Fig. 1: County-level estimates of MMR vaccine uptake and spatial clustering among US children under age 5.

County-level estimates of ≥1-dose MMR vaccine coverage among US children under age 5 were generated using a multilevel regression with poststratification (MRP) framework, based on digital surveillance data (ONM) collected between July 2023 and April 2024 (n = 3,109 counties; one modelled estimate per county). a, Modelled vaccine uptake categorized into five risk levels based on estimated vaccination rate: very high risk (<60%), high risk (60–69%), medium risk (70–79%), low risk (80–84%) and lowest risk (≥85%), relative to the herd-immunity threshold for measles. Because these estimates include children under 6 months who are not yet vaccine-eligible, the upper threshold appears lower than the 92–95% benchmark typically cited for herd immunity. b, Results from a spatial clustering analysis using local indicators of spatial association (LISA), which identifies counties with vaccination rates statistically significantly higher or lower than their geographic neighbours (two-sided permutation test, 499 permutations, P < 0.05 after Benjamini–Hochberg correction). LISA cluster labels denote: high–high (counties with high uptake surrounded by high-uptake neighbours), low–low (low uptake surrounded by low-uptake neighbours), high–low (high uptake surrounded by low-uptake neighbours) and low–high (low uptake surrounded by high-uptake neighbours). Statistically significant clusters are highlighted; counties shown in white did not exhibit statistically significant spatial clustering. Figure adapted from TIGER/Line Shapefiles, US Census Bureau (2022).