Spatiotemporal analysis of heat-related emergency department visits and hospitalizations in Florida (2005–2021): A Bayesian change point detection approach

Abstract

The rise in global temperatures poses significant health risks, underscoring the need for decision-makers to understand the spatiotemporal patterns of heat-related health issues as part of efforts to advance sustainability in cities. However, many existing studies rely on statistical methods and limited time frames, failing to capture full temporal dynamics and trend changepoints. This study addresses these gaps by proposing a novel method using the Bayesian Estimator of Abrupt Change, Seasonality, and Trend (BEAST) algorithm to analyze heat-related emergency department visits (EDVs) and hospitalizations across 67 Florida counties from 2005 to 2021. The method decomposes time series data into trend and seasonality components, with significant trend changes analyzed to determine associations with demographic, health, environmental, and meteorological variables. Results indicated relatively stable or slightly increasing trends in heat-related EDVs and hospitalizations in most counties. Among the 14 analyzed variables, pre-existing health conditions, populations aged 5 years or younger, and those aged 65 years or older were most strongly linked to trend changes, while environmental and meteorological factors played a lesser role. These findings contribute to the understanding of heat-related health vulnerabilities, offering valuable insights for fostering resilience and adaptability in sustainable cities.