Search

From March 10 through October 29, 1975, Listeria monocytogenes infection occurred in seven neonates born at a 401-bed general hospital in Greenville, S.C. In the preceding 19 months, there had been only one case. Six of the seven isolates from the infected infants were of serotype JB. Risk of listeriosis in neonates was associated with being born to mothers of low socioeconomic status and being born to mothers who had had vaginitis during pregnancy. Nosocomial transmission of L. monocytogenes appeared unlikely, but the source of the infection is unknown. Of the media used in this study for isolation of L. monocytogenes from mixed cultures, McBride's agar and McBride's agar with nalidixic acid were most useful.

Bacteria recently recognized as nosocomial pathogens generally fall into three categories: those that grow slowly, those that are fastidious in their nutritional or atmospheric requirements and those that resemble commensals. Each characteristic has contributed to the delay in perceiving their importance. Mycobacterium chelonei and Myco. fortuitum--which grow slowly, although characterized as "rapid-growing" mycobacteria--cause sternal osteomyelitis, pericarditis and endocarditis after cardiac surgery as well as other wound infections after many types of surgery. Myco. chelonei-like organisms have been found to cause "sterile" peritonitis in patients receiving long-term peritoneal dialysis. Legionella pneumophila and L. micdadei are fastidious bacteria that were more difficult to detect because they stain poorly with the Gram method. They cause pneumonia and lung abscess, especially in immunocompromised people. Clostridium difficile is an anaerobe that causes toxin-mediated pseudomembranous colitis in persons given antibiotics that inhibit competing gut bacteria. Chylamydia trachomatis, an intracellular organism that has not been grown in vitro, causes pneumonia and conjunctivitis in young infants who acquire the organism from their mothers at birth. Group JK bacteria cause septicemia in patients whose immune responses have been suppressed and must be distinguished from "diphtheroid" contaminants in blood cultures. Clinicians, microbiologists and epidemiologists must be alert to the characteristics of these organisms that make them easily overlooked and should also anticipate the existence of other bacteria not yet identified.

As of April 30, 1980, 83 nosocomial cases of sporadic legionellosis had been reported to the Centers for Disease Control (CDC). In all 83 cases the patients had pneumonia; the median age of the patients was 56.5 years. All but one patient were hospitalized at the time of onset. Of 71 patients for whom the outcome is known, 22 (31 percent) died of causes directly attributed to their infection. Eleven patients had end-stage renal disease, 28 were receiving systemic immunosuppressive medications, 17 had cancer, 12 had chronic bronchitis or emphysema, 29 were smokers, and four had diabetes mellitus. Risks of acquiring nosocomial sporadic legionellosis for patients with these conditions relative to the general United States population = 340, 26, 11, 3.7, 1.9 and 1.3, respectively. These risk factors are similar to those identified for sporadic community-acquired legionellosis and for epidemic nosocomial legionellosis. Methods for preventing nosocomial legionellosis are not known, but comparing Legionella to other water-associated organisms which have been spread from medical devices to cause pneumonia may be fruitful.

To describe the epidemiology of bacteremia in a large, well defined population, the authors reviewed medical records for residents of Charleston County, South Carolina, who had bacteria isolated from blood in the period 1974 to 1976. The incidence was 80 cases per 100,000 population per year. The most common organisms were Escherichia coli, Staphylococcus aureus, Klebsiella, and Streptococcus pneumoniae. The incidence was highest for neonates, infants, and those 70 years of age and older with annualized attack rates of 1,864,250, and 446 cases per 100,000 population, respectively. The incidence was 3.2 times higher for blacks than for whites and, within races, appeared to be independent of family income. Twenty-five per cent of patients had no clinically apparent focus of infection, 26% had urinary tract infection, and 17% had pneumonia. Thirty-nine per cent of cases were nosocomial, and 30% of patients died.

Three patients with severe pneumonia at a community hospital in Columbus, Ohio, were found to have Legionnaires' disease in late August 1977. A subsequent serologic survey of patients with pneumonia at this hospital identified three additional cases. Among patients with pneumonia, hospital exposure in the 2 weeks before onset of illness was significantly associated with Legionnaires' disease (P = 0.003). Serosurveys of hospital employees with a recent history of upper respiratory illness, healthy employees, and workers at the hospital construction site showed that one of 101, one of 107, and none of 114, respectively, has a single reciprocal titer of greater than or equal to 256 to the Legionnaires' disease (LD) bacterium. Serosurveys of patients with pneumonia at three control hospitals identified five additional patients with Legionnaires' disease, three of them with pneumonia that was apparently hospital acquired in a single renal transplant unit. A fourth patient from that unit without clinical illness had a fourfold rise in titer to LD bacterium.

During September and October 1979, 23 patients admitted to hospitals in the Boston area had systemic Listeria monocytogenes infection. Twenty (87%) of these isolates were L monocytogenes type 4b, whereas only nine (33%) of the isolates serotyped during the preceding 26 months had been 4b. Patients with type 4b Listeria infection during the epidemic period (case patients) differed from patients with sporadic Listeria infection in the preceding two years in that more of the case patients had hospital-acquired infection (15/20 vs 4/18), had received antacids or cimetidine before the onset of listeriosis (12/20 vs 3/18), and had gastrointestinal tract symptoms that began at the same time as fever (17/20 vs 4/18). In addition, more case patients took antacids or cimetidine compared with patients matched for age, sex, and date of hospitalization (12/20 vs 10/40). Three foods were preferred by case patients more frequently than by control patients: tuna fish, chicken salad, and cheese. However, the only common feature appeared to be the serving of these foods with raw celery, tomatoes, and lettuce. The raw vegetables may have been contaminated with Listeria, which was able to survive ingestion because of gastric acid neutralization and subsequently to cause enteritis, bacteremia, and meningitis in susceptible hosts. However, we cannot exclude pasteurized milk as a source of this outbreak.

Legionella pneumophila serogroup 6 was isolated from nine of 16 shower heads in a Chicago hospital ward where three patients had contracted Legionnaires' disease caused by serogroup 6 L. pneumophila. Each patient had showered there 2 to 10 days before the onset of disease symptoms. We also isolated the bacteria in two other hospitals, and found the same serogroups as had been causing Legionnaires' disease in those hospitals: serogroup 1 in Pittsburgh and serogroups 1 and 4 in Los Angeles. However, showers from hospital wards where no patients had contracted Legionnaires' disease also yielded L. pneumophila. Shower heads at the Chicago hospital were sterilized with ethylene oxide but rapidly became recontaminated, suggesting that the potable water at these hospitals may have contained the organism. The question of whether aerosols of shower water or other exposures to potable water containing L. pneumophila may cause nosocomial Legionnaires' disease has not been resolved but deserves further study.

A three-year epidemic of legionnaires' disease in a hospital was dramatically curtailed following hyperchlorination of the potable water supply. The hypothesis that potable water was the source for the outbreak was further supported by isolation of Legionella pneumophila (the agent of legionnaires' disease) from the hospital water supply, observation that a sudden upsurge had occurred in the number of cases following a peculiar manipulation of the hospital water system, and documentation of a 30-fold increase in concentration of organisms in the water when this manipulation was artificially recreated. Thus, potable water may be an important source of epidemic legionnaires' disease and continuous hyperchlorination a method of control.