FDA Has Seventeen Open Foodborne Illness Investigations, 7,320 People Sick and Three Dead. Nine of Them Have No Food Named at All, Eleven Have Never Had an On-Site Inspection, and the Advisory for the Outbreak That Killed Someone Has Not Been Touched Since June 4.

FDA keeps a public list of every outbreak its CORE Response Teams are working, and it is the only place in the federal government where you can see how many foodborne illness investigations are open at one time. The page is called Investigations of Foodborne Illness Outbreaks, it was last updated on August 6, and it lists seventeen active investigations. Most people have heard of two of them.

Run the case counts down — some sit in the table, some are buried in the individual advisories behind a link that just says See Advisory — and the seventeen add up to 7,320 reported illnesses, at least 391 hospitalizations where a hospitalization figure has been published, and three deaths. Two of the dead are in the Cyclosporaoutbreak. The third died in a multi-year Listeria outbreak traced to soft cheese, and I will come back to that one.

Cyclospora is most of the table, and it is worth saying plainly: seven of the seventeen open investigations are Cyclospora, and together they carry 6,536 illnesses, which is 89 percent of every sick person on the page. FDA has named a food in exactly one of the seven. The other six — ref #1398, opened August 5 with ten people and traceback just beginning, then #1392 at 115, #1385 at 25, #1384 at 18, #1381 at 2 and #1375 at 8 — are all listed as Not Yet Identified. And the one file with a food named, the fifteen-state iceberg lettuce outbreak at 6,358 illnesses, 278 hospitalizations and two deaths, is itself a fraction of what the state health departments are reporting. My own fifty-state tally as of yesterday was 26,755.

Here is the whole table as FDA published it, with the counts filled in from the individual advisories rather than left as a link.

Nine of the seventeen investigations have no food identified at all, and those nine hold 294 sick people: six Cyclospora files, a Listeria file with nine, a Salmonella Oranienburg file with ninety-three that has been open since July 8, and an E. coli O157:H7 file with fourteen. FDA’s most interesting sentence this month is attached to one of them. File #1385 grew from ten cases to twenty-five, and the agency explained that it now includes additional illnesses based on similarities in reported exposures. That is FDA telling the public it has found a common exposure pattern in an outbreak whose food it will not name.

Look next at the on-site inspection column, because it is the one nobody reads. Eleven of the seventeen investigations show no on-site inspection initiated, and that includes the two largest outbreaks on the page: the iceberg lettuce, and the Sinaloa jalapeños that have sickened 345 people in twenty-seven states and hospitalized thirty-six. It is not indifference. It is geography. The lettuce was grown and processed in central Mexico and the peppers came out of Sinaloa, and FDA does not walk onto a foreign farm the way it walks into a plant in Indiana. In the frozen blueberry outbreak — twelve sick, four hospitalized, every case in Florida and Georgia — you can watch the substitute in action. FDA conducted what it calls a remote regulatory assessment of Frutas y Hortalizas del Sur S.A. of San Carlos, Chile, collected records from the firm, developed concerns about other lots, and added the company to Import Alert 99-35. A records review conducted over the internet is now what stands in for someone walking the ground where the contamination happened. Every one of the six investigations that did get an on-site inspection involves a firm on American soil.

Then there are the dates. FDA stamps every advisory with the day its content was last made current, and those stamps are the quietest problem on the page. The soft cheese Listeria advisory — the one with the death — has not been updated in sixty-five days. The infant botulism advisory is thirty-three days old. The shell egg advisory is fifteen days old, and that one matters this weekend, because the recalled Midwest Poultry eggs carry best-by dates running through August 17, 2026. Recalled cartons are in refrigerators right now and the page that tells people about them has not moved in two weeks.

The status column has its own problem. Six of the seventeen say the outbreak has ended while the investigation is still carried as Active. On the moringa leaf powder file, the linked advisory says in so many words that CDC has declared the outbreak over and that FDA’s outbreak investigation is complete — 131 people sick in thirty-eight states, thirty-six of them hospitalized, the ingredient imported from India — and the table still lists it as Active. The cantaloupe file was posted on February 25, seventy people sick, outbreak ended, no advisory ever written, and it is still sitting there as Active. If Active means an open file rather than active work, the public has no way to tell those two things apart, and the number seventeen means less than it looks like it means.

The worst paragraph on any of these pages is not on the front table at all. It is inside the infant botulism advisory. Four infants in California, Pennsylvania and Washington were hospitalized after drinking Nara Organics powdered formula, and California detected Clostridium botulinum in an opened can taken from one of their homes. Tracing it back, FDA found that the Nara lots were made with milk supplied by Organic West Milk and spray dried by Dairy Farmers of America — the same two suppliers as the ByHeart infant botulism outbreak in November 2025. FDA had already run a traceforward during that outbreak to find out who else received that milk. Organic West Milk gave the agency a customer list that left Nara Organics off it, and has since told FDA the information it provided was incomplete. Four more babies went to the hospital in the gap. My firm represents families in both of those outbreaks, so read that with whatever discount you think it deserves — but the sentence is FDA’s, not mine, and it is sitting on FDA’s own website.

The eggs deserve one more line. FDA says outright that Midwest Poultry does not account for all the illnesses in that outbreak, and the arithmetic on its own page shows why: ninety-eight people have been reported sick across seventeen states, and the recalled eggs have been confirmed distributed in six. Eleven states have sick people and no confirmed egg. Finding where the rest came from is the part that is still open.

And understand what this table is not. Every number on it is a laboratory-confirmed case that survived a doctor’s visit, a stool test, a state lab and PulseNet, and was then matched to an outbreak. CDC’s own burden work puts roughly twenty-nine Salmonella illnesses in the community for every one confirmed, and roughly eighty-three for Cyclospora. Seven thousand three hundred twenty is not a headcount of the sick. It is a headcount of the counted.

Seventeen open investigations. Nine with no food named. Eleven with no on-site inspection. Six carried as active with the outbreak already over. One advisory untouched since June while the person it is about is still dead. I have been doing this since 1993 and I have never found a better one-page argument for why traceability, transparency and enough inspectors are not abstractions. The page is public. Anyone can read it. Almost nobody does.