Michigan's cyclosporiasis ledger passed ten thousand on July 30: 10,386 probable and confirmed cases, up 309 in a day, with 193 people hospitalized — 33 of them in the past week — and no deaths reported. [1] [2]
The number is large and the increment is slowing. The Detroit News documented the deceleration in daily growth the same day the state crossed the threshold. [3] That pairing — a headline total climbing past a round number while the daily additions taper — is the first of two curves in this outbreak, and it is not the worrying one. The worrying one runs the other direction: hospitalizations steepening against slowing incidence. [2] [3]
The federal ledger, reconciled on the record by Reuters, remains a different instrument measuring a different thing. The CDC counted 6,707 laboratory-confirmed cases nationally plus roughly 11,500 suspected ones as of July 28, across nine states. [1] Michigan's 10,386 is a probable-plus-confirmed count; the CDC's is confirmed-only. The two cannot be summed, and this paper's position on the file — carried from the July 29 ledger that kept Michigan's count and the CDC's on separate clocks — is that movement between incommensurable reports is not a single outbreak number and never becomes one by addition.
Two curves, one outbreak
The analysis layer is where July 30 added genuine information. A Truveta examination of health-record data found July diagnoses running at 11 times the eight-year July average, with emergency-department visit patterns to match. [1] That is an analyst's finding, not a CDC surveillance figure, and it describes magnitude rather than cause. What it does not explain is the divergence between the curves: if incidence is genuinely slowing, why are hospitalizations rising?
Three candidates exist and none is established. The case mix may be aging into severity. Care may be arriving late in a long-incubation illness. Or the system may simply be detecting severe cases it missed earlier, as clinicians learned what to test for. [1] [2] Each has different implications for what next week's ledger should show, and the daily state update cannot adjudicate among them.
What the counting cannot say
The traceback record is equally disciplined. Investigators' work on Taylor Farms shredded iceberg lettuce remains a traceback in progress, not a closed attribution; no second product or supplier has entered the record. [1] Food-safety outrage wants a named culprit; surveillance skepticism wants to discount the counting itself. The July 30 record supports neither impulse fully. The counting is real — 309 new cases in a day is not an artifact — and the culprit is not yet a document. [1]
What the record does support is a distinction both frames flatten. Incidence and severity are different curves, and this outbreak's are moving in opposite directions. A reader who watches only the total sees growth; a reader who watches only the increment sees the end approaching. The hospitalization line says the outbreak's cost is still arriving even as its spread may be slowing — and until the state or the CDC publishes the age mix and onset dates that would explain why, that divergence is the fact of the day. [2] [3]
-- NORA WHITFIELD, Chicago