Keywords: Disease Burden, state space models
We present a state space or hidden markov model of measles disease burden, and estimate parameters of the model from surveillance data at the country level collected by the World Health Organization (WHO). The aim is to predict the un-observed values of the number of individuals in a population who are infected and susceptible each year using the observed information of number of cases reported. From this we hope to provide valuable information on the impact of regular and supplemental vaccine campaigns both historically and in the near future. Fitting the model to each of 193 countries will enable the setting of goals for vaccination programs, allocation of resources, and evaluation of program success on a country by country basis. Our method builds off of earlier work by Chen, Fricks and Ferrari (2012). Like this previous work it combines expert knowledge in the dynamics of measles disease burden with surveillance data, but additionally it incorporates age distribution of the population and modifies the model so that parameter estimation and forward projections are tenable.