A recent article1, ASHRAE guidelines, and the position of many health authorities point to the reality of the dangers inherent in systems populated with Legionella species. However, therein and in other publications, we find the mantra “The mere presence of Legionella . . . will not in itself cause illness. High numbers of Legionella have been noted in cooling towers and other sources with no associated disease” that serves to lessen the impact of the reality. This untenable position is also held by some water treatment firms, some at the Centers for Disease Control (CDC), and a few world health authorities (albeit not all). If numbers imply nothing, why test?

To everyone with a similar view, I pose the question: “If someone were to add live vibrio cholera to a 1,000-gal container at an inoculum <1 colony forming unit per million (cfu/ml), would that person take a drink of water from that tank?”

An emphatic “no” indicates a cautious nature and intelligent position. Why tempt the fates? Likewise when dealing with Legionella. Why run risks at the expense of others? How can one defend the position dictated by the mantra in a court of law when faced with litigants seeking redress for the pain and suffering associated with losing a loved one? How can one face the mirror knowing that failure to address this potential threat caused the loss of a fellow worker or loved one?

The “ . . . mere presence . . .” mantra has been defended by the asbestos “one fiber” argument, wherein the absurdity of inhaling one fiber causing illness is taken to the Nth degree. Should one’s goal be total elimination? This is the basis for the position — why fight floodwaters? One must agree that a fiber of asbestos, a nonliving inorganic compound, is certainly different from a bacterium that lives, multiplies, and mutates within its environment. If the asbestos fiber had the ability to multiply within the lungs then the analogy would hold, but it does not. Overlooked in this argument is that if there is one fiber, there are others.

The reality, condensed for brevity, is that people are falling ill and dying, with full recovery being defined simply as being alive.

The Problem’s “Unseen” Scope

Also overlooked is the fact that Legionella has been implicated in attacks via wound infections. What hvac maintenance mechanic does not have cuts on the hands? What about a janitor washing a mop in a slop sink, or an employee or visitor washing a staple or paper cut under a sink? In Australia, the major killer is L. longbeachae, some of which is found in compost.

Another point is the assumption that illness was not associated with the count. However, it has been asserted that many atypical pneumonia deaths are undiagnosed Legionella fatalities (10,000-100,000 annually in the United States). How many of the at-risk population succumbed to Legionella, yet the death certificate lists atypical pneumonia or congestive heart failure? How many of the at-risk population undergo autopsies to ascertain exact cause of death? An ill elderly parent dies in a nursing home or hospital; will an autopsy be performed? A child ill with chicken pox succumbs to “atypical pneumonia.” Will loved ones demand an autopsy to determine the exact cause of death? Why?

The New York State Department of Health2 suggests that health care facilities conduct surveillance to identify and develop policies to ensure that Legionella is considered in patients that come down with nosocomial (hospital-acquired) pneumonia. However, in cash-strapped health care facilities, these suggestions do not go far. What about pneumonia patients transferred from nursing homes? Let us not overlook the reality that while many victims appear to be healthy, they could be part of the at-risk population and not know it. The 38-year-old nonsmoking, female janitor at the Federal Building in California many years ago that succumbed to Legionnaires’ Disease was certainly not part of the at-risk population, or was she? That outbreak took two lives. The mantra is ill-serving. It provides a cushion for inaction, a hook to hang one’s hat on, a balm for the soul when the grim reaper calls.

Lastly, at what time is action appropriate and what is the appropriate action? Do we wait for an outbreak and then use serotyping techniques to focus on the outbreak source? Is a count of <1 cfu/ml of concern? We need definitions and understanding. Let us direct our attention to the task of testing a system for Legionella species (sp.) and try to understand the scope.

Table 1. Some examples of recent worldwide Legionella deaths.

Testing Methods

Having been involved in early field testing of cooling towers, I often wondered as to the methodology employed to test for contamination. Using, as a field test, Trypticase Soy Agar (TSA) with 10% sheep blood as an indicator of potential problems, I learned that cooling towers were cesspools of blood-devouring bacteria, made more so by the “biodegradable” inhibitors and “biodegradable-”cides added to them. I knew that “total counts” were of no value, which was recently confirmed in a four-year study in Australia. However, high TSA counts at least provide an indication of unhealthy conditions. Everyone searches for an “easy” way to do something. An early, inexpensive method for testing contamination probability in cooling towers was the use of “ Nutrient Agar Dip Slides.” The dip slide total bacteria test is totally worthless as it fails to be an indicator of anything, yet a colleague defended the method stating it was better than nothing. However, a high or low reading on a dip slide implies nothing, so why go to the trouble of using dip slides if meaningful information is not derived from them? There are many “good” bacteria in nature (e.g., lactic group) that will grow on nutrient agar. To reiterate reasons why total counts are not meaningful:

  • Bacteria are finicky as to their nutrients. This is the main problem in early Legionella detection. Nutrient agar must be formulated for the species in question.
  • Bacteria are finicky about their environmental temperature. The human population density at the poles is less than at the equator. Bacteria behave similarly.
  • Bacteria are finicky about environmental pH. If human blood pH varies from its optimum, illness and death will ensue; so with bacterial nutrient and environmental pH.
  • Bacteria are finicky about their environmental air or lack of same.
  • Bacteria are finicky about the makeup of their environmental air. Carbon dioxide, oxygen, and/or nitrogen? What combinations?
  • Bacteria, in nature, do not grow as single colony entities as illustrated on agar plates. Rather, they grow as a metropolitan entity within slimes of interacting bacterial species.

If one were to use a nutrient specific for Legionella sp. (e.g., Buffered Charcoal Yeast Agar [BCYA]) and incubate at the correct temperature range within a chamber at optimum atmosphere, one could draw a “conclusion” of sorts until the testing protocol is challenged. In addition, the count on BCYA may not be all Legionella; the colonies have to be tested and confirmed as to species and serogroup (SG). However, such a count has a greater value than TSA or Dip Slides as an indicator of contamination.

The first step is to obtain a water sample and send it to a laboratory3 for analysis, incubation, and serotyping. A recent innovation that looks promising is the Binax Field Test for Legionella, albeit it is specific for L. pneumophila, SG1 (the major killer in the United States). If that test proves positive ie. >100 cfu/ml for potable water and > 1,000 cfu/ml in cooling towers, then one can take steps to address the issue. Advantages to this test are that results are provided in under 20 minutes with no paper trail, if that is a concern.

This is all well and good. However, one has to look at the sampling technique and ascertain the objective. Are we looking for Legionella sp. per se, or for potential system contamination?

Testing Objectives

The objective should be to root out potential system contamination, especially in areas that expose the at-risk and general population to Legionella. An early method involved taking a sample of circulating water in a cooling tower or hot water system and ascertaining its Legionella count. Finding it to be <1 cfu/ml, the client was placed at ease and probably decided to do nothing since the “charts” showed the system to be at low risk.

However, was the water really at <1 cfu/ml? What did we test for? We tested for the planktonic Legionella population, and were oblivious to the sessile and intracellular (protozonotic) population in the system — in capped ends, pipe threads, Wye strainers, rust barnacle, you name it.

During a recent Legionella seminar in Singapore on March 29, 2000, I had the opportunity to discuss approaches toward testing with Dr. V.L. Yu, chief of infectious disease at the V.A. Medical Center, Pittsburgh. He recommends swabbing surfaces of plumbing systems in hospitals with the objective being to search for sessile populations. While the test is qualitative rather than quantitative, it is certainly applicable within the health care arena.

Testing Procedures

If we want to test a system for Legionella risk, we should attempt to address the total population and take those steps necessary to uncover the true colony count. Following are suggested testing methods.

For cooling towers, evaporative condensers, and closed-circuit coolers, I recommend the following: Take a sample of circulating water in a sterile bottle and have the staff immediately send it via next day air it to a laboratory4 equipped to do complete testing. The alternative is to use a Binax Field Test kit that detects L. pneumophila down to 100 cfu/ml in 15 to 20 minutes. While that is being done, add a low foaming, non-chlorine detergent to the system, at correct dosage for total system water, and take samples at 10, 20, and 30 minutes and have these sent to the same lab. Compare the results. If the count is higher after cleaning, then there is (was) a sessile population in the system that must be addressed. It is understood that the intracellular Legionella population is not accessed with this technique. However, the detergent should help remove this population during flushing as long as it is being done correctly and conscientiously.

For potable hot water systems (potable, not hot water heaters): Testing for planktonic and sessile Legionella population in a potable circulating hot water system is challenging, to say the least. Hospitals, hotels, motels, and nursing homes present unique challenges due to the nature of operation and alleged short cash reserves. However, if we are to address the issue, we must meet the challenges head-on and define the methodology. This includes:

  • Test system for planktonic population; and
  • Test system after an upset without placing

public/staff/residents at-risk.

The first is easy; the second is the problem. If we have two potable hot water storage tanks, we can valve one off and use a separate pump to agitate the first tank with a suitable cleaning agent. Lacking a second tank, we can agitate the tank with air during a very low-use time (e.g., 2 a.m. to 3 a.m.) and increase the temperature to >140˚F shortly thereafter to kill the disturbed Legionella population. Adding some tri-sodium phosphate to aid in the cleaning process can’t hurt. If the counts increase after the cleaning process, we know we have a potential amplifier that needs preventive measures or scheduled maintenance to address the issue. This is something that must not be put off since the sessile and protozonotic populations have been aroused; they are angry and looking to vent their anger on someone.

For home hot water tanks, I’ve suggested that tanks be operated at the maximum temperature setting. Mine can reach 160˚F at the faucet, and yes, I have thermal cutoffs at the bathroom faucets to prevent guests from being scalded. This is especially important in the Northeast, where retirees flock to the South along with the birds, thus reducing temperatures in their hot water tanks to conserve energy. When they return, they will be subjected to whatever has grown inside the tank during the winter months. One should not forget the stagnant potable piping.

Let us also not forget lawn sprinklers that have been dormant for some time. Who knows what can grow therein. Look at the age of the population that plays golf. Look at the age of the population that likes to take casual walks in the evenings.

A Population at Risk

Knowing what we now know, let us look at the data presented in a New York City study5 and see if it truly addresses the at-risk and general population. The hot water study involved samples from 107 commercial buildings, and the cooling tower study involved 51 samples over a four-and-a-half-year period. The sample/time presented does not allow for drawing focused conclusions to address the health of the at-risk and general population. This does not take sample technique or protocol into consideration. To infer, from such a small sampling, that risk is low is counterproductive to the objective of prevention. A recently released four-year study6 (4/1/92 to 3/31/96) of 28,898 environmental samples from cooling towers showed that 51% were <100 cfu/ml; 100-999 cfu/ml came in at 36.7%; and >1,000 cfu/ml at 12.3%. This study showed that 49% were >100 cfu/ml, which is the Australian standard for action. Of these, 71% were L.p SG1; 27% SG 2-14; and 2% other SGs.

At present, there is a test that is used to screen for Legionella exposure in humans (Binax, now). It is fast, taking 15 to 20 min, and could serve to prediagnose symptoms. Is it Legionella or the flu? The downside is that it tests for exposure to Legionella within the last 12 months, not if one is currently ill. But for $45 or so, the test can determine if a trip to the doctor or hospital is warranted. But who has the flu in March? Yes, it is easy to armchair philosophize; presenting prevention techniques is the challenge. Using copper/silver electrodes in treating plumbing systems is interesting, however if one wanted a specific concentration of silver and copper ions in water, it would be easier and more exacting to add solutions of same from containers via a chemical feed pump.

However, this author does not suggest that approach. For a cold potable system, which feeds the hot water system, filtration with rechlorination is still the best avenue, assuming we’ve also met the pH criteria, if ozone is not an option. The hot water loop can be addressed by increasing the temperature above 140˚F. Use of instantaneous potable hot water heaters to maintain 105˚F circulating loops does not address the circulating loop and contamination induced by the tempering cold city water.

The use of ozone for Legionella control appears to be the best avenue yet. It kills, it is site-produced, and it does not form cancer-inducing trihalomethane (THM) in the water supply. The fly-in-the-ointment is the short life of the ozone molecule and the application technology. The techniques available for prevention are too numerous for this paper. However, many can be found in Legionnaire’s Disease, Prevention & Control. If you wish to learn more about Legionnaires disease on an ongoing basis, visit www.egroups.com/group/legionnaires_disease or www.q-net.net.au/~legion/. This is a group of more than 50 interested professionals knowledgeable about Legionnaires disease worldwide.

Protecting the Public

The purpose of this article was to address the “The mere presence of Legionella . . . ” mantra that pervades the hvacr and plumbing industry and provides a “loophole” or excuse for inaction by building owners and operating engineers.

If testing and high counts imply nothing, why test? Why did the Congressional Office of Compliance, informed of high Legionella counts in a cooling tower in the Capitol in May 1999, show concern by issuing a citation for allowing workers to be exposed? There was no outbreak, yet a government agency showed concern for its workers. Are workers elsewhere of less value? Is the exposed public of less value? Why test after an outbreak? This is akin to shutting the barn doors after the horses have escaped. We — water treatment consultants, professional and mechanical engineers, field operators, and others — must cease providing excuses when lives are at stake. Every outbreak is a preventable one, all illness and death needless, tragic, and costly. ES