Nosocomial Legionnaires' disease. Epidemiologic demonstration of cooling towers as a source

Investigation of a recent outbreak of nosocomial legionnaires' disease--initially thought to be due to the documented presence of Legionella pneumophila in the hospital potable water--showed that aerosols from one or more cooling towers were the actual source of infection. From June 27 to Aug 25, 1983, nosocomial legionnaires' disease developed in 15 persons at a hospital in Rhode Island. Twelve (80%) of 15 case-patients occupied rooms in building 1, unit B, compared with eight (28%) of 29 control patients (odds ratio = 10.8; 95% confidence interval = 1.4 to 85.6). Subsequent investigation demonstrated that water in a cooling tower located 100 ft upwind of unit B was heavily contaminated with L pneumophila, serogroup 1, subgroup 1, 2, 4, 5. The same strain was isolated from nine of the patients and from the make-up water for the tower. Active surveillance during the ten months following decontamination of the cooling tower identified no additional cases of nosocomial legionnaires' disease, although the hospital potable water had not been treated. While recommendations have been made for controlling nosocomial legionnaires' disease by heating or hyperchlorination of hospital potable water, this outbreak demonstrates the importance of an adequate epidemiologic-environmental investigation in choosing the appropriate control strategy.