County-Level Measles Risk Is No Longer Confined to Isolated Clusters
Measles is re-emerging in the United States in a pattern that looks different from the outbreaks seen during the first decade of the past 15 years. Annual county-level data from 2011 through 2025 show a clear epidemiologic shift: early outbreaks were usually limited to single counties and only weakly tied to local vaccine coverage, while more recent events have spread more easily across neighboring counties and have become more tightly linked to immunization levels.
This matters because measles transmission depends on a very high bar for population immunity. Once coverage slips, a virus with this degree of contagiousness does not merely linger in one undervaccinated enclave; it can move outward through adjacent communities and begin to behave like a broader population threat. The new pattern suggests that transmission is increasingly sustained beyond socially clustered groups alone, which changes how public health officials should think about surveillance and intervention.
Spatial Spread Became More Important After 2020
The analysis separated the 2011-2020 period from 2021-2025 and found a sharp change in the relationship between county vaccine coverage and outbreak behavior. During the earlier period, outbreak risk was only weakly associated with county-level vaccine coverage. Outbreaks were also more likely to stay localized. In the later period, lower coverage became a much stronger predictor of outbreak size, and spread to neighboring counties increased.
A one-county event carries a very different operational burden from a multijurisdictional cluster. Once transmission crosses county lines, response demands escalate quickly: contact tracing widens, school and community outreach expands, and localized vaccination gaps become regional vulnerabilities. The study’s spatial framing captures that reality better than a simple county-by-county count of cases.
The biologic logic is straightforward. When vaccine coverage erodes, chains of susceptibility lengthen and the virus encounters fewer immune barriers as it moves through connected populations. Measles does not need a large number of openings to create sustained spread; it needs only enough connected pockets of susceptibility for repeated transmission to occur.
Prediction Model Outperformed Coverage Alone
Investigators built a spatial statistical model to estimate outbreak risk and outbreak size across all US counties, then validated the model using data from 2021 through 2025. The model identified counties with at least 10 cases with a sensitivity of 95%, with a range of 75% to 100% across those validation years, and a specificity of 83%, ranging from 64% to 98%.
Those values are not trivial. In a disease like measles, where early recognition of geographic risk can shape vaccination campaigns, school-based outreach, and rapid containment efforts, a model that performs well at the county level is more useful than a blunt national or state-level signal. The same analysis found that mechanistic models using county vaccine coverage alone performed worse, particularly when the target was larger outbreaks. Coverage matters, but coverage alone did not capture the full spatial structure of risk.
The difference likely reflects how measles actually moves through communities. Immunization rates are essential, yet they do not fully represent mobility, adjacency, social mixing, or the way transmission can intensify once a case enters a particular local network. A spatial statistical model can absorb those patterns more effectively than a threshold based only on coverage.
Targeting the Highest-Risk Counties Produced the Largest Gain
When the model was used to test intervention strategies, the most effective approach was to prioritize the areas with the greatest predicted risk. Targeting the top decile of US counties by model-predicted risk outperformed interventions aimed at the top five states or strategies based only on low vaccine coverage.
A small change in coverage produced an outsized effect in those high-risk counties. Even a 5 percentage point increase in vaccination coverage in the top decile of predicted-risk counties reduced long-term measles risk nationwide by more than 90%.
That finding carries immediate public health weight. It suggests that modest improvements in the right places can bend national risk far more effectively than broader, less focused efforts. In a setting where resources for outreach, catch-up vaccination, and community engagement are finite, the placement of interventions may matter as much as the total number delivered.
Re-emergence Looks More Like a Network Problem Than a Single-Cause Problem
The study points to a deeper shift in measles epidemiology. Early in the period examined, outbreaks were concentrated in socially clustered communities, and county vaccine coverage did not strongly predict the size or distribution of events. More recently, outbreak behavior has begun to reflect broader population susceptibility and greater geographic spread across neighboring counties.
This is the kind of pattern that complicates routine elimination-era assumptions. A jurisdiction with apparently acceptable average coverage may still sit near vulnerable counties, and those geographic relationships can turn a local importation into a wider event. Public health teams that focus only on low-coverage counties may miss adjacent places where risk is building through movement and connectivity rather than coverage alone.
The practical consequence is sharper than a conventional warning about declining vaccination. County-level risk stratification may now be necessary to keep measles from re-establishing itself in the United States at scale. A risk map built on spatial modeling can direct outreach before case counts rise, rather than after a cluster has already widened.
Public Health Response Will Need Finer Geographic Resolution
These results support a shift toward spatially targeted prevention. County-level predictions were able to distinguish higher-risk locations, and those predictions translated into better intervention efficiency than state-level targeting. That suggests local public health departments, health systems, and immunization programs may need tools that work below the state line and account for geographic adjacency.
The study also reinforces a sobering point about vaccine-preventable disease control. Elimination status does not erase susceptibility; it only holds as long as coverage remains high enough and responses remain precise enough to catch emerging gaps. Once coverage declines, the virus can exploit the geography of underimmunization with growing efficiency.
The main limitation is embedded in the modeling approach itself: predictive performance was strong, but prediction is not the same as direct causal proof for every county-level pattern. Even so, the validation results through 2021-2025, combined with the stronger performance of spatial modeling over coverage-only approaches, make the central message hard to ignore. Measles re-emergence in the United States is increasingly a spatial problem, and the most effective interventions appear to be the ones aimed where the model says risk is highest.