Global Health Estimates: Modelling and Predicting Health Outcomes

Mathers, Hogan and Stevens describe statistical and mathematical modelling to produce estimates of health indicators that are comparable across populations and/or time. They explain the reasons for choosing to calculate health estimates and review the major classes of models in use. The authors discuss issues around evaluating, communicating and using estimates, drawing examples from work at the global level to model health indicators across countries, regions and time. They discuss reasons for inconsistencies between statistical estimates published by different groups and the issues that this raises for users of health indicators. Finally, the authors examine the relevance of estimates for countries and whether the estimates help or hinder countries to understand the limitations of empirical data and work to improve the availability of quality data.