Why this paper matters
The United States declared measles eliminated in 2000. Elimination does not mean zero cases. It means measles has not spread continuously inside the country for 12 months or longer. New cases are often brought in by travelers and then spread in undervaccinated communities.
That margin has been under pressure. Kindergarten MMR vaccination coverage fell from 95.2% in the 2019 to 2020 school year to 92.7% in 2023 to 2024. Cases rose from 59 in 2023 to 285 in 2024. CDC has now confirmed 4,410 cases in 2026 through October 15, compared with about 2,290 in all of 2025. According to press reports, a commission of the Pan American Health Organization will review the US elimination status in November.
This week's paper was submitted in March 2025, before most of those cases were counted. It asks how close the United States was to the point where measles can spread on its own, and how much vaccination would have to slip to cross it.
Terms used in this review
- —Measles elimination: No continuous spread of measles inside a country for 12 months or longer. Cases can still occur, usually brought in by travelers.
- —Reproduction number (Re): The average number of people one infected person goes on to infect in the current population. If Re is below 1, outbreaks tend to die out. If it is above 1, an outbreak can keep growing.
- —Importation: A case that is brought into a state or country from somewhere else.
- —Outbreak: Three or more related cases, which is the definition used in the paper.
- —MMR vaccine: The vaccine that protects against measles, mumps, and rubella. It is given in two doses.
- —Model: A computer simulation that uses known data and stated assumptions to estimate what could happen.
- —Scenario: A "what if" setting in a model. A scenario is not a prediction.
What they did
Researchers at Columbia University's Mailman School of Public Health built a model with two parts. The first estimated how often measles is brought into each state, using global measles levels and data on air travel and travel between states. The second estimated how easily measles would spread once it arrived, using each state's vaccination coverage and how people of different ages mix.
They ran 1 million simulated outbreaks. They assumed one case in a fully unprotected population would infect 6 to 10 people, lower than the older estimate of 12 to 18, and that one dose of MMR is 93% effective and two doses are 97% effective. For the future, they asked what would happen to Re through 2030 if vaccination coverage stayed the same or dropped by set amounts.
What they found
Risk was higher in winter and spring, when measles is more common abroad and more infected travelers arrive. Risk was highest in states with large international airports, including Florida, California, New York, Texas, New Jersey, Illinois, Georgia, and Virginia, and moderate in neighboring states. The modeled pattern matched where cases were actually reported from January 2024 to April 2025, though only moderately (rank correlation 0.41).
Kindergarten MMR coverage in 2023 to 2024 ranged across states from 76.7% to 96.6% for the first dose and from 79.6% to 98.3% for the second. At those levels, the average Re was 0.67 (95% CI 0.49 to 0.88). If coverage stayed the same through 2030, Re stayed near that level.
If vaccination fell, Re rose over time. With a 10% decline, Re would exceed 1 in at least 5% of states within five years. With a 25% decline, it would exceed 1 in more than half of states around 2028. With a 50% decline, it would happen within two years.
What the numbers actually mean
An Re of 0.67 means each case infects about two thirds of one other person on average, so most chains of spread end on their own. That is how the United States kept measles eliminated even with cases arriving every year. It also shows how thin the margin is. The gap between 0.67 and 1 is the difference between occasional outbreaks and sustained spread, and vaccination coverage is what holds it open.
An average across states can hide the places that matter most. The authors say their risk estimates are probably conservative, because state data leave out homeschooled children and cannot show differences between communities, which could understate the risk in undervaccinated communities. CDC reported in April 2025 that 82% of that year's early cases were tied to one outbreak in close-knit communities with low vaccination coverage in New Mexico, Oklahoma, and Texas. This paper did not model those communities, so I read that pattern as consistent with the authors' concern and not as a test of the model.
The 2026 counts are not a test of the model either. The paper estimated risk and did not forecast case totals, and its scenarios describe what could happen if coverage fell by set amounts. They do not say coverage has fallen by those amounts. What the 2026 numbers do show is the outcome the paper was written to prevent. CDC has confirmed 4,410 cases through October 15, more than 15 times the 285 cases in 2024. About 96% of 2026 cases were in people who were unvaccinated or whose status is unknown, and 42 outbreaks account for about 96% of cases as of October 8.
The elimination question is now on the calendar. According to press reports, a Pan American Health Organization commission will review whether measles has re-established endemic transmission in the United States, and the meeting planned for April was moved to November. This review cannot say what the commission will decide.
Limitations worth knowing
- —The paper is a model. Its results depend on assumptions, including the reproduction number range, vaccine effectiveness, and contact patterns taken from the United Kingdom because US data were not available.
- —Global measles data are underreported, state vaccination data leave out homeschooled children and differences between communities, and the model used only direct flights and 2018 to 2019 travel patterns. The authors say the estimates are conservative.
- —It is a short brief report, submitted in March 2025 and published in June 2025, so it does not include the 2025 and 2026 case totals.
- —The 2026 figures come from CDC and news reports and are preliminary. CDC counts and the Johns Hopkins tracker differ slightly. Reports of 2026 measles deaths differ by source, and CDC says it is developing a standard definition, so this review does not report them.
The bottom line
The paper's main point is that measles elimination depends on a margin, and vaccination coverage keeps that margin open. At 2023 to 2024 coverage the average Re was below 1, and the authors showed how it could cross 1 if coverage slipped, first in some states and then in most. The 2026 case count is the highest since elimination was declared, and a Pan American Health Organization review is expected in November. The model's message does not depend on how that review turns out. Where coverage is lowest, the margin is thinnest.
Paper reviewed
Bidari S, Yuan H, Yang W. "Assessing the Transmissibility and Outbreak Risk of Measles in the United States, 2024-2030." The Journal of Infectious Diseases. 2026;233(1):132-136. Published online 2025 Jun 25. doi:10.1093/infdis/jiaf342. Available free full text at: https://pmc.ncbi.nlm.nih.gov/articles/PMC12811863/
Additional literature and sources referenced
CDC. "Measles Cases and Outbreaks." Updated October 16, 2026: https://www.cdc.gov/measles/data-research/index.html
USAFacts. "How many measles cases are there in the US?": https://usafacts.org/answers/how-many-measles-cases-are-there-in-the-us/country/united-states/
Contagion Live. "Measles Update: October 16, 2026": https://www.contagionlive.com/view/measles-update-october-16-2026
Fast Company. "The U.S. eliminated measles 26 years ago. Now it could lose that status in a matter of weeks." October 16, 2026: https://www.fastcompany.com/91624440/the-u-s-eliminated-measles-26-years-ago-now-it-could-lose-that-status-in-a-matter-of-weeks
CDC. "Measles Update, United States, January 1 to April 17, 2025." MMWR. Published April 24, 2025: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12021071/
All 2026 figures reflect CDC data as of October 15, 2026, and are preliminary.