To sample or not: The Legionella debate

It's nothing without prevention efforts

By Matt Freije, BS

Consultant and author on Legionella issues in health care

In light of the ongoing debate on the issue of sampling hospital water systems for Legionella, infection control professionals may want to consider sampling, based on a thorough review of the facts. (See Hospital Infection Control, June 1996, pp. 73-77.) This approach falls somewhere between "all hospitals should sample routinely" and "never sample in the absence of disease." Consider the following in deciding what's best for your institution:

* A proactive vs. a reactive approach.

The sampling issue is actually a subset of a bigger issue: Should hospitals take a proactive or reactive approach to Legionnaires' disease? A proactive hospital maintains plumbing systems and cooling towers to minimize Legionella growth, and thus may collect samples periodically to see if the preventive measures are working. A reactive hospital does nothing until a case of Legionnaires' disease is confirmed. A huge flaw with the reactive approach is that risk reduction measures are rarely implemented because most nosocomial cases go undetected-- which may be why the incidence of Legionnaires' disease has stayed about the same for the last 20 years.

* What sampling can and cannot tell you.

Centers for Disease Control and Prevention guidelines for prevention of nosocomial pneumonia state that "the relationship between the results of water cultures and the risk of legionellosis remains undefined. ... Data are insufficient to assign a level of risk of disease even on the basis of the number of colony-forming units detected in samples."1 Indeed, sampling results will not give a yes/no answer, but may provide information that will indicate the condition of your water. If Legionella are found at several sites, you will know that your preventive measures are not working, particularly if the counts are high. In other cases, however, the information may not give clear feedback.

Several screenings are usually required to make good decisions. Test results are more accurately interpreted by viewing trends rather than snapshots, just as stock investors generally do better by analyzing 10-week averages rather than selecting a stock based only on its change from one week to the next.

* Sampling cannot replace preventive measures or appropriate patient surveillance.

Legionella can be in sites not sampled, and counts vary from time to time. Therefore, it is presumptuous to assume that a hospital is free of Legionella, and thus risk of disease, even when all samples test negative. The safest way to view results is to listen only to bad news: Consider corrective measures if samples are positive, but don't relax preventive efforts or patient surveillance if the results are low.

Even if environmental samples are consistently negative, the engineering department should take reasonable measures to minimize

Legionella growth in mechanical systems and physicians should maintain suspicion for the disease in high risk patients. Although a hospital collects samples routinely to look for problems, it should not view negative test results as meaning there is no risk of disease. Remember that environmental sampling does nothing to reduce the risk of Legionnaires' disease--only preventive measures do.

* Risk management.

Ronnie Penton, JD, an attorney in Bogalusa, LA, with extensive experience in litigating Legionnaires' disease cases, has stated, "The reasonableness of an institution's actions will be measured by what is known at the time of the outbreak with respect to preventing, detecting and eliminating risks, and identifying cases of disease." If you disinfect systems based on environmental sampling results, and by doing so prevent one or more cases of Legionnaires' disease, you not only save lives, but also save money by taking corrective action under nonoutbreak conditions.

What if all environmental samples test negative but a case of nosocomial Legionnaires' disease still occurs? You failed to prevent Legionnaires' disease, which was your first priority, but you would not have prevented the illness any more effectively by not sampling. And the money spent on sampling will not have been wasted if your file of clean laboratory reports, establishing that you acted reasonably, is useful in defending a lawsuit or negotiating a lower settlement.

* Cost.

A hypothetical 300-bed hospital sampling one cooling tower, two hot-water tanks, and 10 outlets (faucets or showerheads) will spend from $700 to $2,000 per screening if an outside laboratory is used.

If samples are cultured by your in-house laboratory, and if you can schedule your screenings to coincide with your laboratory's slow periods (so that neither overtime pay nor additional laboratory technicians are required), your only expense will be the laboratory materials required. Bear in mind, however, that culturing environmental samples for Legionella is highly specialized. Significant training time may be required before even the most qualified laboratory technicians are effective.

* Sampling may be an impetus for an overall risk reduction effort.

If a hospital is collecting samples routinely, the engineering department will make an effort to keep water systems Legionella-safe, just as a daily weigh-in will make a boxer more cognizant of his eating habits. In hospitals that are already providing the laboratory tests, clinicians will be more likely to maintain a suspicion for the disease if environmental samples are collected routinely, especially if Legionella is being found.

In general, sampling may increase communication between facility management, infection control, and the medical staff. Ideally, the facility manager will be alerted whenever legionellosis is detected in patients so that he or she will know to investigate mechanical systems, and medical staff members will be alerted whenever water samples are positive so that they can be especially watchful for cases of legionellosis.

[Editor's note: Freije has provided indoor environmental consulting services since 1989 and is now an author of guide materials, primarily for health care facilities. Portions of this column were adapted from his new book Legionella Control in Health Care Facilities: A Guide for Minimizing Risk (July 1996). Information on the book can be obtained by fax (fax-on-demand) by calling (800) 332-9430 and selecting document 1070, or on the Internet at www.


1. Centers for Disease Control and Prevention. Hospital Infection Control Practices Advisory Committee. Guideline for prevention of nosocomial pneumonia. Infect Control Hosp Epidemiol 1994; 15:587-627. *