State and county health officials said on Nov. 2 that a child in Lewis County, N.Y., had died of measles. It was New York’s first measles death in more than 30 years and the seventh death reported by a state health department this year.
The Centers for Disease Control and Prevention lists three. The agency’s measles page, last updated Oct. 30, shows 4,852 confirmed cases in 2026 across 47 jurisdictions. That is more than double the 2,289 cases recorded in all of 2025 and the most since 1991. The page puts total deaths at 3, with an asterisk and a note: “This number is subject to change.”
The gap has a known cause. In September, after Health Secretary Robert F. Kennedy Jr. publicly questioned two deaths in Pennsylvania, the C.D.C. stopped publishing the deaths states report to it. It now counts a measles death only when federal death records list measles as the underlying cause, and those records can arrive weeks or months after the death. The agency doesn’t say which three deaths it has counted or where they happened, so it is not clear whether any of Pennsylvania’s six are among them.
The rule reverses what the same agency did a year ago, when it counted a New Mexico death whose cause was still under investigation. It also gives a number below the disease’s usual toll. The C.D.C.’s own surveillance manual says two to three deaths may occur for every 1,000 reported cases. At this year’s case count, that rate works out to roughly 10 to 15 deaths. The states’ seven falls below that range, and the federal three falls well below the states’.
A death in Lewis County
New York has said little about the child. ABC7 New York reported that the child was an infant and quoted Dr. James V. McDonald, who leads the state health department, saying “this loved one was unvaccinated.” Lewis County is a rural county about 93 miles north of Syracuse, The Associated Press reported. It had logged 15 measles cases by mid-October, the fourth-highest count in the state. New York had 178 cases as of Oct. 24, 170 of them outside New York City. Dr. McDonald said the total was now “well beyond 200 cases already this year and growing.”
Gov. Kathy Hochul announced the death on the evening of Nov. 2. “As a mother and grandmother, I cannot imagine what this family is going through,” she said in a statement, Ars Technica reported. Her post on X put the blame on Washington.
I’m heartbroken to learn that a child in Lewis County has died from measles. My heart goes out to their family and everyone who knew and loved them.
This is New York’s first measles-related death in 30 years, a devastating reminder of what’s at stake when Washington deliberately…
A spokesperson for the Department of Health and Human Services did not respond to that charge in a statement to Ars Technica. “Secretary Kennedy and CDC Director Dr. Erica Schwartz continue to make clear that the MMR vaccine is the most effective way to prevent measles,” the spokesperson said. The death came three days after the C.D.C.’s last weekly update. Under the current rule it can join the federal total only after New York files a death certificate, the National Center for Health Statistics codes it and measles is selected as the underlying cause. New York has been under a state disaster emergency since Oct. 5, which lets paramedics, pharmacists and midwives give the vaccine. That order was the first of its kind in the current surge.
How the count changed
The fight started on the morning of Aug. 25. Pennsylvania’s health department notified the C.D.C. of two measles-associated deaths in Lancaster County, the state’s first in 35 years, and then posted the news online. C.D.C. staff accepted the state’s classification, Reuters reported, citing a Pennsylvania official close to Gov. Josh Shapiro. Later that day Mr. Kennedy called Mr. Shapiro to offer federal help. Mr. Shapiro told him his past statements on vaccines were “having a negative impact on communities across America.”
On Aug. 26, Mr. Kennedy accused the governor on X of withholding information and wrote that the deaths “may even have been altogether fabricated by one of the Governor’s hopeful staffers.”
Governor Shapiro’s unseemly rush to Lancaster County to convene a national press conference and his giddy delight in announcing two measles deaths was a morose commentary on the contemporary political dialogue. The performance was a replay of the Democrats’ COVID playbook:…
The C.D.C. postponed its weekly measles update that Friday. When the page came back on Sunday, the Pennsylvania deaths were gone. In their place was a note saying “available information does not establish whether measles caused or contributed to” them. As of Aug. 31, the federal count for 2026 was zero, PhillyVoice reported. On Sept. 3, citing three people familiar with the matter, Reuters reported that Mr. Kennedy had asked the new C.D.C. director, Dr. Erica Schwartz, to take the deaths off the tally and that she did so without objection. H.H.S. called that account “false and misleading,” according to FactCheck.org. Dr. Georges Benjamin of the American Public Health Association told Reuters it was “unheard of” for the C.D.C. to publicly doubt a state’s findings.
The secretary had no business interfering in the normal process
The new rule came next. A note posted on the C.D.C. site on Sept. 4 moved the count to the National Center for Health Statistics. On Sept. 14, an H.H.S. spokeswoman, Emily Hilliard, told The Guardian that the agency would “report measles deaths if and when the National Center for Health Statistics reports measles as the underlying cause of death,” CIDRAP reported. A death with measles listed only as a contributing condition no longer counts.
The N.C.H.S. system was built to settle the record, not to track an outbreak in real time. Provisional cause-of-death data usually come out three to nine months after a death, depending on the cause, according to an analysis by Josh Michaud and Jennifer Kates of KFF. Denys T. Lau and Jennifer D. Schoendorf, who between them spent about 40 years at the center, wrote in STAT that routine certificates are coded within one to two weeks. Rare or complex causes need manual review, which can take “from weeks to even months in some cases,” they wrote. In their view the method “will not provide the surveillance information needed at the beginning of an active outbreak when counts are low.”
The federal number did begin to move, with no word on where the deaths occurred. One death appeared on the tally around Sept. 24 and a second on Sept. 29. A third followed on Oct. 20, after H.H.S. said the statistics center had verified measles as the underlying cause, The Hill reported. Pennsylvania confirmed two more deaths on Sept. 15 and a fifth on Sept. 30, and a county coroner reported a sixth on Oct. 12. Mr. Shapiro announced the Sept. 15 deaths on X. Through that period he accused Mr. Kennedy of having personally intervened to keep them off the federal count.
Today, the Pennsylvania Department of Health confirmed two more measles-associated deaths after local coroners reported these individuals died from complications associated with measles. That’s four total measles-associated deaths in our Commonwealth and the first Pennsylvania…
Six deaths in Pennsylvania
None of the six had been vaccinated. Two were babies: a newborn with congenital measles, whose mother was severely ill when she gave birth, and a 6-week-old too young for the shot. An 18-year-old from Mifflin County, described as “perfectly healthy,” died of acute disseminated encephalomyelitis, a rare neurological complication. A 40-year-old woman in Jefferson County who had severe asthma and C.O.P.D. died of respiratory failure. A 20-year-old with leukemia also died. The sixth was a 39-year-old woman in Franklin County who died on Oct. 10. The coroner gave her cause of death as “a probable acute cardiac event due to possible complications of the measles,” according to The Philadelphia Inquirer.
The state has confirmed five deaths and is still investigating the Franklin County case. As of Nov. 2 the outbreak stood at 1,433 cases and 269 hospitalizations, Reuters reported. A Franklin County judge sealed the records that identify the woman. The Inquirer has asked the court to vacate that order, arguing that the information is public under Pennsylvania’s Right-to-Know Law, according to the Reporters Committee for Freedom of the Press, which represents the paper.
The outbreak may be slowing. Cases peaked in September, and October’s 238 cases were about half the monthly totals for August and September, The Inquirer reported on Nov. 2. Dr. Debra Bogen, the state’s health secretary, was cautious. “A couple weeks of decline doesn’t mean it’s a pattern,” she said.
The record before 2026
The country has seen this before, and the C.D.C. counted differently then. During the resurgence of 1989 to 1991, more than 55,000 cases and 123 deaths were reported, according to the agency’s surveillance manual. A 1992 report in the agency’s weekly journal gave the toll as “approximately 55,000 cases and 132 suspected measles-associated deaths.” That is the same category Pennsylvania uses now.
The United States declared measles eliminated in 2000. Since then the manual has treated it as an immediately notifiable disease. States are asked to report cases to the C.D.C. within 24 hours, and investigators record deaths in the case file, where “Death certificate diagnoses” is one field among several. Measles has been notifiable since 1912, KFF wrote. The current case definition dates from 2013 and says nothing specific about how to attribute a death.
For years, a single death made national news. In July 2015, Washington State reported the first confirmed U.S. measles death since 2003. She was a Clallam County woman in her 20s whose infection was found only at autopsy, and the autopsy found pneumonia due to measles, The Seattle Times reported. In 2019, 1,274 cases were confirmed, the most since 1992 up to that point. No deaths had been reported to the C.D.C. through Oct. 1 of that year, according to an agency report.
Then came 2025. Two unvaccinated school-age children died in Texas, and so did an unvaccinated adult in New Mexico. On March 6, 2025, New Mexico’s health department said the adult’s official cause of death was still under investigation by the state’s Office of the Medical Investigator, though the state laboratory had confirmed measles. The C.D.C. counted the death. Jess Steier, a public health scientist, and Dr. Elana Pearl Ben-Joseph, a pediatrician, both of the group Unbiased Science, noted that precedent in STAT. “Holding measles deaths out of the national update until certificates are final reverses that order,” they wrote.
The case for a stricter count
Dr. Schwartz took office on Aug. 13 and has denied that politics drove the change. In her first interview as director, with Atlanta News First on Sept. 16, she said the agency had never had a standard definition of a measles death. She compared the question to the pandemic: “If someone got hit by a bus, and they happened to have COVID, often that was called a COVID-related death.” She added, “Definitions matter.”
There is a real dispute behind her point, and it centers on one baby. Stephen Diamantoni, the Lancaster County coroner, attributed the newborn’s Aug. 14 death to a “traumatic laceration of the spleen.” He said his pathologist “did not feel it was related to measles,” FactCheck.org reported, though he acknowledged that measles was listed in the part of the certificate for contributing conditions. Josh Parsons, a Lancaster County commissioner, said the state had “one case where a person died WITH measles, but not FROM measles.” Dr. Amesh Adalja of Johns Hopkins faulted the state for saying so little: “Pennsylvania created its own information vacuum and then seemed surprised when someone else filled it.”
Even physicians who object to how the C.D.C. made the change accept its basic principle. Ms. Steier and Dr. Ben-Joseph called for the government to show the public which deaths states have reported, which are under review and what standard applies. Their verdict on the standard itself:
Underlying cause is the right standard for official mortality statistics, and a clearer surveillance definition would help.
Some readers raised the “with, not from” objection too. On Hacker News, a reader responding to Ars Technica’s list of the Pennsylvania dead asked whether most of them had died of measles at all.
So it wasn’t actually measles except for one case?
For most of the six, the record says otherwise. Coroners in Jefferson, Mifflin and Franklin counties tied those deaths to measles, and the 18-year-old died of a known complication of the virus. The Lancaster newborn is the one case with a genuine conflict. Even there, the Amish parents told The Atlantic they believed measles contributed, because the mother was severely ill with it and went into labor about a month early. The Covid comparison also fails on testing. Dr. Jake Scott, an infectious disease specialist at Stanford, wrote for CIDRAP that “Nobody screens for measles.” Hospitals tested nearly every patient for the coronavirus, he wrote, so incidental positives were common. A measles test is ordered only when a doctor suspects measles.
A certificate-only count also misses deaths in a way neither side disputes. Dr. Tom Frieden, a former C.D.C. director, wrote on his Substack that measles erases immune memory and leaves children open to other infections for two to three years. When such a child dies of pneumonia a year later, the certificate says pneumonia. He also cited a Vermont review that found major errors in 51 percent of death certificates.
Errors in measles death counts run mostly in one direction: too low.
The alert that did not go out
On Oct. 30, Lena H. Sun of The Washington Post reported that about a month earlier, C.D.C. scientists had tried to send clinicians an urgent alert on measles in pregnancy through the agency’s Health Alert Network. Two people familiar with the matter told The Post that the director’s office, whose approval the alert normally needs, did not approve it. They did not know why. States had asked for the guidance because clinicians were seeing unvaccinated women of childbearing age with measles, many for the first time in their careers.
Ms. Sun set out the sequence in a thread on X.
CDC scientists first tried to send out a notice thru a rapid communication system that goes to clinicians, public health professionals and many others because states were asking for the information. It was not approved. 2/5
The scientists then turned to the agency’s weekly journal, the M.M.W.R., whose review takes longer, and were considering publishing outside the agency, Ars Technica reported. H.H.S. told Ars that the information was “moving through CDC’s scientific review and clearance process, which informs publication decisions and timing.”
Pennsylvania went ahead on its own. On Oct. 19, the day a C.D.C. epidemic intelligence officer began working with the state, its health department issued its own alert on measles in pregnancy and in newborns. The state had recorded 66 measles cases in pregnant women since April, The Hill reported, and it was also seeing its first congenital cases in decades. Three days earlier, Dr. Bogen had written to Dr. Schwartz asking why the C.D.C. had changed its reporting, The Inquirer reported.
By not acknowledging the deaths, the CDC obscures the seriousness of this disease and the toll of measles across the U.S
What comes next
On Sept. 18, the C.D.C. said it was working with the Council of State and Territorial Epidemiologists on a national definition of a measles death, Reuters reported. On Oct. 7, NPR reported that the definition was expected within two to three months. Pennsylvania published its own definition in September, modeled on the ones used for Covid, flu and R.S.V. deaths. It requires clinical evidence, a positive test and death within 30 days of symptoms, and it excludes unrelated causes such as car crashes. A death also qualifies if the certificate lists measles as an underlying or contributing cause.
Dr. Frieden has argued for publishing both numbers. “Agencies should publish both and explain the difference,” he wrote. In Congress, Representative Frank Pallone Jr. of New Jersey, the top Democrat on the Energy and Commerce Committee, wrote to Dr. Schwartz on Sept. 14 with two colleagues. The letter asked for a briefing by Sept. 28 on why the Pennsylvania deaths had been dropped. On Sept. 23, Mr. Pallone said he was still waiting for an answer, and it was not clear whether a briefing ever took place.
Other questions are still open. The Inquirer’s petition over the Franklin County records has not been decided. The pregnancy report is still in clearance. New York’s emergency order runs through Nov. 4 unless Ms. Hochul extends it.
The C.D.C. updates its measles page once a week, using reports received by noon Thursday. On the version posted Oct. 30, the death total was still 3, with an asterisk.