Abstract by Brittany Russell
Combining Data Sets with Bayesian Missing Data Models: An Analysis of Rheumatic Heart Disease in Samoa
Rheumatic heart disease (RHD) is a widespread problem in developing countries where prevalence rates are high, severe complications are more common, and children are often affected. The chronic health problems that can result from this disease negatively affect individual well-being and economic productivity. Focused treatment and prevention could both improve the lives of individuals and boost economic activity. However, accurate data about RHD is difficult to obtain in many developing countries, including Samoa. I have general data for a large, representative sample of children in Samoa and detailed information from echocardiograms on a nonrandom subsample. I construct a Bayesian missing data model to combine these data sets and make inference about RHD in the population. This will make it possible to answer questions about the prevalence and severity of the disease in Samoa and inform policy makers about the potential benefits of treatment and prevention programs.