Abstract
BACKGROUND: Understanding the spatial dynamics of oral polio vaccine (OPV) transmission will improve resource targeting. Mexico provides a natural laboratory, as it uses inactivated polio vaccine routinely as well as OPV bi-annually. METHODS: Using geospatial maps, we measured the distance and density of OPV vaccinees' shedding in the areas nearest to unvaccinated households in 3 Mexican villages. Comparison of transmission to unvaccinated households utilized a mixed effects logistic regression with random effects for household and time, adjusted for age, gender, area, and running water. RESULTS: The median distance from an unvaccinated household to its nearest OPV-shedding household was 85 meters (interquartile range, 46-145) and the median number of vaccinees shedding OPV within 200 m was 3 (2-6). Transmission to unvaccinated households occurred by day 1. There was no association (odds ratio [OR] 1.04; 95% credible interval [CrI] 0.92-1.16) between the distance from OPV shedding and the odds of transmission. The number of OPV vaccinees shedding within 200 m came close to a significant association with unvaccinated transmission (OR 0.93; CrI 0.84-1.01), but this was not the case for households 100 or 500 m apart. Results were consistent across the 3 villages. CONCLUSIONS: Geospatial analysis did not predict community transmission from vaccinated to unvaccinated households, because OPV use resulted in rapid, low transmission levels. This finding supports the global cessation of OPV.