Abstract
Using the example of Ebola, this article highlights the problems of using population level data to estimate case fatality rates in epidemics