Bayesian spatial modeling of measles outbreak in Texas, USA

Abstract Measles remains a significant public health concern despite the availability of an effective vaccine, with recent outbreaks in Texas illustrating ongoing vulnerabilities in disease control. This study investigates the spatial distribution of measles incidence during the 2025 outbreak period across Texas counties using a Bayesian Zero-Inflated Poisson model with spatial dependence, incorporating a conditional autoregressive prior. The model accounts for the excess zeros in the data, i.e., counties reporting no cases, while capturing spatial correlations and the effects of demographic and environmental covariates, including MMR vaccination rates, population density, temperature, and precipitation. By applying this modeling framework, we estimate key epidemiological metrics such as the standardized incidence ratio, relative risk, and exceedance probabilities to identify high-risk regions. Results highlight a pronounced spatial concentration of elevated measles risk in northwestern Texas, where low vaccination coverage and other contributing factors have led to a severe outbreak.