County Vaccination Data Masks Measles Risk

Pockets of low vaccination rates in some schools help explain why measles outbreaks have been climbing in the U.S., even when county-level vaccination numbers – the data policymakers consult – appear high enough to provide community protections.

A new study, published Aug. 18 in Nature Medicine, helps reveal why measles, which was declared eliminated in the U.S. in 2000, increased to more than 3,500 cases in 2025-26.

The rise began prior to the COVID-19 pandemic but was made worse by it, largely due to vaccine hesitancy, a surge in vaccine misinformation on social media and vaccine exemption policies that vary across states.

In the study, the researchers collected publicly available county, school district and school-level data of measles, mumps and rubella (MMR) vaccinations and ran it through a mathematical model that translated vaccination coverage into transmission risk.

They found that a new case of measles could spread at a school that had children who were unvaccinated or had missed one of the two recommended MMR doses. These risks, however, appearedlargely invisible when looking at statewide and county-level MMR coverage, because in aggregate, vaccination rates can remain near the 95% herd immunity threshold known to protect a population from wider outbreaks.

The data revealed that the average percentage of unprotected students across the country roughly doubled from 5% to nearly 10% following the pandemic. The study highlights the need for more detailed data to understand the true risks posed by measles.

“Before this decrease of vaccination, even with an introduction of measles, because vaccination coverage was so high, you could still have one or two cases and the outbreak would fizzle out,” said Ana Bento, assistant professor in the Department of Public and Ecosystem Health in the College of Veterinary Medicine, and the paper’s senior author. Siyu Chen, a postdoctoral research associate in Bento’s lab, is the paper’s first author.

“Now we’re in a situation where, given the very heterogenous and low vaccine coverage in places, there is a possibility that an outbreak will become a big, localized outbreak after an introduction, which is what we see,” Bento said.

Outbreaks in 2025 and 2026 in Texas, Arizona, New Mexico and South Carolina revealed the same epidemiological pattern, where MMR vaccination rates at the county level looked reassuring, while school-level clusters of unvaccinated children had silently pushed local transmission risks above that epidemic threshold.

In the study, Chen and Bento collected data and created a nationwide database of kindergartenvaccination records that spanned 50,000 schools, 13,000 school districts, 3,000 counties and 45 states plus the District of Columbia, dating back to 2013.

“There was no single national database of school-level vaccination coverage,” Chen said. “We had to build one, state by state, and reconcile more than a decade of records that were never designed to be compared. That groundwork is what let us finally see risk at the scale measles operates on.”

“We can see the risks now,” Bento said. “We just weren’t watching at the right scale and the window to fix that is closing.”

/Public Release. View in full here.