As time has progressed, Legionnaires' disease - a severe form of pneumonia caused by the Legionella bacteria - continues to be a public health concern. The overall trend over the past 20 years has shown that the detection of Legionnaires’ disease, a Center for Disease Control and Prevention (CDC) reportable illness (as legionellosis), has steadily increased. Due to the significant mortality rate, Legionnaires’ disease is under high surveillance by nation-wide public health departments and the CDC, but how often is the disease underreported? The CDC suggests that underreporting is up to 2.7 than was reported annually, while other estimates have placed that value at 8 to 10 fold. An increasing array of changes and delays are responsible for this issue. Using CDC data from illness reporting data, Legionnaires’ disease underreporting phenomena will be illustrated with data below.
Legionnaires Disease: Understanding the Threat
Legionnaires' disease is typically caused by inhaling aerosolized water droplets containing Legionella bacteria. These bacteria thrive in warm water environments, such as cooling towers, hot tubs, showers, and decorative fountains. However, assessments of facilities show that there are other areas of a facility’s water system that could contribute to the presence and growth of Legionella bacteria. Human risk factors exist (e.g., age, gender, immunocompetency) that increase the chances of Legionella impacting the body.
What has been the Recent Legionnaires Disease Incidence Trend?
In order to evaluate that question the overall trend during Legionnaire’s disease tracking should be examined. Figure 1 shows the evolution of tracking Legionnaire’s disease over time since 1976.
Figure 1: With the exception of 2020 as an outlier during COVID-19, It is obvious to see that the number of annual cases has been steadily increasing since 2003, after years of remaining relatively stable.
While the annual numbers tell one story, can weekly CDC reportable totals compared with these tell another? A comparison of weekly reportable can provide insight on how many cases are being underreported when compared with the annual numbers.
How Can Weekly Legionellosis Reporting Illustrate Legionnaires Disease Underreporting?
The CDC’s National Notifiable Diseases Surveillance System (NNDSS) enables public health agencies to share information on infectious diseases (e.g., legionellosis) that the Council of State and Territorial Epidemiologists (CSTE), in consultation with CDC, has designated as nationally notifiable - i.e., Legionnaires' disease.
Through the NNDSS, cases of Legionnaires' disease (as legionellosis) are reported weekly and displayed in tables for ease in tracking the number of cases for each US state and some territories. Weekly reportable disease case reporting began in 2006 and has continued since. Annual tables are released at the end of each year, which show the total number of cases of Legionnaires Disease for that particular year. These annual numbers are often revised and there is a lag in reporting of these values. For example, 2020 is the most recent year that annual tables for reportable illnesses are available.
While CDC warns against using weekly legionellosis for understanding prevalent disease impact, the weekly data can be used in an underreporting analysis to illustrate how often the disease can be overlooked-this is especially important for facilities trying to understand the impacts of the Legionella bacteria on facility water quality and management.
Below is a sample of data from CDC WONDER (Wide-ranging Online Data for Epidemiologic Research) Weekly Infectious Disease Tables (Legionellosis Week 37, 2023):
Figure 2: CDC Wonder Table. The Week 37 reports 56 cases of legionellosis with a cumulative 2023 year to date (YTD) of 4,398 cases. CDC provides a weekly update of this table as the year progresses.
Legionnaires Disease Underreporting Using Weekly Data
For underreporting analysis, the current week totals for each week during the year have been summed and compared to the yearly totals that CDC reports on an annual basis, when available. Weekly data can be used to illustrate the continued and increasing underreporting that is occurring for Legionnaires disease cases. In the figure below, weekly incidence of Legionnaires' disease cases for respective years were totaled and compared with the annual total as reported at the end of each year by CDC. The data in Table one reflect the weekly percentage of the annual total, and the underreporting based on the weekly percentage.
Table 1: Yellow is underreporting at 60% or greater. Red is underreporting at 70% or greater. *= Based on cumulative weekly total because CDC has not finalized annual numbers for these years.
A visual representation of the underreporting using weekly data can be viewed in Figure 3. As illustrated and from Table 1, weekly totals are underreporting the total cases by at least 50% for most of the years data are available. In the years since COVID-19, the underreporting has increased to an average of 75%.
Figure 2: Orange represents the CDC annual totals for legionellosis (i.e., Legionnaires' disease) since 2006 and the blue represents the sum of the weekly totals as reported by CDC- when weekly reporting numbers were first available.
While the weekly summed totals have always lagged the annual numbers, there has been an increasing trend in recent years of annual disease underreporting e versus weekly reporting. Figure 3 below highlights the increasing trend in underreporting since weekly tracking started.
Figure 3: Legionnaires' disease underreporting trend since 2006.
It is clear through the weekly data that underreporting of disease incidence is an increasing issue, especially in recent years. Below the possible contributing causal factors to these increases will be explored.
Legionnaire's Disease Underreporting Causal Factor Analysis
Updated 2019 CSTE Case Definition
The Council of State and Territorial Epidemiologists (CSTE) is an organization consisting of epidemiologists serving public health from states and territories in the United States. CSTE provides technical advice on infectious diseases to the CDC, including CDC adopting the CSTE legionellosis case definitions.
In June 2019, CSTE updated the Legionellosis definition for identifying reportable illnesses. The last case definition update was in 2005 and had legionellosis associated with two clinically and epidemiologically distinct illnesses: Legionnaires’ disease and Pontiac fever. Additionally, CSTE identified only two case classifications in 2005: suspected and confirmed.
With the new 2019 definition, CSTE added a "probable" case classification which is identified as a "clinically compatible case with an epidemiologic link during the 14 days before onset of symptoms". A new "extrapulmonary legionellosis" case description was also added during this update. See Figure 4 for underreporting since CSTE case definition change.
Figure 4: Legionnaires' disease underreporting trend since 2019 CSTE case definition update.
The underreporting trend following the update is marginal, but the impact could be hampered by the short amount of time that has passed - and the impacts of COVID on disease identification- since the updated case definitions.
Implementation of ASHRAE Standard 188
In June 2015, the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) released Standard 188 titled " Legionellosis: Risk Management for building Water Systems". The standard outlines specific assessment and control areas that building designers, owners, and operators shall follow in regards to managing legionellosis risk within building water systems. It also details general requirements necessary for any building that may need to make use of the standard’s guidance in reducing the risk of legionellosis for building occupants. ASHRAE developed this standard specifically aimed at preventing the growth and spread of Legionella. It defines types of buildings and devices that need a water management program as well as devices (e.g., hot tubs, cooling towers) that need to be controlled in order to prevent the growth and spread of Legionella.
Figure 5: Legionnaires' disease underreporting trend since ASHRAE Standard 188 released.
The purpose of ASHRAE 188 was to establish minimum risk management requirements for legionellosis risks in building water systems. That is, implementation of risk management techniques to minimize the incidence of legionellosis within a facility that originated with the building water system. However, based on stratification of the CDC data in Figure 5, underreporting has continued to increase since the implementation of the standard.
Implementation of EPA Disinfectant Byproduct Rule
First proposed in 1994, the Stage 1 and stage 2 Environmental Protection Agency (EPA) disinfection byproduct (DBP) rules were intended to protect public exposure to constituents formed during the disinfection process when an antimicrobial (e.g., chlorine) interacted with organic material (e.g., bacteria or other carbon-containing water contaminants). Stage 1 implementation was required to begin in 2002 for systems servicing>10,000 people, and 2004 was the set implementation deadline for smaller water utilities. Stage 2 implementation was also staggered, with 2012 as the implementation deadline for systems servicing >50,000 people and 2013 as the dead line for smaller water utilities.
While it is not possible to evaluate the underreporting data from 2002 (because the weekly reporting did not start until 2006), the data from the Stage 2 DBP rule can be stratified for an understanding of the impacts the rule had on incidence reporting.
Figure 6: Legionnaires' disease underreporting trend since EPA DBP Stage-2 implemented.
While there is an initial reduction in underreporting (e.g., 2014 and 2015), the overall trend is increasing. An interesting observation in the data following the 2012 DBP implementation for larger water systems is the first instance in the weekly reporting data of more than 100 legionellosis cases (2013: Week 26 - 105 cases).
Legionnaires' Disease Underreporting Discussion
There is no doubt that Legionnaires disease cases are underreported. The analysis above used weekly data to understand underreporting as it related to annual legionellosis incidence. These data can illustrate how delays or changes in external factors (e.g., identification of disease or confirmation via CSTE case definition) can impact our understanding of disease incidence in the moment.
While weekly data stratification can allow an evaluation of the impact that certain changes have had on underreporting, the main culprit may be the bureaucratic process that is imbedded in disease tracking. For example, CDC warns against using weekly data (they identify it as "provisional") and highlight the unreliability of using weekly values to compute a cumulative annual total. From CDC:
"Current Week: For a case to be published in the table under current week, it must have been reported to CDC during that week and assigned by the jurisdiction to that MMWR week. Cases assigned by the jurisdiction to that MMWR week but reported later or cases reported in that week but assigned to a different MMWR week are published in the CDC WONDER weekly tables in the cumulative total column for that year, but are not published in any current week column. As a result, the cumulative sum of cases does not equal the sum of the number of cases published each week."
AND
"...weekly totals should not be added to compute the cumulative count for a year. It is not uncommon that cases are reported after data has been published for the week, or that updates change whether the case meets the case definition or publication criteria. Less commonly, a case definition is modified or the criteria are clarified during the year, resulting in a reevaluation of previously reported cases." (emphasis added)
If so many revisions or delays in reporting are occurring, perhaps there is a process issue that should be addressed. CDC also states, the following, which is in contrast to their statements above:
"Public health surveillance of national notifiable infectious diseases and conditions helps public health authorities monitor the effect of these diseases and conditions, measure the disease and condition trends, assess the effectiveness of control and prevention measures, identify populations or geographic areas at high risk, allocate resources appropriately, formulate prevention strategies, and develop public health policies."
If data needs constant revisions, than these efforts are hindering more than they are helping understand disease prevalence. The analysis of the legionellosis case definition update in Figure 4 show a moderate impact, but there are other factors that cannot be account for in this analysis. For example, the detection of this disease by public health officials in recent years has been hampered by COVID-19 due to similar disease outcomes (e.g., pneumonia, fever, etc.). Because the case definition changed the latter half of 2019, there the impact of the change (and bureaucratic red tape for case confirmation) may be masked by COVID prevalence, surveillance, or recognition, among other factors. This is a plausible outcome because the average underreporting during the COVID years (i.e., 2020-2022) is 75%. (Table 1). This is further supported by the fact that no finalized annual legionellosis numbers have been released since 2020, even though CDC guidance states that final numbers should be published approximately 10 months following year end.
Beyond the layers of reviews and re-reviews that are taking place dur to the changing epidemiological definitions for disease surveillance, other factors are also influence reporting (and thus, underreporting). Ass Dr. David Krause points out in his blog the impact the DBP rules have had on the ability for control of pathogens like legionella in drinking water systems. The weekly data illustrate that impact because there were no weekly reporting of 100+ cases of Legionnaire's disease until the full implementation of the Stage 2 BDP rule:. See Figure 7.
Figure 7: Number of 100+ legionnaire's disease days in each year. Note that the one day reported in 2020 appears to be an outlier, because there were 129 cases reported the last week of the year (i.e., Week 52 in December 2020), which is an unrealistic number for the winter months.
If cases were increasing following the DBP State 2 implementation, why did weeks of 100+ cases cease after CSTE change the case definition? Perhaps the revisions that were necessary were creating such a lag that weekly underreporting was increasing more than before, as the data indicate with 75% underreporting during 2020-2022.
The implementation of ASHRAE 188 (and other efforts, like those taken by Centers for Medicare and Medicaid Services) was to increase understanding, identification, and management of legionellosis risk in facility water systems. Based on Figure 5 above, underreporting is an increasing problem, even since the release of the standard. However, if a side effect of the standard is to increase awareness of the risk, why is the underreporting trend increasing?
Based on the analysis above, there are a number of factors that can and are influencing Legionnaire's disease underreporting. There are also likely factors not analyzed above that are also contributing (e.g., CMS requirements for reducing Legionella risk in healthcare facilities). The data above are not to assign a causative reason in his article, but to illustrate that increasing bureaucratic layers of constant revisions and reevaluating based on case definitions and confirmatory tests is masking true incidence on a weekly basis and leading to revisions and underreporting on an increasing scale. Overall active Legionnaire's disease surveillance and and underreporting of the true prevalence appears to take a back seat to other diseases (e.g., COVID-19) and the trend appears to be increasing.Legio