If you want to know how many Americans measles has killed this year, the answer depends on whom you ask. The Centers for Disease Control and Prevention (CDC) counts one, with an asterisk. Pennsylvania, currently the centre of the country’s worst measles outbreak in a generation, counts four of its own. The gap between those numbers says much about the state of America’s public-health defences.
The outbreak numbers are sobering enough without the arithmetic dispute. The CDC’s latest update, reported by the Center for Infectious Disease Research and Policy (CIDRAP) at the University of Minnesota, puts confirmed cases at 3,659 so far this year, with 171 added in the past week. That brings the total since the start of 2025—when an outbreak in West Texas set the current wave in motion—to 5,948. By comparison, all of 2024 produced 285 confirmed cases, and the five years from 2020 to 2024 produced just 527 between them.
Pennsylvania is the hot spot. Its health department reported 890 cases for the year on September 26th, 129 of them in the previous seven days, spread, by an earlier count, across 39 counties. Lancaster County, where two of the state’s four measles deaths occurred, has been hardest hit, with 342 cases. Officials last week announced two further deaths, of a 40-year-old woman and an 18-year-old. Three Americans died of measles last year; before 2025, the country’s last measles-associated death was in 2015.
The CDC has yet to acknowledge the Pennsylvania deaths in its tally. Its solitary entry carries a footnote: “This number is subject to change; CDC will update its reporting as additional information becomes available and relevant reviews are completed.” The agency’s director, Erica Schwartz, is reportedly working with the Council of State and Territorial Epidemiologists to produce a standardised definition of a measles death. Until that arrives, the official count will understate what state health departments are already reporting.
The deaths also hint that the case count itself is too low. On CIDRAP’s Osterholm Update podcast this week, Michael Osterholm, the centre’s director, argued from historical experience that measles kills on average about one person per 1,000 cases, with the very young and teenagers and older patients at higher risk than typical school-age children. Measured against Pennsylvania’s four deaths, he suggested, the true case count could be as high as 4,000—“could be half that”, he cautioned—because infections in families and across age groups are going unrecorded. It is very likely, he said, that a number of community cases are unreported.
The official statistics already point to the cause. Of the 3,659 confirmed cases, 95% of patients were unvaccinated or of unknown vaccination status; 9% have been hospitalised. Eighteen per cent are children under five, 43% are aged five to 19 and 38% are adults of 20 or older. The CDC counts 40 outbreaks—defined as three or more linked cases—this year, and says 95% of confirmed cases trace back to outbreaks that began this year or in 2025.
Mr Osterholm tipped his hat to Pennsylvania’s health department, which he has consulted, saying it was handling the outbreak “as well as any state could” and has substantially raised vaccination rates in the affected counties over the past month. He was less sanguine about the national picture, warning that as winter lengthens the virus’s transmission season, similar scenarios will play out “state by state by state”. Ohio added nine cases this week to reach 189 for the year; Wisconsin added four to reach 171; Kentucky added ten to reach 67. All are dwarfed by Pennsylvania.
All of which sits awkwardly beside the claim, repeatedly made by Robert F. Kennedy Jr, the health secretary, that America is controlling measles better than any other country in the world. A government that cannot yet say how many Americans measles has killed, in a state where the true case count may be several times the official one, is poorly placed to grade its own performance. The standardised definition now being drafted will be a start; the asterisk, for the moment, is doing more honest work than the number beside it.