Community Legionella outbreak linked to a cooling tower, 2022

Abstract

Background: Thirty-five laboratory-confirmed legionellosis cases were reported to the Simcoe Muskoka District Health Unit (Ontario, Canada) between September 27, 2022, and October 15, 2022, resulting in one death and 29 hospitalizations. This article describes the Legionella outbreak and highlights activities for managing the outbreak, including various environmental and infrastructural controls associated with the public health response and some of the unique challenges and potential solutions to mitigate future outbreaks.

Methods: All cases of legionellosis were reported to and investigated by the local provincial health unit. Within a 6 km radius around the community, 27 cooling towers (CTs) were identified as potential sources of Legionella. Environmental samples were collected from 19 CTs and a long-term care home.

Outcome: Of the 35 cases, 29 (83%) were hospitalized (including three long-term care residents) with two requiring intubation/ventilation. Of the five sputa (clinical isolates) collected from confirmed cases, four tested positive for Legionella pneumophila (one was positive for L. pneumophila serogroup 1—with the same sequence type as one of the CT isolates). Education and recommendations were provided by the local provincial health unit to operators to improve CT operation.

Conclusion: Detection and management of community legionellosis outbreaks associated with CTs involve resources and time to properly identify and control risks. Measures for community risk mitigation included coordinating with provincial and community partners, developing methods to rapidly identify CTs as a likely source of infection and applying operational/maintenance/testing standards for CTs to control bacterial growth and minimize the dispersion of contaminated aerosols.

Introduction

Legionellosis is caused by the Legionella bacteria; most commonly Legionella pneumophila ^{Footnote 1}. While the bacteria are commonly found in natural, freshwater environments, it can become a health concern in human-made water systems (e.g. plumbing systems of large buildings, cooling towers [CT], certain medical devices, decorative fountains) where conditions allow it to multiply. People contract Legionella by inhaling aerosolized water droplets containing the bacteria or, less commonly, by aspiration of contaminated drinking water. Legionellosis may present as Legionnaires’ Disease or Pontiac Fever with symptoms including anorexia, malaise, myalgia, headache, productive cough, fever, pneumonia, confusion, chills, nausea and diarrhea^{Footnote 1}. Ontario provincial data^{Footnote 2} reported a rise in legionellosis incidence from 2012 (rate=1.4/100,000; 191 cases) through 2021 (rate=2.6/100,000; 385 cases) with individuals 50 years of age or older representing most reported cases. Legionellosis is a disease of public health significance under the Government of Ontario’s Health Protection and Promotion Act ^{Footnote 3} and is reportable to the local provincial health unit where the individual resides. No province-wide regulatory framework associated with inventory, design, maintenance, operation or testing of CTs for the purpose of mitigating Legionella outbreaks exists in the province of Ontario as observed in some other Canadian jurisdictions^{Footnote 4Footnote 5Footnote 6}. However, following the 2019 community outbreak, a provincial public health association recommended developing a CT registry, in addition to risk management and maintenance plans for owners and operators^{Footnote 7}. As of the date of this publication, no provincial CT registry exists in the province of Ontario.

Thirty-five laboratory-confirmed cases of legionellosis were reported to the Simcoe-Muskoka District Health Unit (Ontario, Canada) between September 27, 2022, and October 15, 2022. An outbreak of legionellosis in the same community in 2019 resulted in 10 cases (all admitted to hospital) and one death. A CT was linked by laboratory testing with one of the cases. Given the 2019 outbreak, and recognizing that CTs are the most frequent source of large community outbreaks^{Footnote 8Footnote 9Footnote 10Footnote 11Footnote 12}, CTs were investigated in 2022 as a source.

This report describes a community outbreak of legionellosis in the fall of 2022 in Ontario. The article focuses on the epidemiology and investigative processes leading to the outbreak’s conclusion. The article highlights the continued challenges that public health practitioners experience in the identification and control of Legionella in relation to CT units and how a registry could facilitate rapid identification of CTs in a region of interest, as well as how guidance and/or regulations could support ongoing maintenance and monitoring as preventive public health measures.

Methods

Case investigation

The first confirmed case of legionellosis was reported to the health unit on September 27, 2022. On October 4, 2022, four cases of legionellosis were admitted to the local hospital, presenting with pneumonia. Analysis of case movement and exposures revealed that all four cases resided or worked in the same community based on postal code. Given the monthly expected case counts, based on the five-year mean for the public health region is between one and three cases in September and October, the observed case count (n=4) in a single community was aberrant. Accordingly, an outbreak was declared by the health unit on October 4, 2022. An outbreak case was defined as any individual who lived, worked or visited the identified community in Simcoe County, with compatible signs and symptoms of legionellosis, on or after September 5, 2022, and who had laboratory confirmation of legionellosis. The outbreak investigation was initiated on October 4, 2022. Using the seven-day median reporting time from symptom onset plus the maximum incubation period of 15 days, in addition to allowing for two additional days of coordinating communications, the outbreak was declared over on November 8, 2022, which represented 24 days from the last reported case (October 15, 2022). The outbreak lasted 35 days.

Urine specimens were submitted to the Public Health Ontario Laboratory. Urinary antigen tests were performed by the laboratory for all suspect cases. Upon report of a new case, the health unit conducted a case interview and exposure history that included the 14 days prior to symptom onset, identifying home location, work location and any other locations visited during this period. An exposure associated with this outbreak was defined as any location within the area of investigation, where the case spent any amount of time during their period of acquisition (e.g. workplace, home, shopping, day excursion).

Interventions and environmental investigation

Given the hospitalization and laboratory confirmation of five cases of legionellosis with pneumonia at the local hospital from September 27 to October 4, 2022, the health unit initiated an investigation into potential environmental sources. Community and healthcare providers were informed of the outbreak to assist in facilitating case finding, prompt diagnosis and management of persons with potential legionellosis. The public was notified of the outbreak via local media to 1) inform those who may be at risk, 2) describe clinical signs and symptoms, and 3) recommend that those who are symptomatic seek medical attention if their symptoms are severe or do not resolve.

A review of case investigation data revealed that no single exposure site or activity was common among the five cases, other than travelling to, living or working in the community. A line list of cases with exposures was developed and, on review, no common exposures for the cases were identified other than geography suggesting an environmental exposure covering an area defined by a 6 km radius. Through elimination of other potential sources, including municipal drinking water, CTs were identified as the probable source. On October 5, 2022, 15 CTs were identified within the 6 km radius using location data collected during the 2019 outbreak. A 6 km radius was established based on previous literature^{Footnote 13} that described this radius as a reasonable distance for aerosol dispersion. The health unit requested operators of these known CTs to immediately cease operation unless deemed by the operator as essential to the facility’s operation. The CTs that remained operational were required to provide maintenance and testing records from July to October 2022 and directed to arrange for immediate cleaning and disinfection.

The owners/operators for 11 of 15 CTs provided monitoring data collected through routine Legionella maintenance programs. All quantitative Legionella culture results were reported as fewer than 10 colony forming units (CFU)/mL for Legionella species. According to the Canadian guidelines for federal buildings^{Footnote 14}, a value of fewer than 10 Legionella spp. CFU/mL is considered within the acceptable limit. Laboratory data for the remaining four CTs were not available from the CT owner/operator as no sampling program was in place. From October 8 to November 1, 2022, public health inspectors identified additional CTs (total n=27) and collected water and biofilm samples from 19 CTs for submission to the Public Health Ontario Laboratory. Eight other CTs could not be sampled due to seasonal shut down. During the investigation, one operator/owner was ordered to shut down their CT due to unsanitary conditions, lack of routine maintenance and lack of testing. The CT previously identified as the source of the 2019 outbreak could not be immediately sampled by public health officials due to a mechanical failure that rendered it devoid of water.

On October 17, 2022, the health unit received culture results from a CT showing L. pneumophila serogroup 1 levels of 2,575 CFU/mL. The sample was collected on September 27 and tested as part of routine monthly testing. The health unit arranged to have the September 27 isolate transported to the Public Health Ontario Laboratory for sequence-based typing (SBT). Sequence-based typing was performed to determine the relatedness of the clinical and environmental isolates. The SBT method at Public Health Ontario Laboratory is based on the epidemiological typing scheme for clinical and environmental isolates of L. pneumophila, which was developed by members of the European Legionnaires’ Disease Surveillance Network and evaluated for implementation in the investigation of outbreaks of legionellosis caused by L. pneumophila. Culture results can take up to 14 days to be received, which can delay interventions. In this case, no other immediate action was required given the CT had already been shut down on October 7 due to mechanical failures. Between October 17 and November 10, the health unit also conducted quantitative polymerase chain reaction (qPCR) testing using field equipment to quantify results for further decision-making. The Federal Standard guided the corrective actions taken and the interpretation of results. The CTs selected for qPCR testing were prioritized based on 1) lack of maintenance and testing, 2) missing test results or 3) requirement to further evaluate samples from those CTs where qPCR and culture showed the presence of Legionella spp.

A separate environmental investigation took place for a sub-cluster of cases identified at a long-term care home (LTCH) within the 6 km radius. Three residents of the LTCH were hospitalized and determined to be L. pneumophila serogroup 1-positive based on urinary antigen tests. Due to the case investigation, two shower rooms were identified as potential sources and sampled on October 8, 2022. All samples yielded negative results by culture.

Epidemiologic and statistical analyses

Descriptive epidemiology was used to enumerate total cases, the severity of cases and their outcomes. An epidemic curve was created in Microsoft Excel (Figure 1). It displays cases over time, with differentiation between community cases, cases residing in a LTCH, and cases from out of jurisdiction of the local health unit. The environmental investigation results of CTs are displayed above the cases over the same time scale.

To explore sources of disease, ESRI ArcGIS Desktop^® 10.6.1 was used to map cases by home location and assess clustering in investigation, followed by mapping of cases’ exposure locations and 27 known CTs identified during the outbreak period (and the testing results, as available). A hotspot analysis was undertaken using the Optimized Hot Spot Analysis Spatial Statistic Tool using a 6 km radius buffer around the area of investigation and a cell size of 200 m.

Hotspot analysis and exposure history analysis were rerun as new confirmed cases were identified and the outbreak progressed. Upon identification of a sequence type (ST) match between the isolate cultured from a case’s sputum and an isolate cultured from the implicated CT, assessment of each case’s nearest exposure to the CT was conducted.

Results

Thirty-five laboratory-confirmed cases were identified in this outbreak and were associated with having a residence and/or having visited locations within a 6 km radius of a community within the jurisdictional area of the local health unit. Of the 35 laboratory confirmed cases, 29 were hospitalized. One case died, resulting in a case-fatality rate of 2.9%. Twenty-six cases resided in the community, seven cases were LTCH residents, and two cases were residents of other communities who visited the area during their acquisition period. Table 1 shows a breakdown of cases by demographics (age, sex, location), health conditions and severity. Twenty-nine of the 35 cases (83%) were hospitalized, including three of the LTCH residents, with two hospitalized cases admitted to the intensive care unit requiring intubation/ventilation.