Exploration of the health risk-based definition for heatwave: A multi-city study

Background: As heatwaves are expected to be more frequent, longer, and more intense in the future, it is imperative to understand how heatwaves affect health. However, it is intensely debated about how a heatwave should be defined. Objectives: This study explored the possibility of developing a health risk-based definition for heatwave, and assessed the heat-related mortality in the three largest Australian cities. Methods: Daily data on climatic variables and non-accidental deaths for Brisbane, Melbourne and Sydney during the period 1988-2009 were obtained from relevant government agencies. Several local heatwave definitions were tested by using percentiles (e.g., from the 75th to 99th centile) of mean temperature with duration >= 2 days across these cities. We examined the relative risks of mortality associated with heatwaves in each city using Poisson generalised additive model, after controlling for long-term trend, within-season variation, day of the week, and relative humidity. Then, Bayesian hierarchical model with segment-spline was used to examine the threshold for the heatwave-related impacts. Results: A consistent and significant increase in mortality during heatwaves was observed in all three cities. The pooled data show that the relative risk of mortality started to increase around the 95th centile of temperature, increased sharply at the 97th centile and rose alarmingly at the 99th centile. Based on research findings, we proposed tiered health risk-based metrics to define a heatwave. Conclusions: Our findings provide supportive evidence for developing health risk-based metrics to assess the impacts of heatwave. These findings may have important implications for assessing and reducing the burden of heat-related mortality. Crown Copyright (C) 2015 Published by Elsevier Inc. All rights reserved.