Extreme Precipitation and Legionnaires Disease Hospitalizations in Boston, Massachusetts from 2002-2012

Abstract

BACKGROUND AND AIM: Legionnaires disease (LD) and Pontiac Fever are caused by exposure to Legionella spp. bacteria–environmental contaminants commonly found in water and soil. A milder disease, Pontiac Fever is typically self-limiting, and has a short incubation period of hours to days. The estimated incubation period for severe LD pneumonia is two–fourteen days and is the form of Legionellosis most likely to result in a hospital visit. Previous studies report LD associated with extreme weather conditions: higher humidity and increased precipitation. We conducted a retrospective study of air pollution, extreme precipitation, and LD-coded hospital visits in Boston, Massachusetts. METHODS: We analyzed weather and air pollution data: humidity, temperature, precipitation, ozone, and fine particulates (PM2.5) with LD ICD-9 code 482.84 hospital visits. Variables were assessed as continuous variables, at extremes, in different seasons, and at multiple LD case time lags. Using time-series analysis and negative binomial methods, final multivariable models included extreme precipitation events, temperature, and the occurrence of LD-related hospital visits. RESULTS:A total of 466 hospital visits were identified during 2002-2012. The selected model had the lowest AIC and included an autocorrelation term for LD case frequency from lag 1 to 7 days, mean temperature, and extreme precipitation (90th percentile). Extreme precipitation events were associated after 9 and 11 lag days with IRR of 1.54 (95% CI, 1.1379 – 2.0496) and 1.38 (95% CI, 1.0086 – 1.8610) respectively. We did not find an association with temperature, particulates or ozone, and humidity was excluded because of collinearity with precipitation (R = 0.39). CONCLUSIONS:Hospital visits for LD significantly increased at lag days nine and eleven occurred after extreme precipitation events. We did not find a significant association with air pollution. These findings agree with other reports of LD occurrence after wet weather. This abstract does not reflect EPA policy. KEYWORDS: Climate, Water quality, Infectious diseases