Analysis of heat wave effects on health by using generalized additive model and bootstrap-based model selection

It is known that high summer temperature may lead to worsening health conditions among fragile individuals within exposed populations. It is also argued that multiday patterns of high temperature-heat waves-may have relevant effects on health. We discuss the possible measures of intensities of heat waves to be included in a generalized additive model explaining the number of hospital admissions that occurred during summer months in Milan. The issue of variable selection is central to the analysis: a computational method is discussed which may help in assessing the robustness of the model selection method. Eventually, we obtain evidence supporting the relevance of heat waves in driving adverse health episodes. Copyright (c) 2008 Royal Statistical Society.