Combining data from multiple sources improves reliability of county-level health prevalence

Is there a method for the state health departments to leverage the data collected using the U.S. Centers for Disease Control and Prevention’s Behavioral Risk Factor Surveillance System (BRFSS) to plan county-level interventions, allocate scarce resources, and provide information to the general public? New research on this topic is published in Statistical Methods & Applications, “Hierarchical Bayes small area estimation for county-level health prevalence to having a personal doctor.” Westat’s Andreea Erciulescu, PhD, and Tom Krenzke, MS, are among the coauthors, collaborating with colleagues Jianzhu Li of the Financial Industry Regulatory Authority and Machell Town of the Centers for Disease Control and Prevention.

Researchers developed a hierarchical Bayes small area estimation model to predict the prevalence of having a personal doctor for all the counties in the U.S., including those where BRFSS survey data were not available. The model combines BRFSS survey data with data from auxiliary sources, while accounting for various sources of error and nested geographical levels. The estimation method we developed unlocks the essential ability to make the most productive use of the BRFSS data when improving county-level health policies.