Objective: There is a need to develop cost-effective methods to support public health policy makers plan ahead and make robust decisions on protective measures to safeguard against severe impacts of extreme weather events and natural disasters in the future, given competing demands on the social and healthcare resources, large uncertainty associated with extreme events and their impacts, and the opportunity costs associated with making ineffective decisions.
Design: The authors combine a physics-based method known as nonextensive statistical mechanics for modeling the probability distribution of systems or processes exhibiting extreme behavior, with a decision-analytical method known as partitioned multiobjective risk method to determine the optimal decision option when planning for potential extreme events.
Results: The method is illustrated using a simple hypothetical example. It is shown that partitioning the exceedance probability distribution of health impact into three ranges (low severity/high exceedance probability, moderate severity/medium exceedance probability, and high severity/low exceedance probability) leads to the correct estimation of the conditional expected impact in each range. Multiobjective optimization is used to determine the optimal decision option based on the perspective of the policy maker.
Conclusion: This method constitutes a robust generic framework for the quantification of impacts and supporting decision-making under scenarios of extreme and catastrophic health risks.