Three books go into Sunday, and Sunday merges none of them. Kennedy's book says one: the secretary acknowledged the coroner's measles cause for one infant while the national data prints zero measles deaths for 2026. [1] Pennsylvania's book says two: the state confirms two measles-associated deaths, and the Lancaster coroner split them Friday — measles in the 6-week-old girl, who also had Amish lethal microcephaly, and a ruptured spleen, not measles, in the August 14 newborn. [1] The CDC's public book says none: Friday's update still excludes the Pennsylvania-reported deaths. [3] Saturday's file printed one forensic cause plus a still-zero national box. Sunday's files document how the zero got made.

Reuters' exclusive, carried Wednesday by WMBD, names the mechanism: Kennedy asked his new CDC director to remove references to the two Pennsylvania deaths from the agency's online measles tally after questioning whether the disease contributed — and this after CDC staffers had already accepted the state's characterization of the deaths as measles-linked, a Pennsylvania official said. [2] The recently appointed director, Erica Schwartz, agreed with the directive and implemented it without objection, three sources said. [2] Accepted-by-staff on one day, removed-by-directive the next. That sequence is the story whether or not one calls it interference — and public-health experts told Reuters it amounts to extraordinary political interference aimed at minimizing the virus's risks. [2]

The coroner's Friday split is the forensic anchor both books misuse. The girl died of measles; she also carried a lethal genetic condition. [1] The boy died of a ruptured spleen with measles on the certificate but not as the cause. [1] Two infants, two findings, zero ambiguity about which is which — and three official counts anyway, because "associated," "caused," and "recorded by NCHS" are three different tests being graded as one argument. Do not collapse the spleen into a measles cause. Do not collapse Kennedy's one into Pennsylvania's two. Do not treat an NCHS-records test as a finding that deaths were invented.

The political file runs alongside without merging anything. Shapiro announced two "measles-associated" deaths last week and accused Kennedy of sowing doubt, playing politics, directing non-acknowledgment, and crying fabrication. [1] Kennedy answered that emerging facts vindicated waiting for evidence over grandstanding, and that Shapiro politicized the deaths before causes were established. [2] That exchange is real and it is not epidemiology. Meanwhile the outbreak that produced the argument keeps its own count: 540 Pennsylvania cases this year, declining vaccine rates, medical groups citing rising hesitancy. [2] The fight is about two deaths. The denominator is five hundred forty cases and growing.

Friday's CDC update is the operating record until replaced: no Pennsylvania deaths in the national box. [3] The coroner's office, asked about Kennedy's remarks, did not immediately respond in the Friday reporting. [1] The state health spokesman reiterated two and declined to engage the secretary's remarks. [1] Everyone has spoken. Nothing has merged. A reader of only the secretary's post misses the state's two. A reader of only the state's two misses the federal zero. A reader of only the zero misses the removal ask that made it. Keep all three books open. Keep all three books open. That is the whole of the Sunday discipline, and it fits in one line because the alternative — picking a ledger — is how the miscounts started.

Saturday's file caught the secretary citing the coroner directly — the Lancaster finding that measles caused only the second infant's death, with the child carrying Amish lethal microcephaly and living about two months against a typical six — while the state's table still printed two and the CDC's site swapped its asterisk language for an NCHS-records test. That post is the one the Sunday ledgers inherit: acknowledged cause plus zero box, side by side, unreconciled. Nothing since has moved any of the three numbers. The next edition should check whether NCHS finishes its review, whether the state revises its table, and whether the 540-case denominator [2] — the outbreak behind the argument — forces a fourth book nobody has opened yet.

-- NORA WHITFIELD, Chicago