An Open Letter to Public Health and Occupational Health Professionals
The rate of Legionnaires’ disease cases has steadily risen since 2003, with another notable increase in the number of reported cases beginning in 2012. The number of large outbreaks have been rising as well, for example Flint, Michigan in 2014 and 2015, New York City in 2015, the Sheraton Hotel in Atlanta, Georgia in 2019, and the Mountain State Fair Grounds in Asheville, North Carolina in 2019. However, these public-facing events likely represent only the tip of an iceberg since it is recognized that cases of Legionnaires’ disease are significantly underdiagnosed and under reported. (See Our Explanation) Most confirmed cases of Legionnaires’ disease (~95%) are deemed to be sporadic cases unrelated to an outbreak. The threat to public health, disruptive impacts of an outbreak, and costs to remediate building water systems have led to public health agencies recommending comprehensive prevention programs.
The reasons behind this unrelenting rise in reported cases of Legionnaires’ disease has been the subject of many conferences and public health meetings for many years. After an unexplained one-year increase of 70% was observed between 2002 and 2003 CDC staff published an article exploring possible contributing factors. (Hicks et al, 2006) While the research article examined a variety of possible factors that could have impacted the number of reported cases, including possible increases in urine antigen (UA) testing, possible improvements in sensitivity or specificity of the UA test, possible increases in reporting of infectious diseases by physicians, infection control staff, or health departments in general, and possible changes in state or national surveillance methods. Oddly, the authors focused on and concluded that increased rainfall was the likely culprit.
Legionnaires Disease And Rainfall
The average monthly rainfall for 1990–2002 from May to September for the Mid-Atlantic states was 10.4 cm compared to 15.7 cm from May to September 2003. The authors’ model predicted that an increased rainfall of 5.3 cm (2 inches) over 5 months corresponded to an increased Legionellosis risk of 14.6% (RR 1.146, 95% CI 1.067–1.231). Examination of rainfall records during these same periods (May to September) for the years 1993 through 2022 reveals that trends in rainfall have NOT always followed the upward trends in reported Legionnaires’ disease cases. (See Table 1) If the model offered in CDC’s 2006 paper held true and offered any predictive value for Legionnaires’ disease rates, the 5-month rainfall average for the Mid-Atlantic States would in 2011 (16.7 cm) that corresponded to 4,202 cases would have equaled the number of cases in 2018 that peaked at 9,933, when average rainfall was also 16.7 cm. If this model were a valid predictor of Legionnaires’ disease rates, then higher rates of disease should have been observed in 1996 when rainfall in these Mid-Atlantic States exceeded those recorded in 2006, 2012, and 2013. On the contrary, the number of Legionnaires’ disease cases reported in 1996 was 1,198, lower that levels seen any year subsequent to 2001. While rainfall may be a contributing factor to regional Legionnaires’ disease case numbers it is naïve to attribute ongoing and unabated rises in disease numbers to increased rainfall alone. The mechanism of increased rainfall’s influence on Legionnaires’ disease increases may be due to difficulty in treating surface waters that charge drinking water reservoirs. However, ultimately this would translate into lower levels of residual chlorine disinfectant reaching building water systems. One could see how a confluence of factors would complicate both drinking water treatment efforts.
Table 1: Selected Years Rainfall vs. Legionnaires’ Disease Cases
Source: https://www.weather.gov/wrh/climate
The most glaring omission from analysis in the 2006 article by CDC staff was the potential influence of regulating disinfection byproducts in large municipal water systems. The overwhelming importance of residual disinfectant levels and significant impact it has on the likelihood of a building water system becoming colonized by Legionella has since been reported. (Papadakis, 2018)
The Unintended Consequence of Federal Rules and Regulations
An unexamined, and unintended outcome of the disinfectant and disinfection byproducts (DBP) rule, stage 1 in 2002 was the reduction of chlorine disinfectant residual in municipal water systems across the United States. The most direct route to meeting regulatory limits for total trihalomethanes (TTHMs) was to reduce the amount of chlorine disinfectant introduced to drinking water, resulting in less residual chlorine reaching buildings. Low disinfectant levels in the water supplied to a building result in a rapid loss of residual disinfectant once in the premise plumbing, and a greater chance that Legionella will survive, colonize, and amplify to levels that pose a health threat.
Leaving this possibility unexamined in the 2006 article has allowed the conditions to persist and worsen ever since. The unrelenting increase in reported cases since 2003 is apparent in the data, with only a brief interruption by the COVID-19 pandemic societal shutdown and travel interruption. Until public health agencies reconcile the impact of reduced disinfectant levels in drinking water with the 20-year climb in reported cases of Legionnaires’ disease the rise of cases and outbreaks will continue unabated. Setting aside the lost opportunity to change course and revise the DBP Rule to ensure safe and effective levels of chlorine disinfectant residual are in drinking water that reaches buildings, this article should be revised or retracted by its authors and the CDC as it continues to be cited as an authoritative reference and perpetuates a myth that has not borne fruit.
Unfortunately, Legionella monitoring and prevention programs advocated by the CDC and ASHRAE since 2015 have had no discernable impact on the growing number of cases reported each year. A major barrier to effective prevention efforts is the limited arsenal of control methods that can reliably prevent Legionella colonization and growth within premise plumbing and drinking water systems of buildings. Almost all currently available methods, as described in the 2016 US EPA Review of Technologies for Legionella Control in Premise Plumbing Systems, involve adding potentially toxic and corrosive chemicals to water used for drinking, washing, and cooking, many of which are known to produce disinfection byproducts that may be carcinogenic. Federal and State laws regulating the introduction of supplemental disinfectants to building water systems make these control measures even less desirable for many building owners. Ultimately, the tools and technical means of effectively controlling Legionella contamination in buildings and preventing Legionnaires’ disease are very limited, often ineffective, and inherently introduce a number of health hazards by themselves. Supplemental chemical disinfectant injection to a building’s drinking water also requires that water be routinely flushed down the drain in order to maintain effective concentrations throughout the system and at distal fixtures, which unavoidably wastes precious water resources.
Treating drinking water with chemical disinfectants presents a trade-off between the theoretical risk of chronic chemical exposure and a conspicuous rise in waterborne diseases, most notably Legionnaires’ disease. Ideally, we need to find more effective non-chemical control measures to prevent Legionella amplification in building water systems, but until then delivering chlorine and other disinfectants at effective levels throughout municipal water systems is paramount to bringing Legionnaires’ disease rates under control. Ensuring that effective free chlorine levels actually make it to all buildings throughout a municipal water system, and not just to monitoring stations after exhaustive flushing occurs, is critical to see improvements. Beyond that, educating all building owners and operators on the risk of Legionella and what they can do to reduce the risk is needed.
The basis for enacting the Disinfectants and Disinfection Byproducts (DBP) Rule and lowering permitted limits was aimed in part at reducing cancer risks. Whether the benefits outweigh the increased occurrence of Legionnaires’ disease is subject to debate. Since cancers typically take up to 30 years to become diagnosable any benefits from reduced DBPs would not likely be observed until 2032 if at all. On the other hand, the rise in Legionnaires’ disease cases has been observed immediately after the DBP rule was implemented. If this coincidence is determined to be a causal association then public health policy needs to reconcile the risk tradeoffs and determine if reducing cancer risks are worth the resulting infections, outbreaks, and fatalities.
Historically, before 2015, implementing measures to control Legionella was recommended only after an outbreak of disease was detected by public health epidemiological surveillance. This laissez faire public health strategy resulted in cases soaring from 1,280 per year in 1993 to 9,933 cases in 2018. By examining the annual data on Legionellosis cases reported in the CDC Nationally Notifiable Disease Surveillance System, grouped into ten-year periods reveals some remarkable trends preceded by two inflection points.
From 1993 through 2002 the total number of reported LD cases was 12,576, or approximately 1,260 per year on average. Prior to 2002 the trajectory was relatively flat with 1,280 cases in 1993 and ending with 1,321 cases in 2002, with most of the intervening years reporting even fewer cases.
From 2003 through 2012 the total number of reported cases was 30,115, or approximately 3,010 cases per year on average. Starting with 2,232 cases in 2003 and ending with 3,688 cases in 2012. The rate of increase over this ten year period was notable, about 2.4 times more cases than in the preceding 10 years, and apparently led to policy changes by the CDC to endorse prevention programs.
From 2013 through 2022 the total number of reported cases was 69,511, or approximately 6,951 cases per year on average. Starting with 4,954 cases in 2013 and ending with 6,138 cases in 2022. The rate of increase over this ten year period was accelerated beyond that of the previous ten years despite introduction of the ASHRAE Standard 188-2015, a voluntary standard for building owners. By 2022 there were 2.3 times more cases than in the preceding 10 years and 5.5 times more cases than in the 1993 to 2002 period. This last ten year period also includes 2 years of the COVID-19 Pandemic, where travel and working from home have likely reduced the number of Legionnaires’ disease cases that would normally have occurred.
No matter how you parse the data and observe the trends, the number of people getting sick from Legionella bacteria is on the rise and shows no signs of abating.
So, what prompted changes in municipal water treatment and disinfectant levels?
Rules to reduce disinfectant levels in drinking water and mandated water conservation efforts have been introduced over the past 20-30 years. The need for these regulations and mandates to protect the health and safety of the public and to conserve limited energy and water resources is not in dispute. However, policy makers, regulators, utilities, building owners and operators, and the general public must recognize the impacts of these unintended consequences, and develop strategies to address them. Current guidance to monitor and make insignificant adjustments to hot water temperatures or to remove dead leg plumbing lines are not enough to overcome the deficit in disinfectant levels. The following examples constitute just some of the mandates that can increase the risks for Legionella growth in building water systems.
Hot Water Temperatures: The Uniform Plumbing Code and most State Health Regulations prohibit the delivery of water above 122 degrees F. This often means that water heaters are set to operate at temperatures that are ideal for Legionella growth. In order to inhibit Legionella growth or inactivate it, water must be maintained and circulated above 60 degrees C (140 degrees F). Without installing special equipment, called thermal mixing valves, operating water heaters at these temperatures creates a scalding risk. Notably, in 2021 the CDC reversed its earlier recommendations to use thermal pasteurization or heat treatment as a remediation method for building water systems. (CDC Toolkit, 2021) Similar recommendations that focus upon maintaining water temperatures to control or prevent Legionella’s growth remain though. The ineffective and limited impact of increasing heated water temperatures has been recognized for some time and may have been a mechanism for selecting for more “heat-tolerant” strains of Legionella. Public health recommendations to control Legionella by increasing water temperatures alone were not well thought through, contradictory to energy conservation recommendations, and in the end ineffective.
Water Conservation Efforts: The National Energy Policy Act of 1992 established efficiency standards for all toilets, faucets, and showerheads sold and installed in the US. The law mandated low flush toilets and prohibited the installation of toilets that flushed more than 1.6 gallons (6 liters) of water, where toilets installed before 1994 typically used 3.5 gallons per flush. Additional legislation required improvements in the efficiency of washing machines, dishwashers, and other water using appliances. This and other Federal, State, and Local requirements to reduce water usage have been largely successful. Between 1998 and 2015, the average amount of water used per person per day dropped from 172 gallons to 82 gallons, a 52% decrease, excluding industrial and agricultural uses.
When water usage rates drop, the time it takes for water to transit the distribution system and premise plumbing systems increases, referred to as “water age”. Increasing water age results in loss of chlorine residual due to reactions with organic matter, inorganic surfaces, and bacteria. This results in water being delivered to buildings with levels of residual chlorine so low that they are unable to prevent Legionella growth in the premise plumbing systems.
Comparing the Risk of Reducing Disinfectant Byproducts (DBPs): Excessive amounts of disinfectants such as chlorine may be harmful, and even chronic exposure to low levels of disinfection byproducts may be carcinogenic. Implementing the DBP Rule ultimately reduced the levels of disinfectants in drinking water that are needed to control pathogens such as Legionella. Rates of reported disease appear to have risen the following year (2003). Reducing DBPs is believed to offer a reduction in the number of bladder cancers over time, however since the latency of cancer diagnosis may be 30 years from initiation of exposure, we cannot expect to see a benefit from the DBP Rule for another 10 years if at ever. The coincidental timing of the DBP Rule implementation and the dramatic rise in Legionnaires’ disease cases warrants further investigation to evaluate a potential causal link. Despite delays, Stage 2 of the DBP rule was finally implemented in October 2012, which preceded another increase in the rate of Legionnaires’ disease for the next ten years.
In order to control Legionella and other pathogens in domestic drinking water maintaining effective levels of residual disinfectants is critical. In a study of Legionella sources in hotels and resorts that examined the risk of colonization, inadequate residual chlorine levels (<0.20 mg/L) posed a relative risk of 54.78 (95% CI 20.47 – 148.04). This study supported observations and anecdotal findings of many practitioners when it concluded that the risk of Legionella contamination due to inadequate residual chlorine levels far surpassed any other risk factor, such as favorable water temperatures and stagnation. (Papadakis et al, 2018)
The possible additive effect of federal rules, regulations, and programs to reduce DBPs and conserve water and energy described earlier have directly and indirectly lowered the levels of residual disinfectant in domestic water supplied to buildings throughout the United States since 2002. The following twenty years have seen an unprecedented rise in Legionnaires’ disease rates that show no signs of ending. Because the Safe Drinking Water Act (SDWA) prohibits the supplemental treatment of building water systems unless a permit is issued and the facility is regulated as a public water supplier, building owners and operators have few meaningful and effective tools to prevent or control Legionella contamination. Until the unintended downstream effects of the federal rules and regulations are acknowledged and measures to deliver microbiologically safe water that contains effective levels of residual disinfectant to all end users, Legionnaires’ disease rates will continue to rise.