Cyclospora outbreak triples to 6,358 cases in 15 states after the CDC widened what counts as a case

The Centers for Disease Control and Prevention on Wednesday put the cyclospora outbreak linked to recalled iceberg lettuce at 6,358 cases across 15 states, up from 1,947 cases in nine states, with hospitalisations rising from 98 to 278 and the first two deaths added. Almost none of that is new illness. The agency changed who belongs in the count.

The Food and Drug Administration, which publishes the investigation alongside the CDC, states the change at the top of its own page: "Due to additional interview data, CDC’s confirmed cases were expanded to include illnesses in people who reported exposure to Taco Bell or recalled Taylor Farms de Mexico iceberg lettuce. As a result, additional cases and states are being added to the outbreak." It adds, in the same paragraph, that "many of these illnesses began before the Taylor Farms de Mexico iceberg lettuce recall on July 17, 2026." The six states joining are Arkansas, Iowa, Missouri, Nebraska, New Hampshire and North Carolina.

Until Wednesday the number had not moved. The CDC last revised it on 24 July, and every archived copy of the outbreak page taken between then and the evening of 5 August carries an identical Fast Facts block reading "at least 1,947" cases, 98 hospitalisations, no deaths and nine states. That is twelve and a half days in which the outbreak, on the agency’s own public accounting, did not grow at all. On 3 August the page’s own date stamp was changed while the figures beneath it were not.

The wording moved with the number. In mid-July the page said "more than 1,644". From 24 July it said "at least 1,947". On Wednesday the hedge disappeared and the figure became a flat 6,358. The title lost a number too: on 3 August it read "Cyclospora Outbreak Linked to Iceberg Lettuce in 9 States", and it now reads simply "Cyclospora Outbreak Linked to Iceberg Lettuce".

The web address has a history the current page does not mention. Until 17 July, cdc.gov/cyclosporiasis/outbreaks/07-26 was a different page, headed "Cyclosporiasis Outbreak with Unknown Source". It described "a large multistate outbreak of cyclosporiasis occurring in at least four midwestern states" and reported that, as of 13 July, "more than 400 people infected with Cyclospora have been reported to CDC from 4 states in connection with this multistate outbreak: Michigan, Ohio, West Virginia, and Kentucky." It said no specific food had been confirmed as the source, and that sick people had reported symptoms on or after 22 June. The lettuce outbreak now published at that address gives the same 22 June start, and all four of those states are among its fifteen.

This site has twice published the outbreak as a share of the national case count — about 47 per cent on 26 July and about 29 per cent on 31 July, the second under a headline saying the recall accounted for under a third of confirmed cases. Both were computed on what each CDC page said that day. The lower figure was lower mainly because the outbreak number had stopped being revised while the national one kept climbing.

We are not replacing them with a new share, because the two counts are not comparable. The national figure runs to illnesses reported through 3 August and its published onsets stop on 28 July; the outbreak figure carries no data-cut date at all, and the FDA gives its onsets as running through 31 July, so it contains cases the national count cannot yet include. The national count is compiled by the state where a patient lives, while the outbreak count follows where people were exposed. The outbreak page states no case definition of its own.

What the national surveillance page does say is that the CDC has recorded 10,468 laboratory-confirmed domestically acquired cases of cyclosporiasis since 1 May, with 517 hospitalisations, two deaths and cases in 47 states. It is aware of more than 12,255 further cases that are not laboratory-confirmed or that require further investigation, many of them from Michigan. A further 1,341 confirmed cases were acquired abroad and are counted separately.

The share of confirmed cases that ended in hospital has fallen through the season: 7.4 per cent of the 4,173 cases reported by 20 July, 6.3 per cent of 6,707 by 27 July and 4.9 per cent of 10,468 by 3 August. The likelier reading is not that the illness has softened but that testing has widened, and that surveillance now reaches people who would not have been counted in June.

The two deaths are both in Michigan, in people the FDA describes as having "significant underlying health conditions that may have been adversely impacted by cyclosporiasis and dehydration", and both fell ill before the 17 July recall. The CDC attached them to this outbreak in a notice on 3 August, on a day its own outbreak page still reported the outbreak as having caused none.

Both counts will keep moving. The CDC says it estimates a six-week lag between when someone falls ill and when their case reaches the agency, and the FDA says separately that it can take as long as six weeks to establish whether a sick person belongs to a cyclospora outbreak. The median onset date among the national cases is 4 July, which sits inside that window, so the most recent weeks are the least complete part of both numbers.