The effectiveness of heat alerts issued by national weather services in preventing heat-related mortality in Europe

Recent research has elucidated that the implementation of heat-health prevention plans has been effective in reducing heat-related mortality in Europe. However, it remains largely unresolved whether heat alerts issued based on weather forecasts, which constitute a core element of these plans, independently contribute to the observed mortality reductions. In this contribution, we will present preliminary results from a study based on daily heat alert data and all-cause mortality series from major cities in several European countries, including Germany, Poland, and Portugal. We used random forest classification to identify days, which would have likely experienced a heat alert in the time period before the implementation of the early warning system. Subsequently, we applied time-series regression methods, including distributed lag non-linear models, combined with a difference-in-difference approach to assess whether actual heat alerts are associated with a reduction in all-cause mortality. City-specific results were pooled using mixed-effect meta-regression techniques. Based on a reduced dataset analysed so far, we found strong geographic heterogeneity, with evidence for a protective effect of heat alerts seen only in part of the cities. The pooled relative risks were close to 1, suggesting that the implementation of heat alerts alone cannot explain the mortality reductions associated with the introduction of more comprehensive heat-health prevention plans. Future research will need to investigate other elements of heat-health prevention plans to identify the most effective preventive measures as Europe faces further escalating heat due to climate change.