Abstract
Hurricane Harvey made landfall in Southeast Texas in August 2017, causing catastrophic flooding. We examined the association between exposure to Hurricane Harvey and rates of 30-and 90-day mortality among Texas nursing home residents. Using Medicare data, we compared a cohort of residents exposed to the storm in 2017 (n= 18,693) to a cohort of residents not exposed to the storm in 2015 (n=19,688). We fit generalized estimating equations with a sandwich estimator, adjusting for resident demographic and clinical characteristics. Exposure to Hurricane Harvey was not associated with 30-day rates of mortality. However, it was associated with a 15% increase in the odds of dying at 90 days holding resident characteristics constant (adjusted odds ratio: 1.15 [95% Confidence interval: 1.06, 1.25). Health effects due to severe flooding, such as those related to decreased ability to manage chronic conditions, may not be apparent until several weeks after the storm.