Hospital water: A hazard with an unclear solution

Abstract & Commentary

Threat to immunocompromised patients

Synopsis: Both clinical disease and sputum colonization due to Mycobacterium avium complex were traced to the potable water system of a large public hospital.

Source: Tobin-D’Angelo MJ, et al. Hospital water as a source of Mycobacterium avium complex isolates in respiratory secretions. J Infect Dis 2004; 189:98-104.

Investigators at Grady Memorial Hospital in Atlanta examined 161 Mycobacterium avium complex (MAC) isolates from respiratory specimens obtained from 131 patients and 13 MAC isolates from the hospital’s hot water system.

Of the 131 patients, 35% had MAC disease, based upon meeting the American Thoracic Society (ATS) criteria for MAC pulmonary infection or upon isolation of identical MAC strains from blood and sputum. Ninety-six patients (73%) had MAC colonization of sputum in the absence of disease. Ten different MAC clusters of closely related strains were identified by pulsed-field gel electrophoresis.

Of these clusters, eight contained one or more isolates with MAC disease; the majority of patients were HIV-infected. Eighty-eight percent of isolates from patients with MAC colonization belonged to one of the 10 clusters, as did 53% of isolates from patients with MAC disease. The largest cluster contained isolates from 85 patients, 13 (15%) of which were from patients with MAC disease.

The cluster also contained isolates recovered from the hospital hot water system. Tobin-D’Angelo and associates concluded that the hospital hot water system was the source of both sputum colonization with MAC and MAC infection.

Comment by Robert Muder, MD, hospital epidemiologist, Pittsburgh VA Medical Center.

Hospital water systems can be the source of many different nosocomial bacterial pathogens, including Legionella species, nonfermentative Gram-negative bacilli such as Pseudomonas aeruginosa and Stenotrophomonas maltophilia, and nontuberculous mycobacteria (NTM).1 NTM can be isolated from up to 83% of municipal water supplies.2 NTM can multiply in a variety of aquatic environments.

They are resistant to many disinfectants, including chlorine in the concentrations used to treat potable water. Several species, including MAC, can grow at temperatures of 45° and above.3

Colonization of hospital water with NTM has several potential adverse consequences. Pseudo-outbreaks of infection can occur when medical devices, such as bronchoscopes and endoscopes, are cleaned with contaminated water.

Even when no infection results, pseudo-outbreaks can impose a considerable burden on the microbiology laboratory and on the infection control program. The laboratory may be required to process many specimens for NTM; the infection control program will be required to investigate numerous cases of pseudo-infection and determine the source of the outbreak. More ominously, contaminated water can result in nosocomial infection.

Outbreaks of postoperative wound infection, typically involving M. abscessus, have affected cardiac surgery, plastic surgery, and ophthalmologic surgery patients, among others. The resulting infections are particularly problematic, as NTM tend to be resistant to many standard antituberculous medications. Surgical debridement and prolonged multidrug antimicrobial therapy often are needed.

Exposure to NTM is hazardous particularly for HIV-infected patients, as MAC is a major cause of morbidity among those with advanced immunodeficiency. MAC isolates from HIV-infected patients tend to be genetically distinct, reflecting the widespread occurrence of MAC in the environment.4

The occurrence of multiple cases of MAC infection with closely related strains among HIV-infected patients in a single institution is unusual.

In the Grady Memorial report, the majority of isolates (53%) from patients with MAC disease belonged to a cluster of genetically related isolates, and 28% of all diseases among HIV-infected patients were caused by a strain isolated from the hospital water system. Although it seems clear that colonization of hospital water systems by NTM poses a significant hazard to susceptible patients, the solution to the problem is unclear.

There are no established methods for eradicating these organisms from potable water. As noted, chlorine in concentrations usually added to potable water is ineffective. Copper-silver ion treatment has proven effective in eliminating Legionella species from hospital water systems. NTM are susceptible to killing by copper and silver ions in the laboratory,5 but actual clinical experience is lacking. Institutions whose water systems are colonized by NTM should consider providing sterile water for highly immunocompromised patients until a proven water disinfection system is available.


1. Squier C, et al. Waterborne nosocomial infections. Curr Infect Dis Rep 2000; 2:490-496.

2. Carson LA, et al. Prevalence of nontuberculous myco-bacteria in water supplies of hemodialysis centers. Appl Environ Microbiol 1988; 54:3,122-3,125.

3. Wallace RJ, et al. Nosocomial outbreaks/pseudo outbreaks caused by nontuberculous mycobacteria. Annu Rev Microbiol 1998; 52:453-490.

4. Arbeit RD, et al. Genetic diversity among strains of Mycobacterium avium causing monoclonal and polyclonal bacteremia in patients with AIDS. J Infect Dis 1993; 167:1,384-1,390.

5. Lin YE, et al. Inactivation by Mycobacterium avium by copper and silver ions. Water Research 1998; 32:1,997-2,000.