Title : Cooling tower culprit or community clue? A modern approach to legionella surveillance
Abstract:
Introduction: Legionella disease (LD) is primarily transmitted through the inhalation of aerosolized water droplets containing Legionella bacteria, often originating from water sources such as cooling towers. In August and September 2021, an Infection Prevention team at a New York City acute-care hospital identified an outbreak of Legionnaires’ disease affecting 13 patients across two adjacent zip codes in northern New York City. This study evaluates the utility of the Legionella Early Risk Assessment (LERA) tool in diagnosing Legionella-associated healthcare-associated pneumonia (HAP) and its potential role in mitigating risks associated with building water systems and preventing future outbreaks.
Methods: A two-phased approach was implemented. In the retrospective phase, electronic medical record surveillance identified pneumonia cases occurring ≥3 days post-admission, followed by Legionella testing if not previously performed. The prospective phase, from September 2021 to February 2022, employed the LERA tool to identify LD cases. Environmental surveillance included testing three cooling towers, one of which tested positive for Legionella species. Statistical analysis, including Spearman correlation, was performed to assess relationships among variables due to the small sample size.
Results: Significant correlations were observed between the number of patients identified/suspected of HAP and those tested for Legionella (p < .04), as well as between the number of HAP patients tested for Legionella and those tested using the urinary antigen method (p < .04). However, no significant correlations were found between cumulative hospital-wide patient days and the number of patients identified/suspected of HAP (p < .34), the number of HAP patients tested for Legionella (p < .73), or the number of patients tested using the urinary antigen method (p < .91). Additionally, no significant correlation was observed between the number of patients identified/suspected of HAP and those tested using the urinary antigen method (p < .23).
From September 2021 to February 2022, six patients were identified/suspected of HAP, and four were tested for Legionella. All patients identified/suspected of HAP and those tested for Legionella were associated with periods where cumulative hospital-wide patient days exceeded 5100. Notably, in January 2022, despite 6730 hospital-wide patient days, no HAP patients were tested for Legionella, suggesting inconsistencies in testing practices. Environmental testing identified one Legionella-positive cooling tower, but this finding did not explain the community- based outbreak.
Conclusion: The outbreak, initially presumed to be healthcare-associated, was ultimately determined to be community-based. Misclassification of this pseudo-outbreak heightened concern among healthcare and public health stakeholders and delayed appropriate response measures. This study highlights the critical role of the LERA tool in accurately diagnosing LD, guiding antimicrobial stewardship, and enhancing public health surveillance. Importantly, the detection of an infectious disease after hospital admission does not inherently imply hospital acquisition, underscoring the necessity of tools like LERA to account for incubation periods and clarify the origin of infections. Adoption of the LERA tool could prevent misattribution of outbreaks, ensure targeted interventions, and reduce unnecessary hospital resource utilization.
