Validation of VaxEstim, a Bayesian model for rapid estimation of measles-mumps-rubella vaccination coverage at outbreak onset: a cross-sectional study.

Abstract

Accurate estimation of vaccination coverage at the onset of an outbreak is critical for guiding timely public health responses. Conventional methods often rely on delayed or incomplete data that obscure immunity gaps. During the 2022-2023 central Ohio measles outbreak, Columbus Public Health estimated measles-mumps-rubella (<b>MMR</b>) coverage at 80-90% despite incomplete reporting. Martoma et al. developed VaxEstim, a statistical model that generated an early estimate of two-dose MMR coverage in the outbreak-exposed population at 53.0% (95% credible interval [<b>CrI</b>] 21.0-77.0) using limited publicly available case-based data. A subsequent epidemiological investigation by Martoma et al. defined the outbreak-exposed population as children <15 years of age, of Somali descent, residing in Columbus (Ohio, USA), and receiving care within a primary care network (<b>PCN</b>). We conducted a cross-sectional validation study using electronic medical records from the PCN of 133,476 children <15 years of age. This cohort included 9864 children of Somali descent residing in Columbus, who comprised the previously defined outbreak-exposed population. Two-dose MMR coverage was defined as ≥2 valid doses by the outbreak onset date of October 8, 2022. VaxEstim's predicted coverage was compared with observed coverage in this group. Observed two-dose MMR coverage among the outbreak-exposed population was 42.4% (4181 of 9864; 95% CI 41.4-43.4), compared with VaxEstim's early-phase prediction of 53.0% (95% CrI 21.0-77.0). The wide credible interval reflects uncertainty typical of early outbreak phases. Model performance showed a mean absolute error of 0.106 and a mean squared error of 0.0113. This study externally validates the VaxEstim prediction against observed coverage in this outbreak-exposed population. The model accurately predicted substantial underimmunisation, underscoring its potential to guide rapid, targeted public health action. National Institute of General Medical Sciences, National Institutes of Health; National Science Foundation.