Dr. Stephen Diamantoni is the coroner in Lancaster County, Pennsylvania. His office certifies how residents die, and he found out about his county’s two measles deaths the way the rest of us did, from the news. He told reporters his office had examined a newborn who tested positive for measles, and that he had recorded the cause of death as a ruptured spleen and massive blood loss, not measles. He had no record of a measles death anywhere in the county, he said, and had not been told about the second fatality the state was describing at all.
The state was describing quite a lot. On August 25 the Pennsylvania Department of Health announced that two unvaccinated Lancaster County residents had died in what it classified as measles-associated cases, the state’s first measles deaths in 35 years. Governor Josh Shapiro delivered it from a podium. The wires picked it up. It was the headline the whole outbreak had been building toward, a vaccine-preventable disease returning and turning lethal, and the official who signs death certificates in the county at the center of it had certified no such thing.
We have argued about this exact question before, over a different virus, and it cost us. For most of 2020 through 2022 the public-health apparatus treated a positive test near the time of death as close enough to assign the death to COVID, and the fight over whether people had died of the virus or merely with it never got an honest resolution. The tallies swelled. The public noticed. Confidence in official death counts has not come back since. The lesson that should have survived is plain enough: how you count a death is a scientific judgment, and once the count starts serving the argument instead of the other way around, people stop believing it.
Start with why measles-associated is a real category and not just a dodge. Measles rarely kills on its own. It kills by opening the door to something else: pneumonia, which is its leading cause of death in children, encephalitis in roughly one case in a thousand, and a stretch of immune amnesia in which the virus erases part of the body’s existing immunity and leaves a child exposed to other infections for months. A death recorded as pneumonia can still be, in every sense that matters medically, a measles death. Surveillance definitions are written broad on purpose, because the virus does its worst work through proxies.
A broad definition is a tool, though, and a tool can be pointed at a headline. There is a wide gap between a child who died of measles pneumonia and a child who tested positive for measles and then bled out from a ruptured spleen. If the spleen was enlarged by an active infection and gave way, that is a measles death. If it ruptured from trauma and merely carried a positive test, that is a different event wearing the same label. Which of those happened decides how this case should read, and the state’s chosen word, measles-associated, is built to keep you from having to ask.
The CDC, under its new director, declined to add the two deaths to the national tally, saying that available information “does not establish whether measles caused or contributed to the deaths or whether the individuals died from other causes while infected with measles.” As of this week the federal count still reads 0 measles deaths for 2026, even as Pennsylvania books 2.
The mainstream read arrived within hours, and it was uniform. CNN reported that Trump’s new CDC director had ordered a delay in counting the deaths. The Philadelphia Inquirer framed it as the agency leaving the deaths off the count after Health Secretary Robert F. Kennedy Jr. sparred with Shapiro. The implied story is not hard to read: a politicized, vaccine-skeptical administration disappearing inconvenient bodies to make an outbreak look tamer than it is.
Read the CDC’s actual sentence again. It asks whether the disease caused or contributed to the death. That is not censorship. It is the question every death certificate exists to answer, and it is the same question the apparatus waved off for three years when the incentive pointed the other way. When a positive test plus a death equaled a COVID death, these same institutions called it caution. Asking for evidence of causation before you hang a death on a disease is not a partisan trick. It is the floor.
The vacuum where the evidence should be
None of that clears Harrisburg. The failure in Lancaster County is not the CDC asking for an autopsy. It is a state that announced two deaths and then went quiet on the evidence for them. Vinay Prasad, a serving FDA official and no defender of the establishment, walked through the infant case and found that silence where the proof should be.
His account is specific. He says the medical examiner found no enlarged spleen at autopsy and noted exterior trauma marks on the torso. He says it has never been made clear whether the newborn was PCR-positive for an active infection or simply carried its mother’s measles antibodies, the distinction that separates a measles death from a coincidence. And the second death, he says, comes with no public information about age or underlying conditions at all. His verdict on the governor is flat: Shapiro held a press conference and then abdicated the one thing that would settle it, the release of the record. A data vacuum in the middle of a live outbreak does not reassure anyone. It is the most dependable engine of the vaccine hesitancy officials keep saying they want to fight.
The pressure on both sides is easy to understand. Pennsylvania has logged 497 measles cases in 2026, 227 of them in Lancaster County, against a national total of roughly 2,318, the worst year for measles in the United States in 35 years. Two deaths in that setting are not a footnote. They are the emotional core of the argument for aggressive MMR uptake, which is precisely why each side wants them counted its way, and precisely why they have to be counted correctly. A death pinned on measles to win a debate is worth nothing to the next parent deciding what to believe, because sooner or later the coroner talks. This one already has.
What ends this is not a press release or a viral post from the health secretary. It is paperwork: the full autopsy reports, the lab work that separates active infection from inherited antibody, and the certifier’s own cause-of-death findings for both cases. Pennsylvania has not released that record, and the CDC says its review is waiting on that kind of documentation. Whichever office publishes it first, Harrisburg or Atlanta, will have done the only honest thing on the table. Until one of them does, two measles deaths and zero measles deaths are both just assertions with a podium behind them.
Sources
- PhillyVoice – CDC leaves Pennsylvania measles-associated deaths off national count (2026)
- Al Jazeera – US health authorities dispute Pennsylvania’s measles-related deaths (2026)
- Vinay Prasad – “The CDC and Pennsylvania won’t agree if 2 people died of measles” (2026)
- The Philadelphia Inquirer – CDC left two measles-associated Pa. deaths out of its latest count (2026)
- CNN – Trump’s new CDC director ordered delay in counting 2 measles-related deaths (2026)
- CDC – Measles Symptoms and Complications
- BMJ – “Measles: CDC refuses to record Pennsylvania fatalities in ‘measles associated deaths’ row” (2026)