin Infect Dis 1994;18:199–206. 135. Guichard I, Nitemberg G, Abitbol JL, et al. Tunnelled versus non-tunnelled catheters for parenteral nutrition in an intensive care unit: a controlled prospective study of catheter related sepsis. Clin Nutr 1986;5(Suppl 1):169. 136. Nahum E, Levy I, Katz J, et al. Efficacy of Subcutaneous tunneling for prevention of bacterial colonization of femoral central venous catheters in critically ill children. Pediatr Infect Dis J 2002;21:1000–1004. 137. Timsit J, Bruneel F, Cheval C, et al. Use of tunneled femoral catheters to prevent catheter-related infection: a randomized, controlled trial. Ann Intern Med 1999;130:729–735. 138. Timsit J, Sebille V, Farkas JC, et al. Effect of subcutaneous tunneling on internal jugular catheter-related sepsis in critically ill patients: a prospective randomized multicenter study. JAMA 1996;276:1416–1420. 139. Hanley EM, Veeder A, Smith T, et al. Evaluation of an antiseptic triple-lumen catheter in an intensive care unit. Crit Care Med 2000;28:366–370. 140. Veenstra DL, Saint S, Saha S, et al. Efficacy of antiseptic-impregnated catheters in preventing catheter-related bloodstream infection: a meta-analysis. JAMA 1999;281:261–267. 141. Maki, DG, Wheeler S, Stolz SM, et al. Prevention of central venous catheter related bloodstream infection by use of an antiseptic-impregnated catheter. A randomized, controlled trial. Ann Intern Med 1997;127:257–266. 142. Kamal GD, Pfaller MA, Remple LE, et al. Reduced intravascular catheter infection by antibiotic bonding. JAMA 1991;265:2364–2368. 143. Trooskin SZ, Donetz AP, Harvey RA, et al. Prevention of catheter sepsis by antibiotic bonding. Surgery 1985;97:547–551. 144. Sherertz RJ, Carruth WA, Hampton AA, et al. Efficacy of antibiotic-coated catheters in preventing subcutaneous Staphylococcus aureus infection in rabbits. J Infect Dis 1993;167:98–106. 145. Romano G, Berti M, Goldstein BP, et al. Efficacy of a central venous catheter (Hydrocath) loaded with teicoplanin in preventing subcutaneous staphylococcal infection in the mouse. Int J Med Microbiol Virol Parasitol Infect Dis 1993;279:426–433. 146. Darouiche RO, Raad I, Heard SO, et al. A comparison of two antimicrobial-impregnated central venous catheters. N Engl J Med 1999;340:1–8. 147. Sampath LA, Tambe SM, Modak SM. In vitro and in vivo efficacy of catheters impregnated with antiseptics or antibiotics: evaluation of the risk of bacterial resistance to the antimicrobials in the catheters. Infect Control Hosp Epidemiol 2001; xx:640–646. 148. McLean RJ, Hussain AA, Sayer M, et al. Antibacterial activity of multilayer silver-copper surface films on catheter material. Can J Microbiol 1993;39:895–899. 149. Liu WK, Tebbs SE, Byrne PO, et al. The effects of electric current on bacteria colonizing intravenous catheters. J Infect 1993;27:269. 150. Raad I, Zermeno A, Dumo M, et al. In vitro antimicrobial efficacy of silver iontophoretic catheter. Biometrics 1996;17:1055–1059. P.240 151. Longuet P, Douard MC, Arlet G. Venous access port-related bacteremia in patients with acquired immunodeficiency syndrome or cancer: the reservoir as a diagnostic and therapeutic tool. Clin Infect Dis 2001;32:1776–1783. 152. Krzywda EA, Andris DA, Edmiston CE. Treatment of Hickman catheter sepsis using antibiotic lock technique. Infect Control Hosp Epidemiol 1995;16:596–598. 153. Capdevila JA, Segarra A, Planes AM, et al. Long term follow-up of patients with catheter related sepsis (CRS) treated without catheter removal [abstract 257]. In: Programs and abstracts of the 35th Interscience Conference of Antimicrobial Agents and Chemotherapy, San Francisco, 1995. 154. Messing B, Peitra-Cohen S, Debure A, et al. Antibiotic-lock technique: a new approach to optimal therapy for catheter-related sepsis in home-parenteral nutrition patients. J Parenteral Enteral Nutr 1988;12:185–189. 155. Gaillard JL, Merlino R, Pajot N, et.al. Conventional and nonconventional modes of vancomycin administration to decontaminate the internal surface of catheters colonized with coagulase-negative staphylococci. J Parenteral Enteral Nutr 1990;14:593–597. 156. Arnow PM, Kushner R. Malassezia furfur catheter infection cured with antibiotic lock therapy [Letter]. Am J Med 1982;90:128–130. 157. Douard MC, Arlet G, Leverger G, et al. Quantitative blood cultures for diagnosis and management of catheter-related sepsis in pediatric hematology and oncology patients. Intens Care Med 1991;17:30–35. 158. Elian JC, Frappaz D, Ros A, et.al. Study of serum kinetics of vancomycin during the “antibiotic-lock” technique. Arch Fr Pediatr 1992;49:357–360. 159. Cowan CE. Antibiotic lock technique. J Intravenous Nurs 1992;15:283–287. 160. Schwartz C, Henrickson KJ, Roghmann K, et al. Prevention of bacteremia attributed to luminal colonization of tunneled central venous catheters with vancomycin-susceptible organisms. J Clin Oncol 1990;8:1591–1597. 161. Jurewitsch B, Lee T, Park J, et al. Taurolidine 2% as an antimicrobial lock solution for prevention of recurrent catheter-related bloodstream infections. J Parenteral Enteral Nutr 1998;22:242–244. 162. Raad I, Hachem R, Tcholakian RK, et al. Efficacy of minocycline and EDTA lock solution in preventing catheter-related bacteremia, septic phlebitis, and endocarditis in rabbits. Antimicrob Agents Chemother 2002;43:327–332. 163. Chatzinikolaou I, Zipf TF, Hanna HA, et al. Minocycline-ethylenediaminetetraacetate lock solution for the prevention of implantable port infections in children with cancer. Clin Infect Dis 2003;36:116–119. 164. Hospital Infection Control Practices Advisory Committee (HICPAC). Recommendations for preventing the spread of vancomycin resistance. Infect Control Hosp Epidemiol 1995;16:105–113. 165. Douzinas EE, Pitardis MT, Louris G. Prevention of infection in multiple trauma patients by high-dose intravenous immunoglobins. Crit Care Med 2000;28:8–15. 166. Gruson D, Hilbert G, Vargas F. Impact of colony-stimulating factor therapy on clinical outcome and frequency rate of nosocomial infections in intensive care unit neutropenic patients. Crit Care Med 2000;28:3155–3160. 167. McCowen KC, Friel C, Stewrnberg J, et al. Hypocaloric total parenteral nutrition: effectiveness in prevention of hyperglycemia and infectious complications—a randomized clinical trial. Crit Care Med 2000;28:3606–3611.
Воскресенье, 06 Ноябрь 2011 | 998 Просмотры | Печать | PDF
Long-term intravascular devices have become indispensable in the modern medical care of chronically ill patients such as cancer patients, patients with renal failure requiring chronic hemodialysis, or patients requiring organ or bone marrow transplantation. In the 1960s and 1970s treatment of cancer patients through a small peripheral venous catheter used to be complicated by extravasation of vesicant chemotherapeutic agents and thrombosis of peripheral veins, which often limited anticancer chemotherapy. Long-term silicone central venous catheters (CVCs) allowed the extended, safe use of anticancer chemotherapeutic agents as well as the potential for appropriate use of total parenteral nutrition (TPN) fluids, blood products, and other intravenous therapeutic agents (1). For patients with short bowel syndrome, long-term CVCs have become the only source for nutritional support through TPN. Similarly, patients requiring hemodialysis who have had prior failure of arteriovenous fistulas or shunts become totally dependent on intravascular catheter-related access for their hemodialysis. In all of these clinical situations, the long-term CVC becomes an essential device for the maintenance of life. There is no standard agreed-upon definition for long-term catheters in terms of the duration of catheterization. Sherertz (2) defined long-term catheters as those with a duration of placement of an average of >8 days. Rather than using an average duration, we have defined this term in a previous study to signify catheters that remain in place for >30 days (3). We also defined short-term catheterization by duration of placement of <10 days, and intermediate catheterization by duration of placement ranging between 10 and 30 days (3). Long-term intravascular devices can be categorized into one of three groups: (a) nontunneled long-term CVCs [such as peripherally inserted central catheters (PICCs) or subclavian CVCs such as Hohn catheters]; (b) cuffed and tunneled catheters (such as Hickman/Broviac, Groshong, and tunneled Uldall catheters); and (c) implanted subcutaneous central venous ports. NONTUNNELED CATHETERS Traditionally it was assumed that the only method of maintaining long-term intravascular access in chronically ill patients was through surgically implantable CVCs such as the tunneled catheters and implantable ports. Over the last decades, nontunneled long-term silicone CVCs have become more accepted as a cost-effective form of intravascular access. In addition, these catheters could be maintained for a long period, up to 400 days, without complications (4). Nontunneled long-term catheters consist of two types: nontunneled subclavian silicone catheters and PICC lines. Nontunneled subclavian catheters are inserted percutaneously via the subclavian vein into the superior vena cava, in the outpatient nonsurgical setting. The advantage of these catheters is that they are associated with low cost, because their insertion does not require the use of an operating room or a special surgical technique (1). In addition, these catheters can be exchanged over a guidewire and the removed intravascular segment may be cultured if a catheter infection is suspected or if a new catheter needs to be inserted. These catheters are available as single-, double-, or triple-lumen cannulas. The PICC lines are becoming widely used, particularly for outpatient long-term central venous therapy, such as patients requiring intravenous home antibiotics for osteomyelitis or endocarditis, cancer patients, or patients requiring TPN delivery. These catheters are usually inserted in the antecubital space, via the cephalic or basilic vein, and advanced into the central venous system. These catheters are very cost-effective, because they can be inserted in the outpatient clinic by a trained infusion therapy nurse and do not require a physician for their insertion. At our institution, these catheters can be maintained for an average of 3 months and are associated with a low infection rate and cost (4). However, their main disadvantage is a high rate of aseptic thrombophlebitis related to mechanical contact (5). Most of these catheters are made of silicone, although some are made of polyurethane. TUNNELED CATHETERS In 1973, Broviac et al. (6) described the first surgically implanted tunneled catheter to be used in pediatric patients requiring P.242 long-term TPN. Later, Hickman et al. (7) described another long-term tunneled catheter for cancer patients requiring bone marrow transplantation. These catheters are usually tunneled under the skin for several inches until they reach the cannulated vein. Tunneled catheters have a Dacron cuff that is located in the proximal subcutaneous segment 5 cm from the exit insertion site. After insertion, the Dacron cuff becomes enmeshed with fibrous tissue, hence anchoring the catheter and creating a tissue interface mechanical barrier against the migration of skin microorganisms along the external intracutaneous pathway. Tunneled catheters usually exit the body midway between the nipple and the sternum. Another vascular access catheter is the Groshong, which, unlike the Hickman/Broviac, is thin walled and has two slit valves adjacent to a rounded closed end that remains closed unless fluids are being infused or blood is being drawn. This decreases the risk of intraluminal blood clotting or infusion of air when the catheter is not in use. Hence, this type of catheter does not require daily heparin flushes, but rather is flushed with saline on a weekly basis. IMPLANTABLE PORTS To eliminate the migration of skin microorganisms from the skin insertion site in externalized catheters along the intracutaneous pathway, the surgically implanted subcutaneous central venous ports were developed where the whole catheter, including the metallic port, is placed beneath the skin (8). Hence, implantable ports consist of a metal/titanium or plastic port placed beneath the skin and connected to a catheter that enters the cannulated vein. Ports are usually placed in a subcutaneous pocket on the upper chest or, less often, in the antecubital area of the arm (peripheral port). Ports are available as single- or double-lumen catheters with or without Groshong valves and can be accessed as needed with a steel needle. EPIDEMIOLOGY The bloodstream infection (BSI) rates associated with long-term CVCs should be reported using catheter-days as the denominator. The Centers for Disease Control and Prevention (CDC) recommends that rates of catheter-related bloodstream infections (CR-BSI) be expressed per 1,000 device-days. This recommendation takes into consideration the varying risks of CR-BSI over time for the different types of CVCs. According to Crnich and Maki (9), although the rates of CR-BSI per 100 CVCs used are usually higher for long-term devices, the risk per 1,000 catheter-days is usually considerably lower than that for short-term CVCs. In 17 studies reviewed by Press et al. (10), 21 studies reviewed by Decker and Edwards (11), 13 studies reviewed by Clarke and Raffin (12), and 26 studies reviewed by Howell et al. (13), the average infection rate for long-term CVCs in cancer patients ranged from one to two episodes per 1,000 catheter days. Assuming this rate and the fact that half a million long-term CVCs are inserted annually in the United States, the estimated annual number of episodes of catheter-associated bacteremia that occur in the U.S. related to the use of these catheters in cancer patients is between 50,000 and 100,000. Several studies have compared the efficacy of tunneled catheters (such as Hickman/Broviac catheters) with implantable ports. Mueller et al. (14), in a prospective, randomized study, compared the complications of the two types of long-term catheters and found no significant difference in infection rates between the two types of devices. Similarly, Keung et al. (15) conducted a retrospective study of infectious complications in 111 long-term CVCs. Multivariable analysis revealed no significant difference in infection rates between tunneled catheters and implantable ports. On the other hand, there are several studies that suggest that ports may be associated with lower infection rates. Mirro et al. (16) evaluated 266 tunneled catheters and 93 implantable ports in children with cancer, and showed that, when all causes of failure were analyzed including infectious complications, ports had a significantly longer duration of use than tunneled catheters. In a prospective observational study conducted at Memorial Sloan-Kettering on 1,630 long-term CVCs (923 tunneled catheters and 707 ports) Groeger et al. (17) found that the incidence of infection per device per day was 12 times greater with the tunneled catheter than with ports. Therefore, these data might suggest that ports are associated with a lower infection rate than tunneled catheters, even though they are not conclusive. In addition, the data should be analyzed with caution because there could be confounding variables, such as the various uses of the catheters (including the use of TPN), duration of neutropenia, and thrombotic complications that were not taken into consideration. There are very few data in the literature comparing tunneled with nontunneled long-term CVCs in terms of infection rates. In a prospective randomized study, Andrivet et al. (18) showed that the infection rate associated with nontunneled subclavian silicone CVCs were not different from those related to tunneled silicone catheters. However, the lack of a difference could be related to the small sample size. In a prospective study evaluating nontunneled long-term CVCs at the M. D. Anderson Cancer Center, we determined that the infection rates for PICC lines and nontunneled subclavian CVCs was 1.4 per 1,000 catheter days, which was comparable to what was described for Hickman catheters in the literature (4). At the M. D. Anderson Cancer Center, the cost of insertion of nontunneled catheters, including the chest x-ray postinsertion and other related fees, is in the range of $1,190 to $1,326 as compared with more than $6,502 for the Hickman tunneled CVC. The cost of placing an implantable port at our institution is about $7,076. Given the comparable durability of all long-term catheters, the potential marginal difference in infection rates might not justify the wide difference in cost between the tunneled catheters and ports on the one hand and the nontunneled CVC (PICC lines and nontunneled subclavian catheters) on the other. PATHOGENESIS Microbial adherence and colonization of long-term catheters is the by-product of the interaction of several factors: (a) host-derived P.243 proteins, (b) microbial factors, (c) catheter material, and (d) iatrogenic factors. After insertion, a thrombin sheath covers the internal and external surfaces of the catheter, which is rich in host proteins (19, 20). These proteins include fibronectin, fibrinogen, laminin, thrombospondin, and collagen (1, 21, 22, 23, 24, 25). Staphylococcus aureus binds strongly to fibronectin and fibrinogen, whereas coagulase-negative staphylococci bind strongly to fibronectin (21, 22). In addition, Candida albicans has been shown to bind well to fibrin (26). Biofilm formation represents the microbial factor involved in the enhancement of adherence of microorganisms to catheter surfaces. Microorganisms, such as coagulase-negative staphylococci, S. aureus, and even C. parapsilosis, have the potential of undergoing intrinsic phenotypic changes that result in the expression of several enzymes that lead to the production of an exopolysaccharide, thus causing the biofilm to form (27, 28, 29, 30, 31, 32). Microorganisms embed themselves in this layer of biofilm (or microbial slime), and hence protect themselves from antimicrobial agents such as glycopeptides (33, 34). Other microbial factors, such as hydrophobicity and the surface charges of microorganisms contribute to the adherence to catheter materials such as silicone (35, 36). Hydrophobic staphylococcal microorganisms adhere better to silicone surfaces of which most long-term catheters are made, than to the polyurethane or Teflon surfaces of short-term catheters. The material from which the catheters are made plays a role in the adherence of microorganisms to the catheter surface. The physical characteristics of the catheter surfaces, including hydrophobicity, surface charges, irregularities, and defects on the catheter surface and the thrombogenicity of the catheter surface, contribute to the process of microbial adherence (1, 2). Several investigators have shown, for example, that Staphylococcus and Candida species adhere better to polyvinyl chloride catheters than to Teflon catheters (37, 38). Sherertz et al. (39) have demonstrated in a rabbit model that silicone catheters are easier to infect with S. aureus than polyurethane, Teflon, or polyvinyl chloride catheters. This was also shown by Vaudaux et al. (40), who demonstrated that indwelling silicone catheters, after being removed from patients, were more prone to S. aureus adherence than were polyurethane or polyvinyl chloride catheters. This was related to the fact that silicone catheters tend to have a direct toxic effect on neutrophils, alter neutrophil chemotaxis, and cause a localized depletion of complement (41, 42). (For additional information on the pathogenesis of infections associated with implantation of biomaterials, see Chapter 67). Iatrogenic factors associated with medical interventions in high-risk patients entail a higher risk of colonization of catheter surfaces. These consist of the use of TPN fluids and lipid emulsions, interleukin-2, and long-term hemodialysis (1, 2). TPN has been associated with higher rates of infection in tunneled catheters (43). The 25% dextrose and the lipid emulsions have been associated with microbial growth, particularly Candida species and Malassezia furfur (2). In addition, interleukin-2 has also been shown to predispose to catheter colonization and infection by staphylococcal microorganisms (44). It is postulated that interleukin alters neutrophil chemotaxis toward staphylococcal microorganisms, and hence leads to a higher degree of colonization of catheter surfaces with these microbial agents. Finally, chronic hemodialysis patients have a high rate of nasal carriage of S. aureus, ranging from 30% to 65% (45, 46, 47). S. aureus chronic hemodialysis carriers have a threefold higher risk of contracting catheter-related S. aureus BSI when compared with noncarriers (48). The majority (more than 90%) of S. aureus infections in carriers are caused by the same type as that carried in the nares (45). The most common microorganisms causing catheter-associated infections in long-term CVCs are coagulase-negative staphylococci, S. aureus, and yeasts (1, 2). This is related to the fact that staphylococci are skin microorganisms. In addition, staphylococci and Candida adhere well to host proteins found on catheter surfaces and tend to form a microbial biofilm (26, 27, 28, 29, 30, 31, 32). This is in contrast to gram-negative microorganisms, such as Escherichia coli and Klebsiella pneumoniae, that do not adhere well to fibronectin and fibrin and are not known to produce a biofilm. Other microorganisms that have been associated with long-term CVC infections are Bacillus species, Corynebacterium species, Pseudomonas aeruginosa, Acinetobacter species, Stenotrophomonas maltophilia, micrococcus, Achromobacter, rapidly growing mycobacteria, and various other fungal microorganisms such as M. furfur and Fusarium oxysporum (49). For long-term catheters, the lumen seems to be the major site of colonization and source of CR-BSIs. This has been shown for catheters used for long-term hemodialysis and for CVCs used for total parenteral nutrition and cancer treatment (50, 51, 52). Sherertz (2) estimated that the hub/lumen contributed 66% of the microorganisms that caused infections of long-term catheters and that 26% of the microorganisms were from the skin. However, for short-term catheters with an average duration of <8 days, the skin seems to be the major source, followed by the hub/lumen (53, 54, 55). The relative contribution of contaminated infusate, hematogenous seeding from a remote infected source, or extension from a contiguous site of infection seems to be low even in long-term catheters. Using semiquantitative scanning electron microscopy studies, we have determined that the extent of biofilm formation and colonization is greater on the external surface of short-term catheters (<10 days of catheterization) than the internal surface (3). However, for catheters that remain in place for >30 days, this phenomenon is reversed with greater biofilm formation and ultrastructural colonization in the lumen of the catheter versus the external surface. Electron microscopy studies have shown that colonization is universal (3, 56). It involves all CVCs within 24 hours of insertion (56). However, although colonization is universal, only a few catheters are associated with infection. There is a quantitative relationship between the number of microorganisms (particularly free-floating microorganisms) on the catheter and the risk of BSIs. Sherertz et al. (57) studied 1,610 CVCs and found that the greater the number of microorganisms retrieved from the catheters by sonication, the greater the risk of BSI. Therefore, infection could be a function of whether the microorganisms on the catheter surface, particularly those that are free-floating, exceed a certain quantitative threshold due to various risk factors outlined above. P.244 MANIFESTATIONS AND DEFINITIONS The clinical manifestations of a CR-BSI for long-term catheters consist of systemic manifestations such as fever and chills, which are nonspecific, particularly in the immunocompromised patient. Clinical evidence of a local infection at an exit site, tunnel, or port pocket would be necessary to suggest the catheter as the source of the BSI. However, for PICC lines, local catheter site inflammation consisting of erythema and phlebitis could be aseptic in nature and reflect a local mechanical irritation of the vein due to the insertion of a large catheter in the relatively small basilic or cephalic veins (4). Therefore, local catheter-related infection or systemic CR-BSIs should be defined in terms of clinical manifestations associated with microbiologic data implicating the catheter as the source of the infection (Tables 18.1 and 18.2). The following definitions were proposed in a recent guideline by the CDC (58): TABLE 18.1. DEFINITIONS OF COLONIZATION AND LOCAL CATHETER-ASSOCIATED INFECTION Catheter colonization: The isolation of 15 colony forming units (CFUs) of any microorganism by semiquantitative culture (roll-plate method) or 103 CFUs by quantitative culture (e.g., sonication technique), from a catheter tip or subcutaneous segment in the absence of simultaneous clinical symptoms. Local catheter-related infection: Exit-site infection: purulent drainage from the catheter exit site, or erythema, tenderness, and swelling within 2 cm of the catheter exit site. Port-pocket infection: erythema and necrosis of the skin over the reservoir of a totally implantable device, or purulent exudate in the subcutaneous pocket containing the reservoir. Tunnel infection: erythema, tenderness, and induration of the tissues overlying the catheter more than 2 cm from the exit site. TABLE 18.2. DIAGNOSIS OF CATHETER-RELATED BLOODSTREAM INFECTION (CR-BSI), BEFORE OR AFTER CATHETER REMOVAL Probable CR-BSI: Common skin microorganisma isolated from two or more blood cultures or Staphylococcus aureus, enterococci, enteric gramnegative bacilli, or Candida isolated from one or more blood cultures Clinical manifestations of infection (fever and chills) No apparent source of sepsis other than the catheter Definitive CR-BSI: All three of the criteria listed above with any one of the clinical or microbiologic findings listed below: Before catheter removal: Clinical evidence: purulent discharge at the catheter insertion site Microbiologic evidence: differential quantitative blood cultures with 5 : 1 ratio of the same microorganism isolated from blood drawn simultaneously from the catheter and peripheral vein OR positive quantitative skin culture After catheter removal: Clinical evidence: clinical sepsis that responds to antibiotic therapy upon catheter removal after being refractory to therapy in the presence of the catheter Microbiologic evidence: isolation of the same microorganism from the peripheral blood and from a semiquantitative or quantitative culture of a catheter segment or tip aCoagulase-negative staphylococci, micrococci, and Bacillus and Corynebacterium species (except for Corynebacterium jeikeium). Local catheter infection: Local catheter infection could exist in different forms, depending on the type of catheter (nontunneled or tunneled implantable port). Exit site infection: either purulence or erythema, tenderness, or induration within 2 cm of the exit site of the catheter, in the absence of concomitant BSI. Pocket infection: purulent exudate in the subcutaneous pocket containing the reservoir of the port or erythema and necrosis of the skin over the reservoir of a totally implantable device (in the absence of concomitant BSI). Tunnel infection: erythema, tenderness, and induration in the tissues overlying the catheter and >2 cm from the exit site. Systemic catheter infection: CR-BSI is defined by the CDC as the isolation of the same microorganisms (identical species and antibiogram) from a semiquantitative [>15 colony-forming units (CFU)/catheter segment] or quantitative (> 103 CFU/catheter segment) catheter culture and from the blood (preferably drawn from a peripheral vein) of a patient with accompanying clinical symptoms of a BSI and no other apparent source of infection (58). Most CR-BSIs are uncomplicated. However, with virulent microorganisms such as S. aureus, C. albicans, and P. aeruginosa, deep-seated infections can occur, particularly catheter-related septic thrombosis, which consists of CR-BSI with an infected thrombus (59, 60, 61). The clinical course of septic thrombosis is characterized by occasional swelling above the site of the thrombotic vein and persistent BSI on antimicrobial therapy even after the removal of the catheter. Other deep-seated infections associated with complicated catheter-related bacteremias and fungemias consist of endocarditis, osteomyelitis, and retinitis in the case of candidemia (59, 60). DIAGNOSIS Catheter-related infections are often overdiagnosed, resulting in unnecessary antimicrobial therapy and wasteful removal of the CVC. Misdiagnosis is often the result of relying on false-positive microbiologic data, such as positive blood cultures from the CVC or clinical data such as catheter-site inflammation/phlebitis associated with PICC lines in the absence of other confirmatory data. Therefore, the diagnosis of these infections is often difficult and should be the result of integrating clinical and microbiologic findings. A positive nonquantitative blood culture drawn through the CVC with a concurrent negative peripheral blood culture should be interpreted with extreme caution. Bryant and Strand (62) demonstrated that 93% of such cultures are often contaminated with microorganisms that colonize the hub or the lumen, and hence do not reflect an infection. This is particularly true for skin microorganisms such as coagulase-negative staphylococci. It has been demonstrated that the positive predictive value of a single positive blood culture for coagulase-negative staphylococci ranges from 4.1% to 26.4% (63, 64, 65, 66). Therefore, prior to initiating antimicrobial therapy and P.245 considering whether the catheter should be removed, a single positive blood culture yielding a skin microorganism should be interpreted in light of associated clinical and microbiologic data. Because CVCs are universally colonized, a positive blood culture from the CVC could reflect intraluminal or hub colonization. Therefore, attention should be paid to other laboratory findings suggestive of BSI and which consist of (a) multiple positive blood cultures of skin microorganisms, (b) quantitative blood cultures revealing a high colony count (>35 CFU/mL of blood) (66), and (c) the same microorganisms isolated from the quantitative catheter culture and peripheral blood culture (58). All three of these factors should be considered to reflect a catheter-related infection in the setting of concurrent signs of infection such as fever and chills with no other apparent source for the infection other than the catheter. Before or after removal of the catheter, the diagnosis should be made based on the interplay of clinical and microbiologic findings. In this chapter we use the approach of Kristinsson (67) in determining the diagnosis in these two situations. The infection is initially suspected when there is a positive blood culture in a patient with a CVC with clinical signs of infection, such as fever and chills, and no other apparent source for the BSI, such as pneumonia, urinary tract infection, intraabdominal infection, or surgical site infection (Table 18.2). This type of BSI has been termed primary BSI. In this case, the primary BSI is a probable catheter-related infection. The diagnosis becomes definitive in the presence of either confirmatory clinical or microbiologic data. Clinical data consist of (a) local inflammation, such as catheter exit site inflammation or tunneled/port inflammatory signs (Tables 18.1 and 18.2); the presence of purulence at the insertion site, particularly in patients with S. aureus bacteremia, is diagnostic of catheter-related bacteremia; and (b) systemic signs of infection, such as fever and chills, that persist despite appropriate antimicrobial therapy for the BSI but resolve with catheter removal (58). Confirmatory microbiologic data are often not available prior to catheter removal. The three best studied methods to determine the diagnosis prior to catheter removal are simultaneous quantitative blood cultures from the CVC and a peripheral vein, differential time to positivity, and, for nontunneled catheters, quantitative cultures of the skin at the exit site (68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80). In the former case, the diagnosis of CR-BSI is often suggested when the number of colonies isolated from a quantitative blood culture obtained from the catheter is severalfold (at least fivefold) more than that quantitated from a peripheral venipuncture blood culture (69, 70, 71, 72, 73, 74). Differential time to positivity (DTP) is a method that was shown in cancer patients to be a simple and reliable tool for in situ diagnosis of catheter-related bacteremia. Blot et al. (77, 78) defined DTP as the difference in time necessary for the blood cultures drawn from a peripheral vein and through the catheter to become positive. When DTP was >120 minutes, this diagnostic method was shown to be highly sensitive (100%) and specific (96.4%) for the diagnosis of CR-BSI (77). Blot et al. (78) concluded that using DTP as a diagnostic technique is mainly of value for patients requiring long-term catheterization. However, another prospective study found that DTP of >120 minutes was highly suggestive of CR-BSI associated with the use of both short-term and long-term CVCs (79). Another prospective study found that DTP does not seem to be a useful diagnostic tool for the diagnosis of CR-BSI in medical-surgical intensive care unit (ICU) patients (80). The DTP method needs to be evaluated further in different patient populations to better understand the different settings in which it can be utilized accurately. Quantitative skin cultures are most convenient in the setting of a primary BSI (75, 76, 81, 82). Unfortunately, there is no standard method for culturing the skin around the insertion site, and criteria for positive quantitative cultures have varied in the published studies. Depending on the methodology used, a quantitative culture of >15 CFUs for a small surface area (9–10 cm2) (81, 82) or >1,000 CFUs for a large surface area (24–25 cm2) (75, 76) is suggestive of a CR-BSI in a patient with primary bacteremia/fungemia. Several other nonstandardized methods were developed to diagnose catheter-related infection in surgically implanted long-term catheters prior to removal, such as a wire brush or intraluminal brush method, Gram staining or acridine orange staining of blood drawn through the catheter, and culture of the catheter hub (83, 84, 85, 86, 87). After catheter removal, semiquantitative and quantitative catheter cultures have helped in diagnosing CR-BSI. The roll-plate semiquantitative culture method is most commonly used for culturing catheters (88). However, this method is limited in that it cultures only the external surface of the catheter and may not retrieve microorganisms that are well embedded in the biofilm on the catheter surface. The fact that this method does not quantitate microorganisms from the lumen of the catheter is important for long-term indwelling catheters where colonization is mostly luminal. In a study of long-term catheters (nontunneled CVC and PICC lines) at the M. D. Anderson Cancer Center, the sensitivity of the roll-plate technique was 45% compared with 72% for the sonication technique for making the diagnosis of catheter colonization or catheter-related infection by culture of the intravascular segment of the catheter (3). Quantitative catheter cultures, particularly sonication, which retrieves microorganisms from the external and internal surfaces, have been shown to be of higher diagnostic value than the roll-plate technique, particularly for long-term CVC with predominantly luminal colonization (38990). If semiquantitative or quantitative catheter cultures are not done, then a clinical response to catheter removal within 48 hours, after failure of antimicrobial therapy to resolve the infection, with the catheter in situ, is highly suggestive of CR-BSI. If the bacteremia or fungemia persists after catheter removal in spite of the use of appropriate antimicrobial agents, then one has to determine whether the patient has a deep-seated catheter-related infection, such as right-sided endocarditis or septic thrombosis (59, 60). In these situations, a venogram would be useful to rule out septic thrombosis, and a transesophageal echogram might be useful to detect valvular vegetations suggestive of endocarditis. PREVENTION Effective preventive strategies for long-term CVC-related infections should be based on an understanding of the pathogenesis P.246 of these infections. Because luminal colonization is the major source of BSIs in long-term catheters, preventing colonization of the external surface of the catheter during the early phase postinsertion will not decrease the overall rate of infection. One such example is the use of the silver-impregnated cuff, which was shown to interrupt the intracutaneous migration of microorganisms and to decrease the risk of short-term catheter colonization and infection (53). However, this silver-impregnated cuff has failed to protect against infections in long-term tunneled Hickman catheters. Measures that decrease the risk of colonization of the lumen of the catheter have been shown to be of benefit in decreasing the risk of catheter-associated infection for long-term CVCs (91). However, most of the preventive measures suggested for prevention of long-term catheter-associated infections have limited data to support their use with respect to this type of catheter (Table 18.3). TABLE 18.3. MEASURES WITH LIMITED DATA FOR THE PREVENTION OF LONG-TERM CATHETER-RELATED INFECTIONS Maximal sterile barriers Skilled infusion therapy team Tunneling of catheters Flush solutions/antibiotic lock New antiseptic hub Antimicrobial coating of catheters Maximal Sterile Barriers A prospective randomized study was conducted to test the efficacy of maximal sterile barriers in reducing infections associated with long-term nontunneled subclavian silicone catheters with a mean duration of placement of approximately 70 days (92). Maximal sterile barrier precautions (which involve wearing sterile gloves and gown, a cap, and using a large drape during insertion of the catheter) were compared with routine procedures (which involve wearing only sterile gloves and use of a small drape). Maximal sterile barrier precautions decreased the risk of catheter-related bacteremia from 0.5/1,000 catheter-days to 0.02/1,000 catheter-days. Long-term catheters consisted of nontunneled subclavian CVCs and PICC lines. Skilled Infusion Therapy Team In addition to decreasing the catheter-related infection rate by five- to eightfold, an experienced infusion therapy team has been shown to be cost-effective. Most of the studies were done with relatively short-term catheters (93, 94, 95, 96, 97). However, we have reported the finding that the duration of placement of nontunneled, noncuffed silicone catheters (mean duration of catheterization of 109 days) could be prolonged to approach that of the tunneled Hickman catheter with a very low infection rate of 1.4/1,000 catheter-days at the M. D. Anderson Cancer Center (4). This was attributed, at least in part, to the presence of a skilled infusion therapy team at our institution. Tunneling Because tunneled catheters have been associated with long durability and low infection rates, it has been assumed that tunneling decreases the risk of catheter-related infections and is the only safe option for the maintenance of long-term externalized silicone catheters. A prospective, randomized study evaluating the effect of tunneling on long-term silicone catheters was conducted by Andrivet et al. (18), wherein the catheters were used in immunocompromised patients. The risk of catheter-related bacteremia associated with tunneled as compared with nontunneled catheters was 2% and 5%, respectively. The difference was not significant, probably due to the relatively small number of patients in each group (107 and 105 patients, respectively). In another study involving short-term polyurethane catheters placed in the internal jugular vein of critically ill patients, tunneled catheters were associated with a statistically significant lower rate of catheter-related bacteremia than nontunneled catheters, suggesting that tunneling may decrease the risk of infection (98). In a prospective evaluation of more than 14,000 PICCs and 882 tunneled CVCs in cancer patients, the rate of CR-BSI associated with the use of tunneled CVCs was shown to be lower than that of PICCs (0.042 vs. 0.065 per 1,000 catheter-days, p = .004) (99). However, in an evaluation of more than 20,000 nontunneled subclavian CVCs, the CR-BSI rate associated with their use in cancer patients was found to be comparable to that associated with the use of tunneled CVCs (0.073 vs. 0.042 per 1,000 catheter-days, p = .071) (100). The main question, however, is whether tunneling is cost-effective, given the fact that surgery may incur an additional cost of about $3,600 per procedure, and long tunneled silicone catheters can be maintained for a long time with a very low infection rate (1). Ports The lowest rate of CR-BSI has been associated with the use of surgically implanted subcutaneous central venous ports. In a review by Crnich and Maki (9), in which they evaluated the results of 13 prospective studies of subcutaneous central venous ports, the pooled mean of CR-BSI rates associated with ports was at a low 0.2 per 1,000 catheter-days [95% confidence interval (CI) 0.1–0.2] (9). A prospective evaluation of more than 2,000 ports in cancer patients showed that they were associated with CR-BSI rates significantly lower than PICCs and than nontunneled subclavian catheters (0.0074 vs. 0.065 and 0.073/1,000 catheter-days, respectively, p <.0001) (99, 100). Ports are especially useful for intermittent venous access needed for short durations such as with periodic chemotherapy administrations. Antiseptic Dressings A novel chlorhexidine-impregnated hydrophilic polyurethane foam dressing (Biopatch, Johnson & Johnson Medical, Dallas, TX), which can be pressed firmly onto the skin at the catheter insertion site and then covered with a transparent polyurethane dressing, was shown to reduce site skin colonization as well as epidural catheter colonization (101). It also prevented infection at the site of orthopedic traction pins in an animal model (102). The chlorhexidine-impregnated sponge dressings P.247 were evaluated in a multicenter trial involving six neonatal ICU patients (103), where they were found to be similar to gauze and tape combined with periodic skin disinfection with 10% povidone-iodine, in preventing skin colonization and CR-BSI (103). However, the use of these chlorhexidine dressings was associated with 15% incidence of dermatotoxicity in low birth weight neonates (<1,000g). In another prospective randomized study, the chlorhexidine-impregnated dressings were found to decrease CR-BSI by threefold when used with PICCs or arterial catheters (104). Intraluminal Antibiotic Locks This prophylactic measure consists of flushing and filling the lumen of the catheter with antimicrobial agents and leaving the solution to remain in the lumen of the catheter for 6 to 12 hours. Various antimicrobial agents have been used as antimicrobial locks, often following an infection in a surgically implanted catheter in order to treat the infection without removal of the catheter (105, 106). Among the antimicrobial agents used were vancomycin, gentamicin, ciprofloxacin, cefazolin, erythromycin, nafcillin, ceftriaxone, clindamycin, fluconazole, and amphotericin B. Vancomycin in combination with heparin has been used as a daily flushing solution of tunneled CVCs and has been reported to significantly decrease the frequency of catheter-related bacteremia caused by vancomycin-susceptible gram-positive microorganisms colonizing the lumen (107). In a large prospective double-blind study of 126 pediatric oncology patients, the efficacy of different flush solutions was investigated over 36,944 tunneled catheter-days (108). In that study, Henrickson et al. randomized patients to receive one of three prophylactic lock solutions: 10 U/mL heparin, heparin/25 Вg/mL vancomycin, or heparin/vancomycin/2 Вg/mL ciprofloxacin. The rate of total line infections including gram-positive and gram-negative line infections was significantly reduced by either heparin/vancomycin or heparin/vancomycin/ciprofloxacin compared with heparin alone. However, this study did not differentiate between local site infection and CR-BSI. In addition, Crnich and Maki (9) point out that this study also failed to evaluate the impact of using antibiotic lock solutions on nosocomial colonization with microorganisms such as vancomycin-resistant enterococci, methicillin-resistant S. aureus, and fluoroquinolone-resistant gram-negative bacilli. Another randomized double-blind trial in neutropenic cancer patients also found that patients who received a lock solution of 10 U/mL heparin and 25 Вg/mL vancomycin had a significantly lower rate of CR-BSI than those who received a lock solution of heparin alone (p = .05) (109). One study revealed no difference in rates of CR-BSI between children receiving a heparin/vancomycin flush compared with those receiving heparin alone (110). However, with the emergence of resistant microorganisms, it is prudent to avoid using antibiotics that are commonly used in the therapy of BSIs (such as beta-lactam antibiotics, vancomycin, quinolones, and aminoglycosides) for prophylaxis against catheter infections. This is particularly true for vancomycin with the emergence of multidrug-resistant vancomycin-resistant enterococci. A novel catheter flush solution consisting of low concentrations of minocycline and ethylenediaminetetraacetic acid (EDTA) has recently been developed. Minocycline is not commonly used in the treatment of systemic infections and does not have cross-resistance with vancomycin or beta-lactam antibiotics against resistant gram-positive bacteria. A flush solution of minocycline and EDTA (M-EDTA) was shown to have broad-spectrum and often synergistic activity against methicillin-resistant staphylococci, gram-negative bacilli, and C. albicans, and was found to prevent CR-BSIs in several complicated, high-risk patients (111). Also, in a rabbit model, M-EDTA lock solution succeeded more than heparin alone and heparin/vancomycin in preventing catheter colonization, CR-BSI, and phlebitis in all of the study animals (p <.01) (112). In that study, the M-EDTA lock solution also prevented tricuspid endocarditis, as did the heparin-vancomycin lock solution, more effectively than heparin alone (p <.06). In a prospective randomized trial involving patients with long-term hemodialysis CVCs, M-EDTA flush solution significantly reduced rates of catheter colonization (p = .005) (113). Also, in another prospective pediatric cohort study, M-EDTA was used as a lock solution in 14 pediatric cancer patients with ports (114). There were no CR-BSIs, thrombotic events, or adverse events associated with the use of M-EDTA flush solution over a total of 2,073 catheter-days in comparison with a rate of 2.23 infections/1,000 catheter-days in a control group that received heparin flush solution. Taurolidine, a derivative of the amino acid taurine, is an antimicrobial agent, which in high concentrations (250–2,000 Вg/mL) has inhibitory as well as cidal activities against many microorganisms (115). The use of taurolidine lock solution reduced the rate of CR-BSI associated with the use of hemodialysis CVC (116) and other long-term CVCs (117). A combination of taurolidine and citrate-based catheter lock solution (Neutrollin; Biolink Corp., Norwell, MA) reduced bacterial counts in a catheter model by more than 99% (118). The microorganisms affected included S. aureus, S. epidermidis, P. aeruginosa, E. faecalis, and C. albicans. Taurolidine-citrate significantly reduced biofilm on silicone disks in modified Robbins devices more than heparin treatment (by 4.8 logs vs. 1.7 logs, p <.01). The issue of resistance developing with a wide use of prophylactic antibiotic lock solutions needs to be investigated thoroughly. However, currently, the CDC guidelines for the prevention of intravascular catheter-related infections do not recommend the routine use of prophylactic antibiotic lock solutions. The guidelines recommend the use only in special circumstances, such as in treating a patient with a long-term cuffed or tunneled catheter or port who has a history of multiple CR-BSIs despite optimal maximal adherence to aseptic techniques (58). Anticoagulant Flush Solutions The prophylactic use of various anticoagulants, such as heparin, warfarin, and urokinase, as catheter flush solutions have shown efficacy in reducing catheter thrombosis and fibrin deposits on catheters (119, 120, 121, 122, 123). Since the majority of heparin solutions contain preservatives that have antimicrobial activity, it is difficult to attribute with certainty any observed efficacy to the P.248 heparin rather than to the antimicrobial activity of the preservatives (58). Antimicrobial Coating of Catheters Antimicrobial coating of catheters has been shown to be effective in reducing the rate of catheter-related infection in short-term polyurethane catheters. By coating the external surface of catheters with chlorhexidine plus silver sulfadiazine, Maki et al. (124) showed that this combination did decrease the risk of colonization by nearly 50% and decreased the risk of catheter-related bacteremia by at least fourfold. However, the antiinfective efficacy of catheters coated with chlorhexidine and silver sulfadiazine (CH-SS) was not confirmed in three subsequent prospective, randomized studies (125, 126, 127). In addition, these catheters were shown in vitro to have short-term antimicrobial activity that decreased over time, with a half-life of 3 days against S. epidermidis (128). A meta-analysis study analyzed the results of 12 studies investigating the efficacy of catheters impregnated with CH-SS (129). According to this analysis, the mean duration of catheterization with CH-SS catheters was between 5.1 and 11.2 days, and hence their efficacy is only proven for short-term catheterization. A second generation of catheters impregnated with CH-SS, in which the catheters are impregnated both externally and internally, significantly reduced catheter colonization more than uncoated catheters, but failed to reduce the risk of CR-BSI in two prospective randomized trials (130, 131). In a prospective, randomized multicenter study when CVCs impregnated with minocycline and rifampin (M-R) on their external and internal surfaces were compared with first-generation CH-SS catheters, they were shown to be 12 times less likely to be associated with CR-BSI and three times less likely to be colonized. It was shown that the risk of catheter colonization was reduced by threefold and the risk of CR-BSIs was reduced from 5% to 0% (132). No evidence of antibiotic resistance was noted among bacteria recovered from patients with minocycline/rifampin-coated catheters. Preliminary results show that long-term nontunneled silicone catheters impregnated with minocycline/rifampin are also efficacious in reducing CR-BSI in cancer patients (133). Currently, the CDC guidelines for the prevention of intravascular catheter-related infections recommend the use of antimicrobial CVC in adults whose catheter is expected to remain in place for more than 5 days, if rates of CR-BSI remain above the goal set by the individual institution after implementing aseptic techniques, including maximal sterile barrier precautions (58). Elimination of S. aureus in Nasal Carriers S. aureus is the leading cause of long-term catheter-related infections in patients on chronic hemodialysis (134, 135). BSIs in patients on chronic hemodialysis have been associated with nasal carriage of the same microorganism (45, 48). In addition, it was shown that reduction of the nasal S. aureus carrier state resulted in a significant decrease in infection rates with S. aureus (136). Mupirocin ointment 2% applied to the nose two or three times daily for 5 to 7 days has resulted in eradication of the nasal carrier state in more than 80% of patients (136, 137, 138). Therefore, patients on chronic hemodialysis with at least one episode of S. aureus infection should be screened for possible nasal carriage and, if positive, treated with mupirocin. In addition, insertion of the catheter and maintenance care should include close adherence to aseptic technique. Topical application of povidone-iodine to exit sites of chronic hemodialysis catheters at each dressing change has been shown to be useful in decreasing the rate of infection (48). REFERENCES 1. Raad II, Safar H. Long-term central venous catheters. Infectious complications and cost. In: Seifert H, Jansen B, Farr BM, eds. Catheter-related infections. New York: Marcel Dekker, 1997:307–324. 2. Sherertz RJ. Pathogenesis of vascular catheter-related infections. In: Seifert H, Jansen B, Farr BM, eds. Catheter-related infections. New York: Marcel Dekker, 1997:1–29. 3. Raad I, Costerton W, Sabharwal U, et al. Ultrastructural analysis of indwelling vascular catheters: a quantitative relationship between luminal colonization and duration of placement. J Infect Dis 1993;168:400–407. 4. Raad I, Davis S, Becker M, et al. Low infection rate and long durability of nontunneled Silastic catheters: a safe and cost-effective alternative for long-term venous access. Arch Intern Med 1993;153:1791–1796. 5. Cowl, CT, Weinstock JV, Al-Jurf A, et al. Complications and cost associated with parenteral nutrition delivered to hospitalized patients through either subclavian or peripherally-inserted central catheters. Clin Nutr 2000;19(4): 237–243. 6. Broviac JW, Cole JJ, Scribner GH. A silicone rubber atrial catheter for prolonged parenteral alimentation. Surg Gynecol Obstet 1973;136:602. 7. Hickman RO, Buckner CD, Clift RA, et al. A modified right atrial catheter for access to the venous system in marrow transplant recipients. Surg Gynecol Obstet 1979;148:871. 8. Goodman MS, Wickman R. Venous access devices: an overview. Oncol Nurs Forum 1984;11:16–23. 9. Crnich CJ, Maki DG. The promise of novel technology for the prevention of intravascular device-related bloodstream infection in long-term devices. Clin Infect Dis 2002;34:1362–1368. 10. Press OW, Ramsey PG, Larson EB, et al. Hickman catheter infections in patients with malignancies. Medicine 1984;63(4): 189–200. 11. Decker MD, Edwards KM. Central venous catheter infections. Pediatr Clin North Am 1988;35(3): 579–612. 12. Clarke DE, Raffin TA. Infectious complications of indwelling long-term central venous catheters. Chest 1990;97(4): 966–972. 13. Howell PB, Walters PE, Donowitz GR, et al. Risk factors for infection of adult patients with cancer who have tunneled central venous catheters. Cancer 1995;75(6): 1367–1374. 14. Mueller BU, Skelton J, Callender DPE, et al. A prospective randomized trial comparing the infectious complications of the externalized catheters versus a subcutaneously implanted device in cancer patients. J Clin Oncol 1992;10:1943–1948. 15. Keung Y-K, Watkins K, Chen S-C, et al. Comparative study of infectious complications of different types of chronic central venous access devices. Cancer 1994;73:2832–2837. 16. Mirro J, Rao BN, Kumar M, et al. A comparison of placement techniques and complications of externalized catheters and implantable port use in children with cancer. J Pediatr Surg 1990;25(1): 122–124. 17. Groeger JS, Lucas AB, Thaler HT, et al. Infectious morbidity associated with long-term use of venous access devices in patients with cancer. Ann Intern Med 1993;119:1168–1174. 18. Andrivet P, Bacquer A, Vu Ngoc C, et al. Lack of clinical benefit from subcutaneous tunnel insertion of central venous catheters in immunocompromised patients. Clin Infect Dis 1994;18:199–206. 19. Raad II, Luna M, Khalil S-A M, et al. The relationship between the thrombotic and infectious complications of central venous catheters. JAMA 1994;271(13): 1014–1016. P.249 20. Brismar R, Hardstedt C, Jacobson S. Diagnosis of thrombosis by catheter phlebography after prolonged central venous catheterization. Ann Surg 1981;194:779–783. 21. Hawiger J, Timmons S, Strong DD, et al. Identification of a region of human fibrinogen interacting with staphylococcal clumping factor. Biochemistry 1982;21:1407–1413. 22. Kuusela P. Fibronectin binds to Staphylococcus aureus. Nature 1978;276:718–720. 23. Lopes JD, Dos Reis M, Brentani RR. Presence of laminin receptors in Staphylococcus aureus. Science 1985;229:275–277. 24. Hermann M, Suchard SJ, Boxer LA, et al. Thrombospondin binds to Staphylococcus aureus and promotes staphylococcal adherence to surfaces. Infect Immun 1991;59:279–288. 25. Vaudaux P, Pittet D, Haeberli A, et al. Host factors selectively increase staphylococcal adherence on inserted catheters: a role for fibronectin and fibrinogen or fibrin. J Infect Dis 1989;160:865–875. 26. Bouali A, Robert R, Tronchin G, et al. Characterization of binding of human fibrinogen to the surface of germ-tubes and mycelium of Candida albicans. J Gen Microbiol 1987;133:545–551. 27. Christensen GD, Simpson WA, Bisno AL, et al. Adherence of slime-producing strains of Staphylococcus epidermidis to smooth surfaces. Infect Immun 1982;37:318–326. 28. Christensen GD, Simpson WA, Younger JJ, et al. Adherence of coagulase-negative staphylococci to plastic tissue culture plates: a quantitative model for the adherence of staphylococci to medical devices. J Clin Microbiol 1985;22:996–1007. 29. Falcieri E, Vaudaux P, Huggler E, et al. Role of bacterial exopolymers and host factors on adherence and phagocytosis of Staphylococcus aureus in foreign body infection. J Infect Dis 1987;155:524–531. 30. Costerton JW, Irvin RT, Cheng KR. The bacterial glycocalyx in nature and disease. Annu Rev Microbiol 1981;35:299–324. 31. Pfaller MA, Messer SA, Hollis RJ. Variations in DNA subtype, antifungal susceptibility, and slim production among clinical isolates of Candida parapsilosis. Diag Microbiol Infect Dis 1995;21:9–14. 32. Sheth NK, Franson TR, Sohnle PG. Influence of bacterial adherence to intravascular catheters on in vitro antibiotic susceptibility. Lancet 1985;2:1266–1268. 33. Farber BF, Kaplan MH, Clogstron AG. Staphylococcus epidermidis extracted slime inhibits the antimicrobial action of glycopeptide antibiotics. J Infect Dis 1990;161:37–40. 34. Costerton JW, Lappin-Scott HM. Behavior of bacteria in biofilms. Am Soc Microbiol News 1989;55:650–654. 35. Hogt AH, Dankert JJ, Feijin J. Adhesion of Staphylococcus epidermidis and Staphylococcus saprophyticus onto a hydrophobic biomaterial. J Gen Microbiol 1985;131:2485–2491. 36. Reifsteck F, Wee S, Wilkinson BJ. Hydrophobicity-hydrophilicity of staphylococci. J Med Microbiol 1987;24:65–73. 37. Sheth NK, Franson TR, Rose HD, et al. Colonization of bacteria on polyvinyl chloride and Teflon intravascular catheter in hospitalized patients. J Clin Microbiol 1983;18:1061–1063. 38. Rotrosen D, Calderone RA, Edwards JE Jr. Adherence of Candida species to host tissues and plastic surfaces. Rev Infect Dis 1986;8:73–85. 39. Sherertz RJ, Carruth WA, Marosok RD, et al. Contribution of vascular catheter material to the pathogenesis of infection: the enhanced risk of silicone in vivo. J Biomed Mater Res 1995;29:634–645. 40. Vaudaux P, Pittet D, Haeberli A, et al. Fibronectin is more active than fibrin or fibrinogen in promoting Staphylococcus aureus adherence to inserted intravascular devices. J Infect Dis 1993;167:633–641. 41. Marosok R, Washburn R, Indorf A, et al. Contribution of vascular catheter material to the pathogenesis of infection: depletion of complement by silicone elastomer in vitro. J Biomed Mater Res 1996;30:245–250. 42. Lopez-Lopez G, Pascual A, Perea EJ. Effect of plastic catheters on the phagocytic activity of human polymorphonuclear leukocytes. Eur J Clin Microbiol Infect Dis 1990;9:324–328. 43. Fuchs PC, Gustafson ME, King JT, et al. Assessment of catheter-associated infection risk with the Hickman right atrial catheter. Infect Control 1984;5:226. 44. Syndman DR, Sullivan B, Gill M, et al. Nosocomial sepsis with interleukin-2. Ann Intern Med 1990;112:102–107. 45. Yu VL, Goetz A, Wagener M, et al. Staphylococcus aureus nasal carriage and infection in patients on hemodialysis. Efficacy of antibiotic prophylaxis. N Engl J Med 1986;315:91–96. 46. Tuazon CU. Skin and skin structure infection in the patient at risk: carrier state of Staphylococcus aureus. Am J Med 1984;76(suppl 51):166–171. 47. Goldblum SE, Ulrich JA, Reed WP. Nasal and cutaneous flora among hemodialysis patients and personnel: quantitative characterization and pattern of staphylococcal carriage. Am J Kidney Dis 1982;2:281–286. 48. Levin A, Mason AJ, Jindal KK, et al. Prevention of hemodialysis subclavian vein catheter infections by topical povidone-iodine. Kidney Int 1991;40:934–938. 49. Raad I. Intravascular-catheter-related infections. Lancet 1998;351:893–898. 50. Linares J, Sitges-Serra A, Garau J, et al. Pathogenesis of catheter sepsis: a prospective study with quantitative and semiquantitative cultures of catheter hub and segments. J Clin Microbiol 1985;21:357–360. 51. Almirall J, Gonzalez J, Rello J, et al. Infection of hemodialysis catheters: incidence and mechanisms. Am J Nephrol 1989;9:454–459. 52. Cheesbrough JS, Finch RG, Burden RP. A prospective study of the mechanisms of infection associated with hemodialysis catheters. J Infect Dis 1986;154:579–589. 53. Maki DG, Cobb L, Garman JK, et al. An attachable silver-impregnated cuff for prevention of infection with central venous catheters: a prospective randomized multicenter trial. Am J Med 1988;85:307–314. 54. Flowers RH III, Schwenzer KJ, Kopel RF, et al. Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection. JAMA 1989;261:878–883. 55. Mermel LA, McCormick RD, Springman SR. The pathogenesis and epidemiology of catheter-related infection with pulmonary artery Swan-Ganz catheters: a prospective study utilizing molecular subtyping. Am J Med 1991;91(suppl 3B):197S–205S. 56. Anaissie E, Samonis G, Kontoyiannis D, et al. Role of catheter colonization and infrequent hematogenous seeding in catheter-related infections. Eur J Clin Microbiol Infect Dis 1995;14:138. 57. Sherertz RJ, Raad II, Belani A, et al. Three-year experience with sonicated vascular catheter cultures in a clinical microbiology laboratory. J Clin Microbil 1990;28:76–82. 58. Centers for Disease Control and Prevention. Guidelines for the prevention of intravascular catheter-related infections. MMWR March 9, 2002;51(RR-10). 59. Raad I, Narro J, Khan A, et al. Serious complications of vascular catheter-related Staphylococcus aureus bacteremia in cancer patients. Eur J Clin Microbiol Infect Dis 1992;11:675–682. 60. Strinden WD, Helgerson RB, Maki DG. Candida septic thrombosis of the great central veins associated with central catheters. Ann Surg 1985;202:653–658. 61. Benezra D, Kiehn TE, Gold GWM, et al. Prospective study of infection in indwelling central venous catheters using quantitative blood cultures. Am J Med 1988;85:495–498. 62. Bryant JK, Strand CL. Reliability of blood cultures collected from intravascular catheter versus venipuncture. Am J Clin Pathol 1987;88(1): 113–116. 63. Fidalgo S, Vasquez F, Mondoza MC, et al. Bacteremia due to Staphylococcus epidermidis: microbiologic, epidemiologic, clinical and prognostic features. Rev Infect Dis 1990;12:520–528. 64. Ringbert H, Thoren A, Bredbert A. Evaluation of coagulase-negative staphylococci in blood cultures: a prospective clinical and microbiological study. Scand J Infect Dis 1991;23:321–323. 65. Kirchoff LV, Sheagren JN. Epidemiology and clinical significance of blood cultures positive for coagulase-negative staphylococci. Infect Control 1985;6:479–486. 66. Herwaldt LA, Geiss M, Kao C, et al. The positive predictive value of isolating coagulase-negative staphylococci from blood cultures. Clin Infect Dis 1996;22:14–20. P.250 67. Kristinsson KG. Diagnosis of catheter-related infection. In: Seifert H, Jansen B, Farr BM, eds. Catheter-related infections. New York: Marcel Dekker, 1997:31–57. 68. Wing EJ, Norden CW, Shadduck RK, et al. Use of quantitative bacteriologic techniques to diagnose catheter-related sepsis. Arch Intern Med 1979;139:482–483. 69. Raucher HS, Hyatt AC, Barzilai A, et al. Quantitative blood cultures in the evaluation of septicemia in children with Broviac catheters. J Pediatr 1984;104:29–33. 70. Mosca R, Curta S, Forbes B, et al. The benefits of isolator cultures in the management of suspected catheter sepsis. Sepsis 1987;102:718–723. 71. Flynn PM, Shenep JL, Stokes DC, et al. In situ management of confirmed central venous catheter-related bacteremia. Pediatr Infect Dis J 1987;6:729–734. 72. Flynn PM, Shenep JL, Barrett FF. Differential quantitation with a commercial blood culture tube for diagnosis of catheter-related infection. J Clin Microbiol 1988;26:1045–1046. 73. Fan ST, Teoh-Chan CH, Lau KF. Evaluation of central venous catheter sepsis by differential quantitative blood culture. Eur J Clin Microbiol Infect Dis 1989;8:142–144. 74. Capdevila JA, Planes AM, Palomar M, et al. Value of differential quantitative blood cultures in the diagnosis of catheter-related sepsis. Eur J Clin Microbiol Infect Dis 1992;11:403–407. 75. Raad II, Baba M, Bodey GP. Diagnosis of catheter-related infections: the role of surveillance and targeted quantitative skin cultures. Clin Infect Dis 1995;20:593–597. 76. Bjornson HS, Colley R, Bower RH, et al. Association between microorganism growth at the catheter insertion site and colonization of the catheter in patients receiving total parenteral nutrition. Surgery 1982;92:720–727. 77. Blot F, Schmidt E, Nitenberg G, et al. Earlier positivity of central venous versus peripheral blood cultures is highly predictive of catheter-related sepsis. J Clin Microbiol 1998;36:105–109. 78. Blot F, Nitenberg G, Chachata E, et al. Diagnosis of catheter-related bacteremia: a prospective comparison of the time to positivity of hub-blood versus peripheral-blood cultures. Lancet 1999;354:1071–1077. 79. Raad II, Hanna HA, Alakech B, et al. Diagnosis of catheter-related bloodstream (CRBSI): differential time to positivity (DTP) for short-term and long-term central venous catheters (CVC). In: Program and Abstracts of the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, September 17–20, 2000. Washington: American Society for Microbiology, 2000. 80. Rinjnders BJA, Verwaest C, Peeetermans E, et al. Difference in time to positivity of hub-blood versus nonhub-blood cultures is not useful for the diagnosis of catheter-related bloodstream infection in critically ill patients. Crit Care Med 2001;29(7): 1399–1403. 81. Cercenado E, Ena J, Rodriguez-Creixems M, et al. A conservative procedure for the diagnosis of catheter-related infections. Arch Intern Med 1990;150:1417–1420. 82. Guidet B, Nicola I, Barakett V, et al. Skin versus hub cultures to predict colonization and infection of central venous catheter in intensive care patients. Infection 1994;22:43–48. 83. Kite P, Dobbins BM, Wilcox MH, et al. Evaluation of a novel endoluminal brush method for in situ diagnosis of catheter-related sepsis. J Clin Pathol 1997;50:278–282. 84. Tighe MJ, Kite P, Thomas D, et al. Rapid diagnosis of catheter-related sepsis using acridine orange leukocyte cytospin test and endoluminal brush. J Parenteral Enteral Nutr 1996;20:215–218. 85. Tighe MJ, Kite P, Fawley WN, et al. An endoluminal brush to detect the infection central venous catheter in situ: a pilot study. BMJ 1996;313:1528–1529. 86. Moonens F, Alami SE, van Gossum A, et al. Usefulness of gram staining of blood collected from total parenteral nutrition catheter for rapid diagnosis of catheter-related sepsis. J Clin Microbiol 1994;32:1578–1579. 87. Linares J, Sitges-Serra A, Garau J, et al. Pathogenesis of catheter sepsis: a prospective study with semiquantitative cultures of catheter hub segments. J Clin Microbiol 1985;21:357–360. 88. Maki DG, Weise CE, Sarafin HW. A semiquantitative culture method for identifying intravenous catheter infection. N Engl J Med 1977;296:1305–1309. 89. Sherertz RJ, Heard SO, Raad II. Diagnosis of triple-lumen catheter infection: comparison of roll plate, sonication, and flushing methodologies. J Clin Microbiol 1997;35(3): 641–646. 90. Raad II, Sabbagh MF, Rand KH, et al. Quantitative tip culture methods and the diagnosis of central venous catheter-related infections. Diagn Microbiol Infect Dis 1991;15:13–20. 91. Groeger JS, Lucas AB, Coit D, et al. A prospective randomized evaluation of silver-impregnated subcutaneous cuffs for preventing tunneled chronic venous access catheter infections in cancer patients. Ann Surg 1993;218:208–210. 92. Raad II, Hohn DC, Gilbreath BJ, et al. Prevention of central venous catheter related infections using maximal sterile barrier precautions during insertion. Infect Control Hosp Epidemiol 1994;15:231–238. 93. Nehme AE. Nutritional support of the hospitalized patient: the team concept. JAMA 1980;243:1906–1908. 94. Faubion WC, Wesley JR, Khalidi N, et al. Total parenteral nutrition catheter sepsis: impact of the team approach. J Parenter Enteral Nutr 1986;10:642–645. 95. Freeman JB, Lemire A, MacLean LD. Intravenous alimentation and septicemia. Surg Gynecol Obstet 1972;135:708–712. 96. Nelson DB, Kien CL, Mohr B, et al. Dressing changes by specialized personnel reduce infection rates in patients receiving central venous parenteral nutrition. J Parenter Enteral Nutr 1986;10:220–222. 97. Tomford JW, Hershey CO, McLaren CE, et al. Intravenous therapy team and peripheral venous catheter-associated complications. A prospective controlled study. Arch Intern Med 1984;144:1191–1194. 98. Timset J-F, Sebille V, Farkas J-C, et al. Effect of subcutaneous tunneling on internal jugular catheter-related sepsis in critically ill patients. A prospective randomized multicenter study. JAMA 1996;276:1416–1420. 99. Hanna HA, McFadyen S, Marts K, et al. Prospective evaluation of 1.67 million catheter-days of peripherally inserted central catheters (PICCs) in cancer patients: long durability and low infection rate. In: Proceedings of the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy (abstract K-2045), Chicago, Illinois, September 22–25, 2001. Washington: American Society for Microbiology, 2001:439. 100. Raad II, Hanna HA, Marts K, et al. Nontunneled subclavian central venous (NTSC) vs. tunneled central venous catheters (CVCs) and ports in cancer patients. In: Programs and abstracts of the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, Illinois, September 22–25, 2001. Washington: American Society for Microbiology, 2001. 101. Shapiro JM, Bond EL, Garman JK. Use of a chlorhexidine dressing to reduce microbial colonization of epidural catheters. Anesthesiology 1990;73:625–631. 102. DeJong ME, DeBerardino MT, Brooks DE, et al. Antimicrobial efficacy of external fixator pins coated with a lipid stabilized hydroxyapatite/chlorhexidine complex to prevent pin tract infection in a goat model. Journal of Trauma-Injury Infection & Critical Care 2001;50(6): 1008–1014. 103. Garland JS, Alex CP, Mueller CD, et al. A randomized trial comparing povidone-iodine to a chlorhexidine gluconate-impregnated dressing for the prevention of central venous catheter infections in neonates. Pediatrics 2001;107:1431–1436. 104. Maki DG, Mermel LA, Kluger D, et al. The efficacy of a chlorhexidine-impregnated sponge (Biopatch) for the prevention of intravascular catheter-related infection: a prospective, randomized, controlled, multicenter study. In: Programs and Abstracts of the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, September 17–20, 2000. Washington: American Society for Microbiology, 2000. 105. Benoit J-L, Carandang G, Sitrin M, et al. Intraluminal antibiotic treatment of central venous catheter infections in patients receiving parenteral nutrition at home. Clin Infect Dis 1995;21:1286–1288. 106. Krzywda EA, Andris DA, Edmiston CE, et al. Treatment of Hickman catheter sepsis using antibiotic lock technique. Infect Control Hosp Epidemiol 1995;16:596–598. P.251 107. Schwartz C, Henrickson KJ, Roghmann K, et al. Prevention of bacteremia attributed to luminal colonization of tunneled central venous catheters with vancomycin-susceptible microorganisms. J Clin Oncol 1990;8:591–597. 108. Henrickson KJ, Axtell RA, Hoover SM, et al. Prevention of central venous catheter-related infections and thrombotic events in immunocompromised children by the use of vancomycin/ciprofloxacin/heparin flush solution: a randomized, multicenter, double-blind trial. J Clin Oncol 2000;18(6): 1269–1278. 109. Carratala J, Niubo J, Fernandez-Sevilla A, et al. Randomized, double-blind trial of an antibiotic-lock technique for prevention of gram-positive central venous catheter-related infection in neutropenic patients with cancer. Antimicrob Agents Chemother 1999;43(9): 2200–2204. 110. Rackoff WR, Weiman J, Jakobowski D, et al. A randomized, controlled trial of the efficacy of a heparin and vancomycin solution in preventing central venous catheter infections in children. J Pediatr 1995;127:147–151. 111. Raad I, Buzaid A, Rhyne, et al. Minocycline and EDTA for the prevention of recurrent vascular catheter infections. Clin Infect Dis 1997;25:149–151. 112. Raad I, Hachem R, Tcholakian RK, et al. Efficacy of minocycline and EDTA lock solution in preventing catheter-related bacteremia, septic phlebitis, and endocarditis in rabbits. Antimicrob Agents Chemother 2002;46(2): 327–332. 113. Bleyer A, Mason L, Raad I, et al. A randomized double-blind trial comparing minocycline/EDTA vs. heparin as flush solution for hemodialysis catheters. Infect Control Hosp Epidemiol 1999;21:100(abst). 114. Chatzinikolaou I, Zipf TF, Hanna H, et al. Minocycline-ethylenediamine-tetraacetate lock solution for the prevention of implantable port infections in children with cancer. Clin Infect Dis 2003;36:116–119. 115. Torres-Viera C, Thauvin-Eliopoulos C, Souli M, et al. Activities of taurolidine in vitro and in experimental enterococcal endocarditis. Antimicrob Agents Chemother 2000;44(6): 1720–1724. 116. Sodemann K, Polaschegg HD, Feldmer B. Two years' experience with Dialock and CLS (a new antimicrobial lock solution). Blood Purif 2001;19:251–254. 117. Jurewitsch B, Lee T, Park J, et al. Taurolidine 2% as an antimicrobial lock solution for prevention of recurrent catheter-related bloodstream infections. JPEN J Parenter Enteral Nutr 1998;22:242–244. 118. Shah CB, Mittleman MW, Costerton JW, et al. Antimicrobial activity of a novel catheter lock solution. Antimicrob Agents Chemother 2002;46(6): 1674–1679. 119. Bern MM, Lokich JJ, Wallach SR, et al. Very low doses of warfarin can prevent thrombosis in central venous catheters: a randomized prospective trial. Ann Intern Med 1990;112:423–428. 120. Randolph AG, Cook DJ, Gonzalez CA, et al. Benefit of heparin in central venous and pulmonary artery catheters: a meta-analysis of randomized controlled trials. Chest 1998;113:165–171. 121. Boraks P, Seale J, Price J, et al. Prevention of central venous catheter associated thrombosis using minidose warfarin in patients with haematological malignancies. Br J Haematol 1998;101:483–486. 122. Franschini G, Becker M, Bruso P, et al. Comparative trial of urokinase (uk) vs. heparin (h) as prophylaxis for central venous ports. In: Programs and abstracts of the 27th annual meeting of the American Society of Clinical Oncology (ASCO). Orlando, FL, ASCO: 1991:337(abst). 123. Ray CE Jr, Shenoy SS, McCarthy PL, et al. Weekly prophylactic urokinase instillation in tunneled central venous access devices. J Vasc Interv Radiol 1999;10:1330–1334. 124. Maki DG, Stolz SM, Wheeler SJ, et al. Prevention of central venous catheter-related bloodstream infection by use of an antiseptic-impregnated catheter. A randomized, controlled trial. Ann Intern Med 1997;127:257–266. 125. Pemberton LB, Ross V, Cuddy P, et al. No difference in catheter sepsis between standard and antiseptic central venous catheters. A prospective randomized trial. Arch Surg 1996;131:986–989. 126. Ciresi D, Albrecht RM, Volkers PA, et al. Failure of antiseptic bonding to prevent central venous catheter-related infection and sepsis. Am J Surg 1996;62:641–646. 127. Heard SO, Wagle M, Vijayakumar E, et al. The influence of central venous catheters coated with chlorhexidine/silver sulfadiazine on catheter-related infections. Crit Care Med 1997;25(suppl):A117(abst). 128. Raad I, Darouiche R, Hachem R, et al. The broad-spectrum activity and efficacy of catheters coated with minocycline and rifampin. J Infect Dis 1996;173:418–424. 129. Veenstra DL, Saint S, Saha S, et al. Efficacy of antiseptic-impregnated central venous catheters in preventing catheter-related bloodstream infection: a meta-analysis. JAMA 1999;281:261–267. 130. Rupp ME, Lisco S, Lipsett P, et al. Efficacy of chlorhexidine/silver sulfadiazine coating on microbial colonization of central venous catheters in a multicenter trial. In: Program and abstracts of the 41st annual meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, Illinois, September 22–25, 2001. 131. Brun-Buisson C, Nitenberg G, Doyon F, et al. Randomized controlled trial of antiseptic-coated (ACC) central venous catheters (CVC). In: Program and abstracts of the 42nd Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, California, September 27–30, 2002. 132. Darouiche RO, Raad II, Heard SO, et al., for the Catheter Study Group. A comparison of two antimicrobial-impregnated central venous catheters. N Engl J Med 1999;340(1); 1–8. 133. Hanna H, Banjamin R, Chatzinikolaou I, et al. The role of long-term silicone central venous catheters impregnated with rifampin and minocycline (S-CVC-RM) in preventing catheter-related infections: a prospective randomized study. In: Program and abstracts of the 38th Conference of the American Society of Clinical Oncology (ASCO), Orlando, Florida, May 18–21, 2002. 134. Sherertz RJ, Falk RJJ, Huffman KA, et al. Infections associated with subclavian Uldall catheters. Arch Intern Med 1983;143:53–56. 135. Boelaert JR, Van Landyt HW, Godard CA, et al. Nasal mupirocin ointment decreases the incidence of Staphylococcus aureus bacteraemia in haemodialysis patients. Nephrol Dial Transplant 1993;8:235–239. 136. Holton DL, Nicolle LE, Diley D, et al. Efficacy of mupirocin nasal ointment in eradicating Staphylococcus aureus nasal carriage in chronic haemodialysis patients. J Hosp Infect 1991;17:133–137. 137. Watanakunokorn C, Brandt J, Durkin P, et al. The efficacy of mupirocin ointment and chlorhexidine body scrubs in the eradication of nasal carriage of Staphylococcus aureus among patients undergoing long-term hemodialysis. Am J Infect Control 1992;20:138–141. 138. Reagan DR, Doebbeling BD, Pfaller MA, et al. Elimination of coincident Staphylococcus aureus nasal and hand carriage with intranasal application of mupirocin calcium ointment. Ann Intern Med 1991;114:101–106.
Воскресенье, 06 Ноябрь 2011 | 1340 Просмотры | Печать | PDF
HISTORY Legionnaires' disease made its debut in July 1976 as an explosive outbreak of community-acquired pneumonia. The outbreak of pneumonia was among attendees of the American Legion Convention at a hotel in Philadelphia, Pennsylvania (1). Six months later, the causative agent was isolated from the lung tissue of Legionnaires' cases by workers at the Centers for Disease Control and Prevention (CDC), Atlanta, Georgia (2). The microorganism, an aerobic gram-negative bacterium, was named Legionella pneumophila. The first known epidemic of nosocomial Legionella pneumonia was in July 1965 at St. Elizabeth's Hospital, a psychiatric institution in Washington, D.C. (3). In this outbreak, 81 patients were afflicted, with an attack rate of 1.4%. It was not until 1980 that hospital water distribution systems were first implicated as the source for nosocomial Legionnaires' disease. Tobin (4) isolated Legionella from showerheads in the hospital room of a patient with nosocomial Legionnaires' disease (4). Shortly thereafter, the microorganism was isolated from potable water distribution systems of numerous hospitals experiencing outbreaks of Legionnaires' disease (5, 6, 7, 8, 9). Among cases of Legionnaires' disease reported annually to the CDC from 1980 to 1998, the proportion of cases identified as nosocomial Legionnaires' disease varied from 25% to as high as 45% (10). Twenty-eight percent of these cases were associated with an outbreak of nosocomial Legionnaires' disease. MICROBIOLOGY The Legionellaceae family has been characterized as one monophyletic family belonging to the gamma subdivision of the class Proteobacteria (11, 12). Although a single genus and species (Legionella pneumophila) was originally proposed for the family Legionellaceae (13), the Legionellaceae family now contains more than 48 species and 70 serogroups in the genus Legionella (14, 15). Approximately half of these Legionella species have been implicated in human disease. Among the species, L. pneumophila is responsible for 90% of infections (Table 35.1) (14). These microorganisms are saprophytic water bacteria that can become opportunistic pathogens. Most cases of Legionellosis are caused by L. pneumophila serogroups 1, 4, and 6 (16). TABLE 35.1. LEGIONELLA PNEUMOPHILA IS RESPONSIBLE FOR THE MAJORITY OF INFECTIONS DUE TO LEGIONELLACEAE First Author, Year (Reference) Percentage of cases Serogroups L. pneumophila 1 Other Other Species Benin, 2002 (10) 91 51 9 9 Yu, 2002 (243) 92 84 7 9 Marston, 1994 (61) 91 57 13 9 Reingold, 1984 (244) 85 52 24 15 Other species implicated in human infection include L. micdadei (the Pittsburgh Pneumonia Agent), L. bozemanii, L. dumoffii, L. tucsonensis, L. cincinnatiensis, L. feeleii, L. longbeachae, and L. oakridgensis (17, 18, 19, 20). Most patients with nonpneumophila Legionella species infections have been severely immunocompromised due to corticosteroid therapy, organ transplantation, or malignancy (21, 22). Legionella species are small (0.3 to 0.9 Вm in width and approximately 2 Вm in length) faintly staining gram-negative rods with polar flagella (except L. oakridgensis). They generally appear as small coccobacilli in infected tissue or secretions, whereas long filamentous forms (up to 20 Вm in length) can be seen when they are grown in culture media. Legionellaceae are obligately aerobic slow-growing nonfermentative bacteria. They are distinguished from other saccharolytic bacteria by their requirement for L-cysteine and iron salts for primary isolation on solid media and by their unique cellular fatty acids and ubiquinones. Differences among species have been assessed by phenotypic (23, 24) and chemotaxonomic tests. Phenotypic tests include composition of lipopolysaccharides (LPS) (24), electrophoretic protein profiles (25), monoclonal antibodies (26), fatty acid composition (27), and cellular carbohydrates (28). Genotypic tests include random amplified polymorphic DNA profiles (RAPD) (29, 30), heteroduplex analysis of 5S ribosomal RNA (rRNA) gene sequences (31), and computer-assisted matching of transfer DNA (tDNA)-intergenic length polymorphism (ILP) patterns (32). The microorganism can be visualized, with some difficulty, with Gram stains of clinical specimens taken from normally sterile sites (e.g., pleural fluid). Both the Gram and Gimenez stains can be used for clinical specimens, whereas silver impregnation stains, including the Dieterle and Warthin-Starry stains, can be used for paraffin-fixed tissue sections. L. micdadei (Pittsburgh pneumonia agent) can stain weakly acid-fast in tissue with Kinyoun and Fite stains and on smears with a modified acid-fast stain in tissue or sputum specimens. These microorganisms are nutritionally fastidious and do not grow on standard bacteriologic P.604 media, which explains why the microorganism was so difficult to isolate in the original American Legion outbreak. PATHOGENESIS Pneumonia is the presenting clinical syndrome in almost all cases of nosocomial legionellosis. Although rare, extrapulmonary Legionella infection has been documented (33) Legionnaires' disease can be acquired by the inhalation of aerosols containing Legionella or by aspiration of water or respiratory secretions containing Legionella (34, 35). Other possible modes of transmission include direct inhalation or hematogenous dissemination from other foci of infection. Cigarette smokers, patients with chronic pulmonary disease, and alcoholics—all conditions in which mucociliary clearance is impaired—are at increased risk for Legionnaires' disease. This barrier to entry can be overcome by adherence of the microorganism to respiratory epithelial cells. Legionella does possess pili that are known to mediate adherence to epithelial cells (36). Symbiosis has also been shown in vitro between oropharyngeal flora and Legionella (37). Legionella is an intracellular pathogen both in humans and in aquatic environments (38). Legionella survives and multiplies as parasites of single-celled protozoa in fresh water and moist soil (39). Virulence may be increased by replication in amebae (40). In humans, Legionella replicates within mononuclear phagocytes, primarily monocytes, and alveolar macrophages (41, 42). Cell-mediated immunity plays the central role in host defense against L. pneumophila as it does against other intracellular pathogens. Although the resident alveolar macrophage normally degrades most microorganisms, Legionella is able to subvert this host defense. The macrophage readily phagocytoses Legionella, a process that is more avid in the presence of specific opsonizing antibody. Attachment of Legionella to epithelial cells or macrophages may be related to the expression of pili. A pilin mutant showed a 50% decrease in attachment to human macrophages and epithelial cells (36). Once inside the cell, the microorganism evades phagosome-lysosome fusion, converts to a replicative form that is acid-tolerant, and multiplies until the cell ruptures (38, 43, 44). Presumably, the liberated bacteria are phagocytosed by newly recruited cells, and the cycle of ingestion, multiplication, and liberation with cell lysis begins anew. Intracellular multiplication of Legionella within human monocytes depends on the availability of iron (45). The lymphokine interferon-Оі (IFN-Оі) stimulates human alveolar macrophages and monocytes to resist Legionella infection by upregulating reactive oxygen production and downregulating cellular iron content. Other cytokines and hemopoietic growth factors, such as interleukin-10 (IL-10) and granulocyte-macrophage colony-stimulating factor (GM-CSF), have not been shown to enhance anti-Legionella activity (46, 47). Significant rises in the Th-1 cytokines IFN-Оі and IL-12 were detected in the serum of patients with Legionnaires' disease, supporting the importance of cellular immunity in this disease (48). Neutrophils are less important, and neutropenic patients are not at undue risk for Legionnaires' disease. Nevertheless, L. pneumophila is susceptible to oxygen-dependent microbicidal systems in vitro. Neutrophils inhibit Legionella growth but lack the capacity to kill L. pneumophila. Lysis of infected macrophages by lymphokine-activated killer (LAK) cells or natural killer (NK) cells may also be an important cell-mediated immune function for eliminating intracellular Legionella. Humoral immunity is also important. For patients with Legionnaires' disease, type-specific antibodies are measurable within several weeks of infection. Moreover, immunized animals and patients develop a specific antibody response with subsequent resistance to Legionella challenge. A number of factors have been postulated to contribute to the virulence of L. pneumophila(38): type I and type II secretion systems, a pore-forming toxin, type IV pili, flagella, a Legionella toxin (49), a 24-kd protein called Mip (50), a zinc metalloprotease (51, 52, 53), and proteases (54, 55) including enzymes that scavenge reduced-oxygenated metabolites (56). Strains of L. pneumophila differ in virulence. L. pneumophila, serogroup 1, is known to cause most cases of Legionnaires' disease. Although multiple strains of L. pneumophila serogroup 1 may colonize water distribution systems, only a few strains are likely to cause disease in patients exposed to the water (8, 57, 58). Monoclonal antibody subtyping of strains of L. pneumophila, serogroup 1, have shown that a surface epitope recognized by one particular monoclonal antibody (MAB-2) may be associated with virulence. The immunodominant part of this virulence-associated epitope has been identified as the 8-0-acetyl group of the 0-specific polysaccharide chain of the LPS (59). Legionella species other than L. pneumophila appear to be less virulent and occur almost exclusively among immunocompromised hosts. They also respond more readily to antibiotic therapy (20, 60). P.605 EPIDEMIOLOGY Although legionellosis is a reportable disease in many countries including the United States, the extent of this infection is still uncertain. Underestimates are likely due to cases that are overlooked because of the persistent lack of availability of specialized laboratory tests. From 1980 through 1998, the median number of cases reported per year to the CDC was 360 (10, 61, 62). If the incidence of Legionnaires' disease is estimated to be a minimum of 8,000 to 14,500 cases, then less than 5% of cases are being reported. Approximately 35% of the reported cases met the definition for nosocomial infection (10). Nosocomial legionellosis has been classified as “endemic” and “hyperendemic.” These terms have become less useful with the recognition that they merely designate different prevalences of the disease over a spectrum. Thus, although hospitals may be labeled as experiencing sporadic disease (implying frequent cases of Legionnaires' disease scattered over a long period), the possibility is that only a proportion of actual cases are diagnosed. These cases surface because of a combination of circumstances: improved diagnostic methods, clinical suspicion of Legionnaires' disease by an individual physician, or isolation of the microorganism from open lung biopsy or postmortem lung culture (22). The introduction of a diagnostic test for Legionnaires' disease, the urine antigen test, was responsible for the detection of a recurrent outbreak of nosocomial Legionnaires' disease at a hospital in Connecticut (63). From 1987 to 1996, routine testing for Legionnaires' disease at autopsy identified eight cases of nosocomial Legionnaires' disease at a regional transplant center in the southwestern U.S. (64). The occurrence of three cases in early 1996 led to a retrospective review, which suggested that nosocomial transmission had occurred for more than 17 years. An additional 14 cases were identified for a total of 25 culture-confirmed cases of nosocomial Legionnaires' disease. Thus, situations labeled as sporadic or nonepidemic may merely represent chance discovery of disease occurring at a low endemic level. Likewise, situations labeled as “epidemic” may merely represent a cyclical peak at a hospital with endemic but previously undiscovered disease. The reported nosocomial Legionella infection rates vary widely from 1% to 40% (65, 66, 67, 68). This variable infection rate reflects a dependence on multiple variables. These include a contaminated potable water system with Legionella, exposure of the host to the contaminated water, susceptibility of the patient exposed, and recognition of the disease by the physician. Consistently identified risk factors for Legionnaires' disease include advanced age, male gender, smoking, alcohol abuse, chronic pulmonary disease, and immunosuppression (malignancy, corticosteroid use). Males are affected at two to three times the rate of women; this may be related to cigarette smoking or underlying medical conditions (e.g., chronic obstructive pulmonary disease). Attributable mortality for Legionnaires' disease is approximately 20%; however, the likelihood of death from Legionella infection increases in patients who are elderly or male, with nosocomial infection, renal disease, malignancy, or immunosuppression (61, 64). Mortality can be as high as 40% for hospital-acquired cases (61). Nosocomial infections due to Legionella occur most frequently in immunosuppressed hosts. The patients at highest risk are organ transplant recipients (69). During an outbreak in an acute care hospital, 55% (5/9) of all patients undergoing kidney transplantation developed Legionnaires' disease over a 5-month period (70). Nosocomial Legionella infection has been reported in transplant recipients of kidneys (70, 71), hearts (64, 72), livers (73, 74), and bone marrow (64, 74, 75). Corticosteroids are an important independent risk factor. Neoplastic disease, diabetes, and renal failure are often cited as risk factors. The broader use of diagnostic testing may result in more patients being identified without these classic risk factors. A retrospective review of over 400 cases of Legionnaires' disease in the Pittsburgh area showed that 25% of reported cases did not have the classic risk factors (76). There is a striking association of Legionnaires' disease with surgery. Up to 40% of cases reported in the literature occurred in surgical patients (77). Nosocomial Legionella infection increased with use of general anesthesia and endotracheal intubation (64, 66, 78, 79). Surprisingly, neutropenic or leukemic hosts appear to have an attack rate no higher than that of the general population. The exception are patients with hairy cell leukemia (80, 81). Likewise, the risk of Legionella infection in the HIV-infected patient appears to be no greater than other high-risk populations, with reports of less than 1% to 4%. However, these patients are prone to extrapulmonary manifestations, bacteremia, and lung abscesses. Increasing use of diagnostic tests for Legionella has led to new risk groups of patients being discovered as susceptible victims for Legionnaires' disease. They include immunocompromised children in pediatric hospitals colonized with Legionella and elderly patients residing in long-term-care facilities and rehabilitation centers colonized by Legionella. Nosocomial cases have been reported in immunosuppressed children (74, 82, 83, 84, 85, 86, 87) and children with underlying pulmonary disease (88, 89). In three hospitals in which epidemiologic investigations were conducted (82, 88, 90), a link to the hospital water supply was made. At least three outbreaks of Legionnaires' disease have been reported in long-term-care facilities (91, 92, 93). The investigation identified Legionella in the potable water supply as the source for two of the outbreaks. In a third outbreak, only limited environmental sampling was performed. Aspiration was presumed to be the mode of transmission for most of these outbreaks. In one outbreak, eating pureed food was a significant risk factor for Legionella, consistent with aspiration originating from a swallowing disorder (93). In another prospective study, L. pneumophila serogroup 1 was isolated from a newly constructed long-term-care facility (94). Six cases of Legionnaires' disease were diagnosed over 2 years. DNA subtyping established that the patient isolates were identical to the environmental isolates from the water supply. Reservoir The environmental ecology of Legionella is particularly pertinent in that Legionnaires' disease is a pneumonia that theoretically could be prevented with eradication of the microorganism P.606 from its reservoir. The natural habitat for Legionella appears to be aquatic bodies including rivers, streams, and thermally polluted waters, although L. longbeachae has been isolated from moist soil in Australia (95). Natural aquatic bodies contain only small numbers of Legionella. Since Legionella tolerates chlorine, the microorganism easily survives the water treatment process and passes into water distribution systems but, again, only in small numbers (96, 97, 98). Subsequent growth and proliferation occur in man-made habitats, especially water distribution systems, which provide favorable water temperatures (25В to 42ВC), physical protection (biofilm), and nutrients (99). The single most important factor appears to be temperature. The microorganism is most readily found at the bottom of hot water tanks—a relation that parallels its propensity for colonization in thermally polluted rivers. Interestingly, bacteria populating hot water tanks were more likely to demonstrate a symbiotic relationship with L. pneumophila than bacteria populating cold water tanks (100). Bacteria, protozoa, and amoeba also colonize water pipe surfaces, some of which have been shown to promote Legionella replication (100, 101, 102). Legionella and other microorganisms attach to surfaces and form biofilms on pipes throughout the water distribution system (103). Water pressure changes that disturb the biofilm may dramatically increase the concentration of Legionella (104). Hospitals with hot water distribution systems colonized with L. pneumophila were significantly more likely to have lower water temperatures (<140ВF), have a vertical configuration, be older, and have elevated calcium and magnesium concentrations in the water (105, 106). Cold-water sources, such as ice machines, have also been implicated as a source of nosocomial infection (97, 107, 108). The role of Legionella-contaminated potable water distribution systems as a source for nosocomial Legionnaires' disease has been well established. The British Communicable Disease Surveillance Centre reported that 19 of 20 hospital outbreaks of Legionnaires' disease in the United Kingdom from 1980 to 1992 were attributed to such systems (109, 110). Cooling towers and, to a lesser degree, evaporative condensers were implicated in the earlier outbreaks prior to recognition of potable water as a reservoir. Surprisingly, air conditioners have never been directly implicated as a source of Legionnaires' disease, despite widespread belief that they are. The role of cooling towers in the dissemination of Legionella has been challenged (111). Following the recognition in 1982 that potable water distribution systems were a source (6), reports of cooling towers as reservoirs for nosocomial legionellosis have essentially disappeared. One exception was a report published in 1985 of a Rhode Island hospital in which cooling towers were cited as the source (112); this now appears to be a typical scenario of water distribution system contamination in which the original epidemiologic investigation was flawed (104). Subtyping of L. pneumophila with molecular methods has proven invaluable in elucidating environmental sources, permitting application of rational methods for prevention (113). In fact, application of subtyping provided the first concrete evidence that water distribution systems rather than cooling towers were the actual sources of infection (9, 114). The subtype of Legionella isolates taken from patients were identical to the isolates taken from putative environmental reservoirs. Both phenotypic and genotypic methods have been used to demonstrate identity among strains of Legionella pneumophila in epidemiologic investigations. These methods include serotyping, monoclonal antibody subtyping, isoenzyme analysis, protein and carbohydrate profiling, plasmid analysis, restriction endonuclease analysis, restriction fragment length polymorphism (RFLP) of rRNA (ribotyping) or chromosomal DNA, amplified fragment length polymorphism (AFLP), restriction endonuclease analysis of whole-cell DNA with or without pulsed-field gel electrophoresis (PFGE), and DNA fingerprinting using polymerase chain reaction (PCR) (113, 115, 116, 117, 118, 119). However, PFGE has been the most widely applied (117, 120, 121, 122). Maximum discrimination among isolates is achieved by combining both monoclonal antibody subtyping and PFGE (117, 123). Modes of Transmission Multiple modes have been identified for transmission of Legionella to humans; there is evidence for aerosolization, aspiration, or even instillation into the lung during respiratory tract manipulation. Aspiration of contaminated water or oropharyngeal secretions appears to be the major mode of transmission in the hospital setting (35). Colonization of oropharyngeal flora by L. pneumophila is a theoretical possibility (124, 125, 126, 127, 128). The evidence for aspiration is impressive. Legionella was found to be the most common cause of nosocomial pneumonia in a population of oncologic head and neck surgery patients (65); these patients had a propensity for aspiration as a result of their oral surgery and extensive cigarette smoking. Nasogastric tube placement has been shown to be a significant risk factor for nosocomial legionellosis in intubated patients in three studies; microaspiration of contaminated water was the presumed mode of entry (34, 129, 130). It should be noted that, in the original 1976 outbreak, consumption of water at the implicated hotel was associated with acquisition of disease—an association that has been generally overlooked (1). Contaminated ice and water from an ice machine have been implicated as the source of nosocomial infection (97, 108, 131). Healthcare personnel frequently use tap water to rinse respiratory apparatus and tubing for use in mechanical ventilation machines. If the tap water is contaminated with L. pneumophila, the microorganism can be instilled directly into the lung (132, 133, 134). In numerous studies, patients with Legionnaires' disease underwent endotracheal tube placement significantly more often or had a significantly longer duration of intubation than patients who had other causes of pneumonia (64, 66, 78, 135, 136). The use of a nasogastric tube, the presence of immunosuppression, and ventilator use were highly correlated with the acquisition of nosocomial Legionnaires' disease in a hospital in Halifax, Nova Scotia (129). Use of sterile water for all nasogastric suspensions and for flushing tubes has been recommended to prevent nosocomial Legionnaires' disease. Intermittent positive pressure ventilators have been associated with nosocomial legionellosis, or more likely, the tubing attached to these ventilators. The use of such equipment was epidemiologically linked to Legionnaires' disease in 18 hospital patients over a 2-year period; again, it was noted that the equipment was rinsed with tap water P.607 between treatments (34). Three cases of nosocomial L. pneumophila pneumonia were acquired from contaminated transesophageal echocardiography (TEE) probes (137). Again, contaminated tap water had been used to rinse the probes. Investigators from the CDC presented the first evidence to support the aerosolization theory when reporting the Legionnaires' disease outbreak in Memphis (138). Tracer smoke studies indicated that aerosols from an auxiliary air conditioning tower could have reached an air intake supplying certain patient rooms. However, the attack rate for patients occupying rooms supplied with air from the air intake was not higher than the attack rate for patients occupying rooms in the same wing but receiving air from other sources (111). Cases also occurred in hospital wings having no relationship to the cooling towers. Water was not cultured since this investigation antedated the discovery that drinking water could be the source for Legionnaires' disease. Because the first environmental isolation of L. pneumophila was from a showerhead (4), it has been widely thought that aerosols from showers may be an important means for dissemination of this microorganism. However, simulation studies show that only small numbers of Legionella are aerosolized and only for short distances (132, 139). Although a few retrospective studies have suggested showers as a potential source (140, 141), an epidemiologic link between showering and acquisition of disease has never been shown in prospective studies; in fact, prospective studies have consistently shown that showers are not a risk factor (34, 64, 142, 143, 144, 145). Aerosolization by respiratory tract devices including the humidifier of oxygen therapy equipment, nebulizers, and room humidifiers has been documented (146, 147). Humidifiers are water-filled devices that add water vapor to air, oxygen, or other gases without producing particulate water. Guinea pigs exposed to a room humidifier contaminated withLegionella experienced subclinical infection as demonstrated by seroconversion. In a hospital setting, a portable room humidifier filled with Legionella-contaminated tap water disseminated the microorganism up to distances of 300 cm. Furthermore, recovery of aerosolized Legionella increased with proximity to the humidifier, and seroconversion of exposed animals was directly proportional to the concentration of Legionella in humidifier water. Humidifiers have been implicated in transmission of Legionnaires' disease in humans. Five of eight patients with nosocomial Legionnaires' disease in an Italian hospital had been exposed to bubble diffuser humidifiers filled with water containing L. pneumophila (148). An immunosuppressed patient at the University of Chicago Hospital acquired Legionnaires' disease after exposure to a room humidifier that had been filled with contaminated tap water for 15 days (149). The statistical association between disease and humidifier exposure was highly significant. Use of a room humidifier was also associated with 18 cases of nosocomial Legionnaires' disease in a 2-year period in a limited retrospective study (150). In all three of these studies, the humidifiers had been filled with tap water (148, 149, 150). A postlaryngectomy patient died from pneumonia following exposure to a room humidifier. L. pneumophila serogroups 4 and 5 were isolated from the patient's lung and from the tap water and containers used to fill the humidifier reservoir (133). Distilled water in humidifiers has also been linked to hospital outbreaks of Legionella infection; one patient with L. dumoffii was exposed to a room humidifier presumably filled with contaminated distilled water (151). Nosocomial pneumonia in a neonate was linked to the presence of Legionella in the humidifier of the incubator (152). In one French hospital, the use of contaminated tap water to fill the humidifier of oxygen therapy equipment and for aerosol delivery of drugs led to five cases of Legionnaires' disease caused by L. pneumophila, serogroup 1 (147). Nebulizers are devices that generate aerosols of uniform particulate size. Ultrasonic nebulizers can produce water particles ranging in size from 0.9 to 10 Вm; water droplets of 1 to 2 Вm in diameter can reach the alveoli. Medication jet nebulizers have been shown to aerosolize water particles containing L. pneumophila when the nebulizer water was seeded with the microorganism (153); these particles were less than 5 Вm in diameter, so it is likely they could bypass the pulmonary defenses and reach the alveoli. Jet nebulizers have been epidemiologically linked to nosocomial Legionnaires' disease (149). Inhalation of contaminated tap water aerosols from jet nebulizers was found to be a highly significant risk factor for four patients who acquired nosocomial Legionnaires' disease. In addition to filling nebulizers with tap water, rinsing the chambers of hand-held medication nebulizers has been suggested as a source of contamination. In one study of 13 patients with nosocomial Legionnaires' disease due to L. pneumophila, serogroup 3, there was a trend toward more frequent use of nebulizer medications in patients with Legionnaires' disease. It was subsequently established that jet nebulizers were often rinsed with tap water (153). Medication nebulizers have also been implicated in one of the few reports of pediatric nosocomial Legionella infection (82). Two children with Legionnaires' disease received nebulizer treatments using equipment likely to have been rinsed under tap water. Aerosolization via excavated soil was suggested as a possible mode of transmission for the outbreaks at the Wadsworth Veterans Administration (VA) Medical Center and St. Elizabeth's Hospital; in retrospect, contaminated water distribution systems were probably the actual reservoirs. Finally, person-to-person transmission has not been demonstrated (154). CLINICAL MANIFESTATIONS Legionella infection presents as two clinical entities: Pontiac fever and pneumonia (Legionnaires' disease). Pontiac fever is an acute, self-limiting illness. Chills, high fever, headache, and myalgias are typical. Pneumonia is not seen, and nosocomial cases of Pontiac fever have not been reported. Pneumonia is the predominant clinical syndrome in Legionnaires' disease. The incubation period for Legionnaires' disease usually ranges from 2 to 10 days. One report demonstrated the onset of disease 63 days after discharge from the hospital, and molecular typing linked the hospital water supply as the source. This led to the speculation that oropharyngeal colonization with Legionella had occurred (127). Subsequent studies have not been successful in demonstrating oropharyngeal colonization with Legionella (125, 155). P.608 Legionnaires' disease encompasses a broad spectrum of illnesses ranging from mild cough and low-grade fever to stupor, rapidly progressive pneumonia, and multiorgan system failure. Nonspecific symptoms including malaise, myalgias, anorexia, and headache are common in the first 48 hours. Fever is virtually always present, and temperatures in excess of 40ВC should lead to the consideration of Legionnaires' disease. Relative bradycardia has been emphasized by some investigators in earlier studies, but we have found this to be a nonspecific finding (156). Initially, the cough is mild and only slightly productive. The character of the sputum is often nonpurulent. Although the sputum may be streaked with blood, gross hemoptysis is rare. Chest pain, often pleuritic, is common, and when coupled with hemoptysis, can masquerade as pulmonary infarction. Gastrointestinal symptoms are more prominent in community-acquired pneumonia, but less so in nosocomial pneumonia; diarrhea, nausea, vomiting, and abdominal pain are common. The most common neurologic finding in Legionnaires' disease is change in mental status, although a wide variety of findings, including encephalopathy, have been reported (157, 158). L. pneumophila microorganisms can disseminate from their pulmonary niche to various extrapulmonary sites including spleen, liver, kidney, bone marrow, myocardium, and lymph nodes. Dissemination apparently occurs via the hematogenous or lymphatic system. Extrapulmonary nosocomial Legionella infections occurred in cardiothoracic surgical patients at Stanford University (33, 79, 159). Seven patients presented with Legionella prosthetic valve endocarditis, three had sternal surgical site infections, and one patient manifested both infections. L. pneumophila, serogroup 1, and L. dumoffii were isolated from clinical samples as well as from the potable water system of the hospital. The origin of the sternal surgical site infections was contaminated tap water used to remove the povidone-iodine solution from the operative site. Other reports have implicated tap water as the source for extrapulmonary Legionella infections. In one patient, an open hip wound infection due to L. pneumophila was linked to colonized water from a Hubbard tank used for rehabilitation (160). Nosocomial extrapulmonary legionellosis involving hemodialysis fistula infections (two cases) (161) and a perirectal abscess (162) were probably secondary to hematogenous seeding from confirmed Legionella pneumonia; however, direct inoculation by contaminated water or equipment could not be excluded. Detection of the microorganism at extrapulmonary sites is problematic. Since selective media must be used to isolate the microorganism, the clinician must think of the possibility of Legionella as the cause of the infection. Other bacteria may also be isolated, thereby confounding the diagnosis. LABORATORY DIAGNOSIS The prompt diagnosis of Legionnaires' disease in the hospital setting can save lives. Not only has early initiation of appropriate therapy been associated with improved outcome, but the diagnosis of a single case of hospital-acquired Legionnaires' disease can prompt the recognition of endemic Legionnaires' disease at the facility (63, 163). For patients with severe pneumonia, the Infectious Diseases Society of America recommends diagnostic tests for Legionella (164). The diagnosis of Legionnaires' disease based on a syndromic approach has been suggested (158); however, most studies have shown that the clinical manifestations of Legionnaires' disease are nonspecific (156). Laboratory abnormalities including abnormal liver function tests, elevated creatinine phosphokinase, hypophosphatemia, hematuria, hemolytic anemia, and thrombocytopenia have been reported. Hyponatremia with a serum sodium of less than 130 mEq/L occurs significantly more often in Legionnaires' disease than in other pneumonias; it appears to be more common in nosocomial Legionnaires' disease than in the community-acquired disease. This syndrome probably is caused by salt and water loss rather than inappropriate antidiuretic hormonal secretion (V. L. Yu, unpublished data). Specialized diagnostic laboratory tests are the key feature for diagnosing Legionnaires' disease, because the clinical presentation is nonspecific. Most hospitals, including university and tertiary care hospitals, often do not have the most sensitive tests available, namely culture on selective media and urinary antigen (110), and up to 40% of hospitals send samples off-site for testing (165). Data from a CDC survey showed that hospitals where Legionella diagnostic tests were available on-site were more likely to identify nosocomial Legionnaires' disease (165). Urinary Antigen Among case reports of Legionnaires' disease submitted to the CDC, there has been a significant increase in the proportion of patients reported with a positive urine antigen test result (10). The Legionella urine antigen test has a high sensitivity (90%), high specificity (99%), and relatively low cost, and the results can be available within hours of submission of the test (166, 167, 168). The urine antigen test is available as an enzyme immunoassay (EIA) or an immunochromatographic (ICT) test. The EIA test is available commercially from two U.S. suppliers (Wampole Laboratories, a division of Carter-Wallace Inc., Cranbury, NJ; and Bartels, Issaquah, WA) and includes the Binax Legionella Urinary Antigen EIA and the Bartels Legionella Urinary Antigen EIA (Intracel, Frederick, MD). The Binax NOW Legionella urinary antigen test is a rapid ICT membrane assay for the qualitative detection of L. pneumophila serogroup-1 antigen (Binax, Inc., Portland, Maine). A swab is dipped in urine and inserted into the test device, and the reagent is added. The reaction is read after 15 minutes as the presence or absence of a visually detectable pink-purple colored line that results from the antigen-antibody reaction giving the result (Fig. 35.1) (169). The EIA and ICT tests have been shown to have comparable sensitivity and specificity (166, 170). Figure 35.1. The NOW immunochromatographic (ICT) test is performed by dipping a swab in urine and inserting it into the test device. Two drops of a reagent are added and the card is closed and allowed to react for 15 minutes. A positive result is the presence of a visually detectable pink-purple colored line (next to the “Patient” line) resulting from the antigen-antibody reaction. The fact that test positivity can persist for days, even during administration of antibiotic therapy, makes it useful in those patients who receive empiric anti-Legionella therapy (171). A shortcoming of the test is that it can detect only serogroup 1 of L. pneumophila (172). Since the other serogroups of L. pneumophila and other Legionella species are less common, this test is still extremely useful. There has been some evidence that the urine antigen test can detect other serogroups and species; however, this requires further validation (172). In addition, it is P.609 often easier to obtain a urine specimen than an adequate sputum specimen. A positive urinary antigen test, along with culture positivity and seroconversion, is now one of the criteria for a definitive diagnosis of Legionnaires' disease (173). Early diagnosis and treatment has resulted from the increased use of the rapid urinary antigen test. This, in combination with the increasing empiric use of quinolones for hospital-acquired pneumonia, may explain the decline in Legionnaires' disease-related mortality in the U.S. The case-fatality rate for nosocomial Legionnaires' disease has decreased from 46% in 1982 to 14% in 1998 (10). Culture on Selective Media When Legionnaires' disease is suspected, both a urinary antigen test and Legionella culture of a respiratory specimen should be ordered. The single most important diagnostic test for Legionnaires' disease is isolation of the microorganism by culture. The availability of the clinical isolate from culture can be critical for subsequent epidemiologic investigations (174). Another reason not to rely exclusively on the urine antigen test is that the urinary antigen test may be negative if the infecting strain is not serogroup 1 or when the infecting strain is serogroup 1 but MAB-2 negative (Dresden Panel MAB-3/1 negative). Among 317 culture-proven cases of Legionnaires' disease, 67 (21%) were nosocomial cases. Only 45% of these cases were urine antigen positive, because 22% of the cases were caused by the MAB-2 negative serotype (175). To achieve a high yield from sputum, multiple media containing antibiotics and dyes are required (169, 176, 177, 178). Buffered charcoal yeast extract (BCYE) agar is the primary medium used for isolation of these microorganisms. The culture media can be made more selective by incorporating antibacterial agents (cefamandole, polymyxin B, vancomycin, aztreonam), antifungal agents (anisomycin), and inhibitors (glycine) into the media to suppress competing microflora. Pretreatment with acid is extremely useful for respiratory tract and environmental specimens, because Legionella microorganisms are acid-resistant, whereas most other bacteria are not. The addition of dyes to the media enhances the visibility of the colonies, because Legionella takes up the dye preferentially. The dye-containing media are especially important in detection of the nonpneumophila species (178). The microorganism grows slowly, taking up to 5 days for visible colonies to develop. Under a dissecting stereomicroscope, the colony surface shows a characteristic ground glass appearance. Legionella culture is performed only when specifically requested. A physician often orders a Legionella urinary antigen test, and only a routine microbiology culture. As a result, when the urine antigen test is positive, no sputum is available for Legionella culture. We refrigerate all respiratory specimens for 7 days by placing them in bins marked by the days of the week. This practice allows for subsequent retrieval of the specimen for Legionella culture if a urine test is positive. The isolate from the patient is now available if an epidemiologic investigation is performed to determine the source of the infection. Transtracheal aspirate specimens that bypass contaminating oropharyngeal flora can achieve a sensitivity as high as 90% (176). Sputum obtained by bronchoscopy can be useful but does not provide any higher yield than a good sputum specimen. If sputum is not available, however, bronchoalveolar lavage can yield the microorganism. Bronchial washings, in which the volume of fluid instilled is notably lower than that of lavage, appear to be less sensitive. Transbronchial biopsy can yield the microorganism P.610 in tissue by direct fluorescent antibody stains and culture and has been successful in identifying Legionella when sputum and bronchial washings were unrevealing. Percutaneous needle aspiration of a lung abscess has yielded the microorganism in culture from a patient who had negative sputum and bronchoscopy cultures. Bacteremia is actually common in severely ill patients. The microorganism can be isolated from blood by biphasic BCYE agar bottles, a radiometric system (Bactec, Johnston Laboratories, Towson, MD), or VACUTAINER tube (Becton Dickinson, Rutherford, NJ). In one study, 38% of cases of Legionnaires' disease had positive blood cultures when subcultures from Bactec bottles were plated onto buffered charcoal yeast extract agar (179). At the Pittsburgh VA Medical Center, an aliquot (0.1 mL) from all negative blood culture bottles is plated to BCYE prior to being discarded. Direct Fluorescent Antibody (DFA) Stain The reported sensitivity of direct fluorescent antibody stains has ranged from 25% to 75% (180). It is highly specific, and the monoclonal antibody test (MONOFLUO, Bio-Rad Laboratories, Redmond, WA) has eliminated the rare occurrence of cross-reactivity with other gram-negative bacilli. Due to low sensitivity compared to culture, we do not perform the DFA on a specimen unless the direct culture is overgrown with competing flora and acid pretreatment of the specimen is required. Polyclonal DFA reagents are available from a number of suppliers for definitive identification of isolates of Legionella (Monoclonal Technologies, Atlanta, GA; Meridian Diagnostics, Inc., Cincinnati, OH; Zeus Technologies, Raritan, NJ). Serology Antibody tests have become less important with the advent of rapid diagnostic tests. Because the definitive criterion for diagnosis is a fourfold rise in antibody titer, repeat serology is required 4 to 6 weeks after onset of infection. Sensitivity in the 1976 outbreak was 91% (181), but sensitivity in studies of nosocomial pneumonia has been less than 50% (34, 182, 183). Maximal sensitivity requires detection of both immunoglobulin G (IgG) and IgM antibody (14). Effective antibiotics and suboptimal timing of specimen collection are possible reasons for the decrease in sensitivity. Diagnosis of Legionnaires' disease by serologic testing has decreased significantly from 1980 to 1998 (10). Polymerase Chain Reaction (PCR) DNA amplification by PCR of Legionella has been reported from patients with pneumonia using throat swab specimens, bronchoalveolar lavage (BAL), urine, and serum (79, 184, 185, 186). Primer sequences of the macrophage infectivity potentiator (mip) gene of L. pneumophila and the 5S rRNA or 16S rRNA have been utilized in PCR assays. A real-time quantitative PCR assay has been used to detect L. pneumophila in respiratory tract secretions (187). One PCR kit was used successfully to detect Legionella in both clinical and environmental samples (188, 189), but it is no longer commercially available. Although Legionella DNA has been detected in urine and serum samples from patients with legionellosis (190), clinical experience has not shown PCR to be more sensitive than culture. Therefore, the CDC does not recommend the routine use of genetic probes or PCR for detection of Legionella in clinical samples (180, 191). PREVENTION It is now well established that there is a direct relationship between colonization of hospital water systems with L. pneumophila and the occurrence of nosocomial Legionnaires' disease (192, 193, 194, 195). Legionella species have been shown to colonize between 12% and 85% of hospital water systems (193, 196). Prospective studies have demonstrated cases of nosocomial Legionnaires' disease in colonized hospitals after environmental and clinical surveillance were initiated (193). Knowledge of this relationship is the first step to prevention. Unfortunately, there is a lack of consensus with respect to the utility of environmental monitoring for Legionella as part of a prevention strategy (193). There are essentially two approaches to prevention of hospital-acquired Legionnaires' disease. One approach suggests maintaining a high index of suspicion for Legionellosis with the use of diagnostic testing in patients with healthcare-associated pneumonia (191, 197). Routine culturing of the hospital water system for Legionella is not initiated unless one case of definite or two cases of possible hospital-acquired pneumonia have been identified. We consider this approach to be “re-active,” and it is favored by many public health authorities including the CDC. An alternate and “proactive” approach has been advocated by Pittsburgh investigators and the Allegheny County Health Department in Pittsburgh, Pennsylvania, for many years (193). This approach recommends proactively culturing the hospital water system as the initial step in making a risk assessment of the facility. Guidelines for Legionella prevention from the Allegheny County Health Department and from the state of Maryland specifically recommend routine environmental monitoring of the hospital water system (198, 199) (Table 35.2). If any outlets yield L. pneumophila, diagnostic tests for Legionella are made available in-house. The presence of L. pneumophila serogroup 1 in the water supply necessitates the on-site availability of the urinary antigen test. If greater than 30% of outlets are culture-positive for L. pneumophila, the Allegheny County guidelines recommend that the facility consider disinfection of the water system (198). TABLE 35.2. GUIDELINES FOR CONTROL OF LEGIONELLA IN HEALTHCARE FACILITIES IN PENNSYLVANIA AND MARYLAND RECOMMEND ROUTINE ENVIRONMENTAL CULTURING OF THE HOSPITAL WATER SYSTEM FOR LEGIONELLA The Texas Department of Health has also issued guidelines that recommend environmental surveillance for Legionella only if a risk assessment indicates that the facility has a significant risk of legionellosis transmission (200). For example, a high-risk facility could be a multistory facility with multiple water distribution systems, supplied with water treated with chlorine, stored hot water at 51ВC (124В F) and delivered at 43ВC (110ВF), and housing bone marrow or solid organ transplant recipients or cancer patients undergoing chemotherapy. Proactive approaches mandating routine environmental cultures within hospitals have now been adopted in Denmark, the Netherlands, France, and Taiwan. P.611 The “Guideline for Prevention of Nosocomial Pneumonia” from the CDC's Healthcare Infection Control Practices Advisory Committee (HICPAC) (201) is under revision and cites a number of important issues that remain unresolved, including the role of routine culturing of water systems for Legionella species in healthcare facilities. Opposition to routine environmental cultures in the absence of documented disease is often based on the premise that Legionella colonization is ubiquitous, that Legionella can colonize water distribution systems without causing disease, and that environmental culturing is expensive (191, 202, 203). However, these assertions have been refuted based on studies in both the U.S. and the U.K. (193, 204, 205). As part of a comprehensive strategy to prevent Legionnaires' disease in transplant units, HICPAC recommends that facilities with solid organ transplant programs or hematopoietic stem cell transplant recipients perform periodic culturing for Legionella in the transplant unit's potable water supply. This recommendation also appears in the “Guidelines for Prevention of Opportunistic Infections in Bone Marrow Transplant Recipients” (206). If Legionella species are detected in the unit's water system, corrective measures (disinfection) should be performed until no Legionella is cultured. No such recommendation is made for healthcare facilities treating nontransplant patients, or for disinfection of areas serving these patients. One problem with this approach is that many cases of hospital-acquired Legionnaires' disease occur in nontransplant patients. In fact, not a single one of the patients in our original report of endemic hospital-acquired Legionnaires' disease were transplant recipients, and Legionnaires' disease constituted 22.5% (32/142) of the cases of hospital-acquired pneumonia (67). In a Swedish hospital, 31 patients with hospital-acquired Legionnaires' disease were diagnosed over a 14-month period; eight were from surgical wards, 16 from internal medicine or geriatric wards, three each from psychiatric and physiotherapy units, and one was from the maintenance department (207). Environmental Culturing Culture of hospital hot water tanks and selected showerheads and faucets (especially in transplant wards or intensive care units) is performed by collection of either swab and/or water samples from water outlets throughout the facility. Swabs of distal sites are used because the yield is higher than for water specimens (208) (Fig. 35.2). If a water sample is collected, at least 100 mL should be collected, and it should be from the hot water system. The sample should be collected immediately after turning on the faucet or shower. A minimum of ten outlets plus the hot water storage tank should be cultured for the average 250-bed hospital (198). If Legionella is isolated, specialized laboratory P.612 tests are made available in-house. The urinary antigen is especially recommended as a cost-effective test if the Legionella isolated is serogroup 1. The infection control professional then begins prospective surveillance of all nosocomial pneumonias (198, 209, 210). It has been well documented that, unless the hospital laboratory can isolate Legionella, nosocomial cases can be overlooked (193). It has also been suggested that surveillance for nosocomial legionellosis can be targeted to select high-risk patients for cost-effectiveness (68, 209); high-risk patients include transplant recipients, immunosuppressed patients, patients with underlying pulmonary disease, and intensive care unit patients. Surveillance could be expanded to all patients with nosocomial pneumonia if cases of legionellosis were uncovered in the high-risk group. It is important to point out that if the frequency of contaminated sites is low, disinfection of the water supply is not necessarily required. Legionnaires' disease is readily treatable; macrolides and quinolones can be used effectively to treat hospital-acquired pneumonias of uncertain etiology. Antibiotic prophylaxis of transplant patients with macrolides or quinolones has even been used to stem outbreaks (211). If the level of contamination increases, the option to disinfect the water supply can be exercised. Figure 35.2. Environmental cultures obtained by swabbing distal sites yield considerably more Legionella microorganisms than culturing water does. Culture of water only yields few Legionella microorganisms on the selective culture plate (left). Rotating a swab upward about the circumference of the faucet yields many Legionella microorganisms (middle). Culture of the water after the sediment within the faucet has been dislodged by the swab yields moderate numbers of Legionella microorganisms (right). Contaminated Respiratory Devices The use of sterile water for filling and rinsing humidifiers, nebulizers, and all other respiratory equipment is recommended. We have banned portable room humidifiers from our hospital. Even rinsing respiratory device tubing with tap water may create a secondary reservoir for Legionella. Subsequently, reattachment of the device to the patient could directly instill Legionella-containing respirable droplets into the respiratory tract. Devices such as medication nebulizers may retain water 12 hours after rinsing (153). DISINFECTION OF WATER DISTRIBUTION SYSTEMS Nosocomial legionellosis has been effectively controlled by disinfection of hospital water distribution systems that are colonized by Legionella. There are two basic types of disinfection systems: focal and systemic. Focal disinfection is directed at only a portion of the water distribution system, usually the incoming water or individual outlets, but not at the entire water distribution system. Systemic disinfection is directed at the entire water distribution system and the biofilm throughout the system. Focal disinfection modalities are modular and easy to install, but are notably less effective if the water distribution system is extensive or if the plumbing is heavily colonized with Legionella (212). Focal modalities include ultraviolet light, instantaneous heating systems, and ozone (213). Focal modalities are not effective if the water distribution system has preexisting Legionella colonization, because the Legionella in the water distribution system remains unaffected. Focal modalities may work best in a virgin water distribution system (e.g., in a new hospital) (214). For maximal effectiveness, a heat and flush sterilization or shock chlorination prior to activation and intermittently thereafter is advisable. Localized disinfection of faucets or showers by physical cleaning and or chlorination has a short-lived effect and is not effective (215). Systemic modalities provide a disinfectant residual that is bacteriostatic or bacteriocidal throughout the water distribution system; these modalities include hyperchlorination, copper/silver ionization, and chlorine dioxide (216, 217, 218, 219). Superheat and flush is a systemic modality that cannot be applied continuously; however, maintaining hot water temperatures at 140ВF (60ВC) minimizes recolonization (145, 213, 220, 221). In some hospitals with endemic legionellosis and a high-risk population (especially transplantation patients), multiple disinfection modalities may be needed so that, if one modality fails because of human error or mechanical failure, the other modality can serve as a safety net (75, 216). Furthermore, a focal modality (ultraviolet light) can be combined with two systemic modalities (superheat and flush, copper-silver) to ensure maximal kill of Legionella. Routine continual surveillance with environmental cultures is critical, since mechanical failures and human error are expected with any system. Cultures performed at 2-month intervals are recommended. The endpoints for disinfection should be realistic and clinically relevant. Total sterility is extremely difficult to achieve with any disinfection modality, and zero positivity is not required to prevent nosocomial Legionnaires' disease (222, 223). The efficacy of some modalities may vary depending on water use. For example, if superheated water or water containing metallic ions or chlorine cannot reach a site because the faucet is unused, disinfection cannot occur. Although the disinfection modality may remove the larger portion of the biomass of Legionella, small pockets of Legionella in protected niches may still be present but in insufficient amounts to cause infection. At our institution, Legionella infections in the hospital setting did not occur until the percentage of colonized sites exceed 30% (7). The cut point of 30% distal site positivity as an indicator of increased risk of transmission of Legionella has not been universally applicable to all hospitals. However, it does demonstrate that the concept of correlating environmental monitoring with predicting increased risk of disease is valid for Legionnaires' disease. In a study by the CDC, increased risk was associated with the extent of colonization (percentage of outlets positive) and not the concentration of Legionella recovered from a given outlet (224). The precise figure depends not only on the extent of Legionella colonization, but also on the susceptibility of patient populations to Legionella infection. For example, patients on a transplant ward may become infected with Legionella with a much smaller inoculum of Legionella in the water than would ambulatory patients on a psychiatric ward. This may be the basis for the more stringent recommendations from the CDC for monitoring and disinfection of bone marrow transplant units (206). Options for Disinfection It is important to apply a scientific method to the evaluation of disinfection methods. We have proposed that any disinfection method should be subjected to a standardized evaluation with the following steps: (a) demonstrated efficacy in vitro against P.613 Legionella microorganisms; (b) anecdotal experience of efficacy in controlling Legionella contamination in individual hospitals; (c) controlled studies of prolonged duration (years, not months) of efficacy in controlling Legionella growth and in preventing cases of hospital-acquired Legionnaires disease in individual hospitals; and (d) confirmatory reports from multiple hospitals with prolonged duration of follow-up (validation step) (222). Given the current reality of economic constraints, disinfection modalities should also be selected with the long-term goals of sustained efficacy at reasonable costs (Table 35.3). Important factors include the area requiring disinfection (one building or multiple buildings, number of floors), the number of hot water heating systems in place (one vs. multiple), the extent of colonization, and the age of the facility. Older hospitals generally pose a more formidable task in disinfection than newer hospitals because of accumulation of scale and Legionella within biofilms (106). Disinfection efforts that target the hot water system have been effective in controlling Legionella. This would suggest that treating the cold water supply may not be necessary. Given the public health implications, any commercial vendor's history of experience and service commitment in Legionella disinfection should be reviewed. It would be prudent to obtain assessments from other hospitals that have used the vendor's product. TABLE 35.3. LEGIONELLA DISINFECTION METHODS: A COST COMPARISON Method Startup Cost Annual Operating Cost Copper-silver ionization $20,000–$40,000 $2,000–$4,000 Thermal disinfection $5,000–$20,000 Repeating costs Chlorine dioxide $15,000–$20,000 $2,000–$10,000 Hyperchlorination (includes silicate injection) $50,000–$80,000 $10,000–$20,000 Ultraviolet light units $10,000–$20,000 $1,000–2,000 Estimates based on a 250 to 500-bed hospital. It should be emphasized that appearance, degree of cleanliness, and regular preventive maintenance of the system have not been shown to minimize Legionella contamination (106). Plumbing modifications including “dead-leg” removal and cleaning or replacing showerheads have been overemphasized. Nevertheless, many engineering guidelines have advocated such unvalidated approaches despite evidence that they are tedious and ineffective (106, 213, 225). The only way to be certain that a system is free of Legionella is to obtain samples for environmental cultures. Finally, a strong infection control program is critical if the approach is to be cost-effective and scientifically valid. We advise that each hospital evaluate the utility of its modality scientifically. Baseline cultures prior to disinfection over an adequate period is critical, so that the efficacy of a new disinfection modality can be adequately evaluated. Copper-Silver Ionization Ionization is the only disinfection method that has fulfilled all four evaluation criteria (222). The systems (Tarn-Pure, T.P. Technology, Buckinghamshire, UK; Liqui-Tech, Bolingbrook, IL; Enrich Products, Pittsburgh, PA) use copper/silver electrodes that generate ions when an electrical current is applied. The positively charged ions form electrostatic bonds with negatively hypercharged sites on bacterial cell walls. The distorted cellular permeability coupled with protein denaturation leads to cell lysis and death. Copper-silver ionization provides residual protection throughout the system. Theoretically, microorganisms are killed rather than suppressed, which should minimize the possibility of recolonization. Controlled studies have shown that this modality is highly effective in eradicating Legionella (217, 218, 223, 226). This system can be used in concert with ultraviolet light and chlorine (216). Two hospitals that switched from thermal eradication (superheat-and-flush) to copper-silver ionization reported that ionization was more effective for reducing the recovery of Legionella from the hospital water system (218, 223). Among the first 16 hospitals to use ionization for Legionella disinfection, 75% had attempted disinfection with other methods (222) (Fig. 35.3). All 16 hospitals were successful in preventing nosocomial Legionnaires' disease after installation of ionization systems. Although elevated pH can adversely effect the action of copper (227) and there has been speculation of ion resistance (228, 229), these hospitals reported satisfactory control of Legionella within the hospital hot water supply. The systems had been in place from 5 to 11 years. Cost depended on the number of systems installed, but the average cost was $20,000 to $40,000 (Table 35.3). Figure 35.3. Among 16 hospitals that implemented Legionella water system disinfection practices, various methods were attempted and failed prior to installation of copper-silver ionization systems. All 16 hospitals reported no cases of nosocomial Legionnaires' disease after installation of the ionization system. (From Stout JE, Yu VL. Experiences of the first 16 hospitals using copper-silver ionization for Legionella control: implications for the evaluation of other disinfection modalities. Infect Control Hosp Epidemiol 2003;24:563–568, with permission.) Chlorine Dioxide Although this technology has been used to control Legionella in European hospital water systems for many years, it has only recently been introduced into the U.S. healthcare market for this application (219, 230, 231). New technology now allows for the safe generation of chlorine dioxide on a small scale. This generation unit utilizes an electrical source and membrane technology to directly oxidize sodium chlorite (Halox, Inc., Bridgeport, CT, a unit of IDEX, Corp.) (Fig. 35.4). These generators typically provide 5 g/hour to 2.4 kg/day of chlorine dioxide. The chlorine dioxide can be fed into the water system at various points (cold water supply, hot water supply, reservoir) depending on where disinfection is desired. The required maintenance involves changing the membrane-containing cartridges. As chlorine dioxide is generated, these cartridges slowly lose their oxidizing ability and require replacement (typically after 2,000 operating hours). Preventative maintenance includes replacing various filters and tubing. Figure 35.4. Chlorine dioxide is a Legionella disinfection option that has been used extensively in Europe, but has recently been under evaluation in the United States. Chlorine dioxide is generated electrochemically from a sodium chlorite precursor within a self-contained unit. The electrochemical reaction occurs within removable cassettes. The unit generates a solution of concentrated chlorine dioxide (approx. 500 mg/L), which is injected into the water stream to achieve a 0.5 mg/L target concentration. There are minimum allowable levels of chlorine dioxide and its by-product chlorite. The U.S. Environmental Protection Agency (EPA) requirements are set forth in the National Primary Drinking Water Standards. The maximum residual disinfectant level for chlorine dioxide is 0.8 mg/L, and the maximum contaminant level for chlorite is 1.0 mg/L (232). Installations using chlorine dioxide as a supplemental disinfectant may be required to implement the same monitoring programs as primary water treatment operators. Potential users should check with their local environmental protection agency for regulatory requirements. Two controlled evaluations of chlorine dioxide have been P.614 performed in the U.S. (219, 230), and both have shown that chlorine dioxide at a concentration of <0.8 mg/L was effective in reducing Legionella species in the hospital water system. In both studies there was a significant reduction in the percentage of positive outlets; however, Legionella persisted at a low level in the treated systems and months were required to reach these levels. Difficulties were encountered in maintaining an adequate chlorine dioxide residual in the hot water system; the residual in the hot water was often <0.1 mg/L (230). This was attributed to a combination of loss of residual with increased distance from the injection point and increased decay of chlorine dioxide at higher water temperatures. Prospective studies of sufficient duration from different institutions are required to validate these results. Cost depends on the number of systems installed, but the average cost was $15,000 to $20,000 plus installation costs (Table 35.3). Superheat and Flush If Legionella must be eradicated from the water distribution system immediately, the superheat and flush method warrants primary consideration. The basic method requires that hot water tank temperatures be elevated to greater than 70ВC (158ВF) followed by flushing of all faucets and showerheads to kill L. pneumophila colonizing these sites (7, 213). All hot water tanks are shut down, drained, descaled with high-pressure steam, and then chlorinated to 100 parts per million (ppm) for 12 to 14 hours. The chlorinated water is drained and the tank flushed with water to remove the residual chlorine. The tanks are then placed back on line and the temperature is elevated to 70В to 80ВC (158В to 176ВF) for 72 hours. All distal water sites in patient care wards are flushed once a day for 2 days, whereas those sites located on patient units housing high-risk patients (intensive care units and transplant wards) are flushed once a day for 3 consecutive days. The outlets are flushed for 30 minutes. It is critical that temperatures of the flushed water be monitored to ensure that the temperature exceeds 60ВC (140ВF) distally. On the fourth day, selected distal sites are recultured; if no Legionella microorganisms are recovered, the procedure is considered completed. If Legionella is still isolated, the entire heat and flush protocol is repeated. Both maximum temperature and duration of the flush are important for successful decontamination. Hospitals that have used shorter flush times have failed to eradicate Legionella (233). Unfortunately, a minimum flush time of 5 to 10 minutes has been erroneously recommended by HICPAC (191); although the 30-minute flush is tedious, it will be more successful than the 5- to 10-minute flush. Recolonization can be delayed and minimized by maintaining hot water tank temperatures at 60ВC (140ВF). At the Pittsburgh VA Medical Center, the heat and flush method was required only once every 2 to 3 years, making this method a cost-effective one. The costs are low except for personnel time; if overtime is required, the costs can quickly escalate. We used volunteers, when possible, for the flushing process. One hospital reported overtime costs of approximately $20,000 (Table 35.3) (205). Ultimately, we abandoned this method of control in favor of the less labor-intensive copper-silver ionization system (223). The main disadvantage is that numerous personnel are involved to monitor distal sites, water tank temperatures, and flushing times. Scalding can occur, although such incidents have not been reported in numerous hospitals using this method. It should be noted that the Joint Commission on Accreditation of Healthcare Organizations has rescinded its earlier standard for a maximum water temperature of 110ВF and allows each hospital to establish its own maximum temperature. However, many P.615 states have regulations for rehabilitation and long-term-care institutions that prohibit temperature in excess of 43ВC (110ВF) at the tap (234). Hyperchlorination Hyperchlorination has proven disappointing as a long-term solution due to high expense, pipe corrosion (235, 236), introduction of carcinogenic by-products into the drinking water (237, 238, 239), and difficulty in maintaining high concentrations (2–4 ppm) of chlorine to sustain efficacy. The EPA instituted stricter standards on January 1, 2002, because of concern about chlorination by-products. Ultraviolet Light Ultraviolet light kills Legionella by disrupting cellular DNA. These systems have proven to be effective if disinfection can be localized—for example, to a transplant or an intensive care unit (75, 143, 240, 241). Because ultraviolet sterilization provides no residual protection, areas distal to the sterilizer must be disinfected following installation and startup. One effective approach is to use superheat and flush to disinfect most of the system and then to introduce chemical disinfection (metallic ion or chlorine) as an adjunct. Prefiltration is necessary to prevent the accumulation of scale on the ultraviolet lamps. One hospital reported successful control of Legionella in the water system after installation of ultraviolet units on the main water supply to a newly constructed hospital (214). GUIDELINES The control and prevention of Legionnaires' disease crosses many disciplines, and as such there are numerous guidance documents and resources for physicians, infection control professionals P.616 (ICPs), engineers, and industrial hygienists. Unfortunately, many recommendations including those that emphasize maintenance by engineers and prohibition of showering are not evidence-based, leading to adoption of ineffective methods that are tedious and expensive. Many of these documents are available via the World Wide Web (Table 35.4). The quality of Legionella-related Web sites maintained by private and state institutions, universities, professional organizations, and individuals has been reviewed (242). TABLE 35.4. INTERNET WEB SITES ARE VALUABLE RESOURCES FOR INFORMATION ON ALL ASPECTS OF LEGIONNAIRES' DISEASE REFERENCES 1. Fraser DW, Tsai T, Ornstein W, et al. Legionnaires' disease: description of an epidemic of pneumonia. N Engl J Med 1977;297:1189–1197. 2. McDade J, Shepard C, Fraser D, et al. Legionnaires' disease: isolation of a bacterium and demonstration of its role in other respiratory disease. N Engl J Med 1977;297:1197–1203. 3. Thacker SB, Bennet JV, Tsai T. An outbreak in 1965 of severe respiratory illness caused by Legionaires' disease bacterium. J Infect Dis 1978;138:512–519. 4. Tobin JO. Legionnaires' disease in a transplant unit: isolation of the causative agent from shower baths. Lancet 1980;2:118–121. 5. Fisher-Hoch SP, Tobin JO, Belson AM. Investigation and control of an outbreak of Legionnaires' disease in a district hospital. Lancet 1981;1:932–936. 6. Stout JE, Yu VL, Vickers RM, et al. Ubiquitousness of Legionella pneumophila in the water supply of a hospital with endemic Legionnaires' disease. N Engl J Med 1982;36:466–468. 7. Best M, Yu VL, Stout J, et al. Legionellaceae in the hospital water supply—epidemiological link with disease and evaluation of a method of control of nosocomial Legionnaires' disease and Pittsburgh pneumonia. Lancet 1983;2:307–310. 8. Plouffe JR, Para MF, Maher WE, et al. Subtypes of Legionella pneumophila serogroup 1 associated with different attack rates. Lancet 1983;2:649–650. 9. Nolte FS, Conlin C, Roisin A. Plasmids as epidemiological markers in nosocomial Legionnaires' disease. J Infect Dis 1984;149:251–256. 10. Benin AL, Benson RF, Beser RE. Trends in Legionnaires' disease, 1980–1998: declining mortality and new patterns of diagnosis. Clin Infect Dis 2002;35:1039–1046. 11. Hookey JV, Saunders NA, Fry NK, et al. Phylogeny of Legionellaceae based on small-subunit ribosomal DNA sequences and proposal of Legionella lytica comb. nov. for Legionella-like amoebal pathogens. Int J Syst Bacteriol 1996;46:526–531. 12. Fry NK, Warwick S, Saunders NA, et al. The use of 16S ribosomal RNA analysis to investigate the phylogeny of the family Legionellaceae. J Gen Microbiol 1991;137:1215–1222. 13. Brenner DJ, Steigerwalt AG, McDade JE. Classification of the Legionnaires' disease bacterium: Legionella pneumophila, genus novum, species nova, of the family Legionellaceae. Ann Intern Med 1979;90:656–658. 14. Fields BS, Benson RF, Besser RE. Legionella and Legionnaires' disease: 25 years of investigation. Clin Microbiol Rev 2003;15:506–526. 15. Adeleke AA, Fields BS, Benson RF, et al. Legionella drozanskii sp. nov., Legionella rowbothamii sp. nov. and Legionella fallonii sp. nov.: three unusual new Legionella species. Int J Syst Evol Microbiol 2001;51(3):1151–1160. 16. Benson RF, Fields BS. Classification of the genus Legionella. Semin Respir Infect 1998;13:90–99. 17. Schousboe M, Gibbons S, Chereshsky A. Community-acquired pneumonia due to Legionella feeleii serogroup 2. N Z Med J 1995;108:279. 18. Harris A, Lally M, Albrecht M. Legionella bozemanii pneumonia in three patients with AIDS. Clin Infect Dis 1998;27:97–99. 19. McNally C, Hackman B, Fields BS, et al. Potential importance of Legionella species as etiologies in community acquired pneumonia (CAP). Diagn Microbiol Infect Dis 2000;38:79–82. 20. Muder RR, Yu VL. Infection due to Legionella species other than L. pneumophlla. Clin Infect Dis 2002;35:990–998. P.617 21. Fang GD, Yu VL, Vickers RM. Disease due to Legionellaceae (other than Legionella pneumophila): historical, microbiological, clinical and epidemiological review. Medicine 1989;68:116–139. 22. Knirsch CA, Jakob K, Schoonmaker D, et al. An outbreak of Legionella micdadei pneumonia in transplant patients: education, molecular epidemiology, and control. Am J Med 2000;108:290–295. 23. Waite R. Confirmation of identity of legionellae by whole cell fatty-acid and isoprenoid quinone profiles. In: Harrison TG, Taylor AG, eds. A laboratory manual for Legionella. Chichester, UK: John Wiley, 1988:69–101. 24. Sonesson A, Jantzen E, Bryn K, et al. Composition of 2, 3-dihdroxy fatty acid-containing lipopolysaccharides from Legionella israelensis, Legionella maceachernii, and Legionella micdadei. Microbiol 1994;140:1261–1271. 25. Lema M, Brown A. Electrophoretic characterization of soluble protein extracts of Legionella pneumophila and other members of the family Legionellaceae. J Clin Microbiol 1983;17:1132–1140. 26. Brindle RJ, Bryant TN, Draper PW. Taxonomic investigation of Legionella pneumophila using monoclonal antibodies. J Clin Microbiol 1989;27:536–539. 27. Diogo A, Verissimo A, Nobre MF, et al. Usefulness of fatty acid composition for differentiation of Legionella species. J Clin Microbiol 1999;37:2248–2254. 28. Fox A, Lau PY, Brown A, et al. Capillary gas chromatographic analysis of carbohydrates of Legionella pneumophila and other members of the family Legionellaceae. J Clin Microbiol 1984;19:326–332. 29. Lo Presti F, Riffard S, Vandenesch F, et al. Identification of Legionella species by random amplified polymorphic DNA profiles. J Clin Microbiol 1998;36:3193–3197. 30. Bansal NS, McDonell F. Identification and DNA fingerprinting of Legionella strains by randomly amplified polymorphic DNA analysis. J Clin Microbiol 1997;35(9):2310–2314. 31. Pinar A, Ahkee S, Miller RD, et al. Use of heteroduplex analysis to classify legionellae on the basis of 5S rRNA gene sequences. J Clin Microbiol 1997;35(6):1609–1611. 32. Gheldre YD, Maes N, Lo Presti F, et al. Rapid identification of clinically relevant Legionella spp. by analysis of transfer DNA intergenic spacer length polymorphism. J Clin Microbiol 2001;39(1):162–169. 33. Lowry PW, Tompkins LS. Nosocomial legionellosis: a review of pulmonary and extrapulmonary syndromes. Am J Infect Cont 1993;21:21–27. 34. Blatt SP, Parkinson MD, Pace E, et al. Nosocomial Legionnaires' disease: aspiration as a primary mode of transmission. Am J Med 1993;95:16–22. 35. Yu VL. Could aspiration be the major mode of transmission for Legionella? Am J Med 1993;95:13–15. 36. Stone BJ, Kwaik YA. Expression of multiple pili by Legionella pneumophila: identification and characterization of a type IV pilin gene and its role in adherence to mammalian and protozoan cells. Infect Immun 1998;66:1768–1775. 37. Stout JE, Best M, Yu VL, et al. A note on symbiosis of Legionella pneumophila and Tatlockia micdadei with human respiratory flora. J Appl Bacteriol 1986;60:297–299. 38. Swanson MS, Hammer BK. Legionella pneumophila pathogenesis: a fateful journey from amoebae to macrophages. Annu Rev Microbiol 2000;54:567–613. 39. Fields BS. The molecular ecology of legionellae. Trends Microbiol 1996;4:286–290. 40. Cirillo JD, Falkow S, Tompkins LS. Legionella pneumophila in Acanthamoeba castellani enhances invasion. Infect Immun 1994;62:3254–3261. 41. Shuman HA, Purcell M, Segal G, et al. Intracellular multiplication of Legionella pneumophila: human pathogen or accidental tourist? Curr Top Microbiol Immunol 1998;225:99–112. 42. Horwitz MA. Toward an understanding of host and bacterial molecules mediating L. pneumophila pathogenesis. In: Barbaree JM, Breiman RF, Dufour AP, eds. Legionella: current status and emerging perspectives. Washington, DC: American Society of Microbiology, 1993:55–62. 43. Swanson MS, Fernandez-Moreia E. A microbial strategy to multiply in macrophages: the pregnant pause. Traffic 2002;3:170–177. 44. Hammer BK, Tateda ES, Swanson MS. A two-component regulator induces the transmission phenotype of stationary-phase Legionella pneumophila. Mol Microbiol 2002;44:107–118. 45. Byrd TF, Horwitz MA. Lactoferrin inhibits or promotes Legionella pneumophila intracellular multiplication in nonactivated and interferon gamma activated human monocytes depending upon its degree of iron saturation. Iron-lactoferrin and nonphysiologic iron chelates reverse monocyte activation against Legionella pneumophila. J Clin Invest 1991;88:1103–1112. 46. Yamamoto Y, Klein TW, Tomioka M, et al. Differential effects of granulocyte/macrophage colony-stimulating factor (GM CSF) in enhancing macrophage resistance to Legionella pneumophila vs. Candida albicans. Cell Immunol 1997;176:75–81. 47. Park DR, Skerrett SJ. IL-10 enhances the growth of Legionella pneumophila in human mononuclear phagocytes and reverses to protective effect of IFN-Оі. J Immonol 1996;157:2528–2538. 48. Tateda K, Matsumoto T, Ishii Y, et al. Direct evidence of interleukin-12 as a critical mediator in patients with Legionella pneumonia (G-38). 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, Canada, 1997. 49. Lochner TC, Bigley R, Iglewski BH. Defective triggering of polymorphonuclear leukocyte oxidative metabolism by L. pneumophila toxin. J Infect Dis 1985;151:42–46. 50. Cianciotto NP, Eisenstein BI, Mody CH, et al. A mutation in the mip gene results in an attenuation of Legionella pneumophila virulence. J Infect Dis 1990;162:121–126. 51. Williams A, Rechnitzer C, Lever MS et al. Intracellular production of Legionella pneumophila destructive protease in alveolar macrophage. In: Barbaree JM, Breiman RF, Dufour AP, eds. Legionella—current status and emerging perspectives. Washington, DC: American Society for Microbiology, 1993:88–90. 52. Conlan JW, Williams A, Ashworth L. In vivo production of a tissue-destructive protease by Legionella pneumophila in the lungs of experimentally infected guinea pigs. J Gen Microbiol 1988;134:143–149. 53. Black WJ, Quinn FD, Tompkins LS. Legionella pneumophila zinc metalloprotease is structurally and functionally homologous to Pseudomonas aeruginosa elastase. J Bacteriol 1990;172:2608–2613. 54. Muller HE. Proteolytic action of Legionella pneumophila on human serum proteins. Infect Immun 1980;27:51–53. 55. Nolte FS, Hollick GE, Robertson RG. Enzymatic activities of Legionella pneumophila and Legionella-like organisms. J Clin Microbiol 1982;15:175–177. 56. Pine L, Hoffman PS, Malcolm GB, et al. Determination of catalase peroxidase, and superoxide dismutase within the genus of Legionella. J Clin Microbiol 1984;20:421–429. 57. Dournon E, Bibb WF, Rajagopalan P, et al. Monoclonal antibody reactivity as a virulence marker for Legionella pneumophila serogroup 1 strains. J Infect Dis 1988;157:496–501. 58. Stout JE, Joly J, Para M, et al. Comparison of molecular methods for subtyping patients and epidemiologically-linked environmental isolates of L. pneumophila. J Infect Dis 1988;157:486–494. 59. Helbig JH, Luck PC, Knirel YA, et al. Molecular characterization of a virulence-associated epitope on lipopolysaccharide of Legionella pneumophila serogroup 1. Epidemiol Infect 1995;115:71–78. 60. Alli OA, Zink S, von Lackum NK, et al. Comparative assessment of virulence traits in Legionella spp. Microbiol 2003;149:631–641. 61. Marston BJ, Lipman HB, Breiman RF. Surveillance for Legionnaires' disease. Risk factors for morbidity and mortality. Arch Intern Med 1994;154:2417–2422. 62. Hoge CW, Breiman RF. Advances in the epidemiology and control of Legionella infections. Epidemiol Rev 1991;13:329–340. 63. Lepine LA, Jernigan DB, Butler JC, et al. A recurrent outbreak of nosocomial Legionnaire's disease detected by urinary antigen testing: evidence for long-term colonization of a hospital plumbing system. Infect Control Hosp Epidemiol 1998;19:905–910. 64. Kool JL, Fiore AE, Kioski CM, et al. More than ten years of unrecognized nosocomial transmission of Legionnaires' disease among transplant patients. Infect Control Hosp Epidemiol 1998;19:898–904. P.618 65. Johnson JT, Yu VL, Best M, et al. Nosocomial legionellosis uncovered in surgical patients with head and neck cancer: implications for epidemiologic reservoir and mode of transmission. Lancet 1985;2:298–300. 66. Muder RR, Yu VL, McClure J, et al. Nosocomial Legionnaires' disease uncovered in a prospective pneumonia study: Implications for underdiagnosis. JAMA 1983;249:3184–3188. 67. Yu VL, Kroboth FJ, Shonnard J, et al. Legionnaires' disease: new clinical perspective from a prospective pneumonia study. Am J Med 1982;73:357–361. 68. Marrie TJ, Macdonald S, Clarke K, et al. Nosocomial Legionnaires' disease: lessons from a four year prospective study. Am J Infect Control 1991;19:79–85. 69. Chow J, Yu VL. Legionella: a major opportunistic pathogen in transplant recipients. Semin Respir Infect 1998;13:132–139. 70. Bock B, Edelstein P, Snyder K, et al. Legionnaires' disease in renal transplant recipients. Lancet 1978;1:410–413. 71. Prodinger WM, Bonatti H, Allerberger F, et al. Legionella pneumonia in transplant recipients: a cluster of cases of eight years duration. J Hosp Infect 1994;26:191–202. 72. Fuller J, Levinson MD, Kline JR, et al. Legionnaires' disease after heart transplantation. Ann Thoracic Surg 1985;39:308–311. 73. Singh N, Muder RR, Yu VL, et al. Legionella infection in liver transplant recipients: implications for management. Transplantation 1993;56:1549–1551. 74. Kugler JW, Armitage JO, Helms CM, et al. Nosocomial Legionnaires' disease: occurrence in recipients of bone marrow transplants. Am J Med 1983;74:218–228. 75. Matulonis U, Rosenfeld CS, Shadduck RK. Prevention of Legionella infections in a bone marrow transplant unit: multifaceted approach to decontamination of a water system. Infect Control Hosp Epidemiol 1993;14:571–575. 76. Squier C, Krystofiak S, McMahon J, et al. Decrease in nosocomial Legionnaires' disease in Allegheny County: guidelines for prevention work (abstract). Association of Practitioners of Infection Control, S