Hospital-acquired Legionella pneumonia has a fatality rate of 28%, and the source is the water distribution system. Two prevention strategies have been advocated. One approach to prevention is clinical surveillance for disease without routine environmental monitoring. Another approach recommends environmental monitoring even in the absence of known cases of Legionella pneumonia. We determined the Legionella colonization status of water systems in hospitals to establish whether the results of environmental surveillance correlated with discovery of disease. None of these hospitals had previously experienced endemic hospital-acquired Legionella pneumonia.

Cohort study.

Twenty US hospitals in 13 states.

Hospitals performed clinical and environmental surveillance for Legionella from 2000 through 2002. All specimens were shipped to the Special Pathogens Laboratory at the Veterans Affairs Pittsburgh Medical Center.

Legionella pneumophila and Legionella anisa were isolated from 14 (70%) of 20 hospital water systems. Of 676 environmental samples, 198 (29%) were positive for Legionella species. High-level colonization of the water system (30% or more of the distal outlets were positive for L. pneumophila) was demonstrated for 6 (43%) of the 14 hospitals with positive findings. L. pneumophila serogroup 1 was detected in 5 of these 6 hospitals, whereas 1 hospital was colonized with L. pneumophila serogroup 5. A total of 633 patients were evaluated for Legionella pneumonia from 12 (60%) of the 20 hospitals: 377 by urinary antigen testing and 577 by sputum culture. Hospital-acquired Legionella pneumonia was identified in 4 hospitals, all of which were hospitals with L. pneumophila serogroup 1 found in 30% or more of the distal outlets. No cases of disease due to other serogroups or species (L. anisa) were identified.

Environmental monitoring followed by clinical surveillance was successful in uncovering previously unrecognized cases of hospital-acquired Legionella pneumonia.