Social and spatial variation in heatwave-related emergency department visits in Australia

Abstract

BACKGROUND AND AIM: Heatwaves are extreme weather events that cause more urgent medical care than any other hazard in Australia. Heatwave impacts vary between populations and locations and empirical evidence concerning vulnerable places and people at a local level is critical for effective policy, planning, and interventions, especially in a warming climate. This study aimed to examine heatwave-related morbidity in Australia (using emergency department (ED) presentations) at a small-area level and identify susceptibility by neighbourhood characteristics. METHODS: ED presentations data for eight Australian cities from 2013-2018 during the warm-season (October- March) were extracted from a national database. In each city, we examined the association of heatwave severity (defined using Excess Heat Factor) with ED visits at the Statistical Areas 2 geographical unit using a time-stratified case-crossover approach. Effect modification by neighbourhood characteristics were examined and city-level estimates were pooled using random-effects meta-analyses. RESULTS:A total of 23,233,374 all-cause ED presentations were included in the analysis. Pooled ED relative risk estimates indicate that ED visits increased by 2.9% (95%CI: 1.5-4.2%) and 3.4% (95%CI: 1.2-5.5%) during low-intensity and severe/extreme heatwaves, respectively. There were 67,974 excess ED presentations during heatwaves with 82% of these occurring during low-intensity events which are more frequent than severe/extreme heatwaves. The effects varied substantially by city with the highest number of excess ED presentations observed in Adelaide (28,198) and Sydney (19,899). Heat-related conditions such as dehydration and heat exhaustion/stroke were significantly greater during heatwaves and the effect increased from low-intensity to severe/extreme heatwaves. Analysis of neighbourhood characteristics on heatwave-related ED presentations revealed greater health impacts for areas with high social vulnerability e.g., household composition, language and culture, and housing quality. CONCLUSIONS:Overall, ED presentations are sensitive to heatwaves across all severity classes and the effects are stronger for individuals with certain health conditions and neighbourhood contexts. KEYWORDS: Heatwaves, integrated data, vulnerability, excess heat factor