Peter Heudtlass, Niko Speybroeck, Debarati Guha-Sapir
Institut de Recherche Santé et Société (IRSS), Université catholique de Louvain, Brussels, Belgium
Summary Points
• Forced migrants are an extremely vulnerable population, but reliable data on all-cause mortality are chronically scarce.
• Almost all available estimates are based on data that are collected by humanitarian organizations and often suffer a lack of precision and potential publication bias.
• With two examples from South Sudan and Iraq, we demonstrate how these shortcomings can make audiences jump to conclusions too quickly, based on too little data. We propose an approach borrowed from clinical research to mitigate this problem.
• There are ongoing efforts by humanitarian organizations to improve the quality of data.
Methodological innovations like this could complement these by providing tools to better deal with small samples of data coming from multiple agencies.