Summary
An analysis of US county data from 2011 to 2025 found that measles outbreaks have become more closely associated with local vaccination coverage and have spread more often across county borders. A spatial model identified counties at higher risk, and projected that a five-percentage-point coverage increase in the highest-risk counties could reduce long-term national measles risk by more than 90%.
A preprint analysing annual measles cases and vaccine coverage across US counties from 2011 to 2025 finds that outbreaks have become more closely tied to county vaccination levels and more likely to spread across county borders. The researchers also developed a spatial model to identify places at risk of larger outbreaks and estimated that targeted increases in vaccination coverage could sharply lower long-term national risk.
The analysis describes a change over time. From 2011 to 2020, county vaccination coverage was only weakly associated with outbreak size, and outbreaks were usually confined to one county. During 2021–2025, the association between coverage and outbreak size became stronger, while spread into neighbouring counties increased. The authors interpret this pattern as evidence that measles transmission is increasingly reaching beyond socially clustered communities.
How well the model identified outbreak risk
The researchers used a spatial statistical model to estimate outbreak risk and size for counties across the United States. In validation covering 2021–2025, it identified counties with at least 10 cases with 95% sensitivity, with yearly values ranging from 75% to 100%. Its specificity was 83%, ranging from 64% to 98%. Sensitivity describes how often counties meeting the case threshold were identified; specificity describes how often counties below it were correctly classified.
Models based on county vaccination coverage alone performed less well, particularly when predicting larger outbreaks. This comparison suggests that coverage levels by themselves may be less useful for anticipating where large outbreaks will occur than a model that also represents spatial patterns.
Targeting vaccination increases
The researchers then used their model to compare intervention strategies. Prioritising counties in the highest-risk group—such as the top 10% nationally—was projected to have a greater impact than focusing on the five states with the highest risk or selecting areas solely on the basis of low vaccine coverage.
In the model, a five-percentage-point increase in vaccination coverage in counties in the top risk decile reduced long-term measles risk nationwide by more than 90%. This is a projection from a modelled scenario, not an observed result from a vaccination campaign. It points to a possible public-health strategy: use outbreak-risk estimates to concentrate efforts in places where local conditions and connections between counties make larger outbreaks more likely.
The study is a preprint, so its findings have not yet been peer reviewed. Its analysis is also county-level: the reported relationships describe patterns between counties, not individual vaccination status or personal risk. The authors report that the study was approved by Stanford University’s Institutional Review Board.