The cyclosporiasis season closed on August 31. The counting did not. CDC now reports 18,445 laboratory-confirmed cases of domestically acquired cyclosporiasis since May 1, with 990 hospitalizations and two deaths across forty-nine states and the District of Columbia. The comparison figure for the same four months of 2025 is 1,180. The total will keep climbing for weeks, because CDC says it takes about six weeks from the start of symptoms to the day a case reaches the agency, so illnesses that began in late August will not finish arriving until October.
The Taylor Farms lettuce is not the whole of it, and that distinction is where the next several months of work sit. CDC, FDA and state officials have linked 11,458 of those illnesses in twenty states, with 495 hospitalizations and both deaths, to processed iceberg lettuce from Taylor Farms de Mexico, recalled on July 17. That leaves close to seven thousand laboratory-confirmed cases with no food named at all. The two counts carry different as-of dates, so treat the subtraction as an estimate rather than a figure and add to it the 3,491 confirmed infections CDC is still sorting into domestic or travel-related and the at least 13,505 further cases the agency says it is aware of. Epidemiologists will be working through the unattributed cases into the fall and the winter, and so will I.
Every one of them is a person who ate something. The open question is what.
Congress spent the same week doing something else. On September 1 the House cleared the Senate’s continuing resolution by 370 to 48, and the President signed it the next day. It funds the federal government at fiscal 2026 levels through December 11, thirty-eight days after the midterm elections. Section 780, the provision barring FDA from spending a dollar to administer or enforce the Food Traceability Rule before July 20, 2028, is a fiscal 2026 provision. No member voted to keep it this month. It survived because a stopgap carries the prior year forward, and because nobody asked.
While that was happening, Taylor Farms posted on its own website that it fully supports and endorses the Food Traceability Rule, that it implemented traceback protocols ahead of the compliance deadline, and that each of its products can be traced to grower, ranch, lot, harvest date and time. Read that beside the paragraph above. The company at the center of the largest foodborne outbreak of the year says it can already do the thing Congress has forbidden the FDA to require.
I am, by disposition, a hopeful person. It is getting harder. Eighteen thousand people are confirmed sick, two are dead, an industry has lost a season of consumer confidence, and the federal answer has been to keep the lights on until December.
Here are ten things that would have prevented this outbreak or made it smaller. Not one of them requires an invention. They run roughly in order of how much illness they would have stopped, which is not the same as the order of how easily they can be done.
What would have kept the parasite out of the lettuce.
1. Fix sanitation in the growing region. Humans are the only known host of Cyclospora cayetanensis. Every oocyst that reached an American plate came out of a person. That makes this a question of field toilets, handwashing, sewage and irrigation water, which is uncomfortable to say and not remotely mysterious.
2. Treat agricultural water against the actual hazard. The leafy greens marketing agreements require an antimicrobial validated against generic E. coli, a bacterial indicator scored by a bacterial test. FDA has said plainly that traditional fecal-indicator testing will not find this parasite. We are measuring the wrong organism and reporting the result as safety.
3. Verify the foreign supplier, which was already the law. An importer owes a duty under the Foreign Supplier Verification Programs rule to analyze the known or reasonably foreseeable hazards in what it brings in, and FDA’s own list of those hazards names parasites first. Cyclospora on imported produce has been known and foreseeable since the 1996 raspberries, and this company’s own 2013 outbreak was notice in writing. That obligation needed no rulemaking, no compliance date and no act of Congress. It has been on the books since 2017.
4. Make import surveillance of Mexican leafy greens permanent. After the 2013 Cyclospora outbreak traced to the same company, FDA announced increased surveillance. Thirteen years later it announced increased surveillance again. It worked both times. What did not exist was the thirteen years in between.
5. Fund the research. Two of the three federal Cyclospora research projects were defunded in fiscal 2026, and the third is being moved out of Beltsville. There is no culture method and no animal model for this parasite. That work is the only route to a validated kill step, and it was cut in the year of the largest outbreak ever recorded.
What would have made the outbreak smaller once it started. This is the group you and I can still move.
6. Put the Food Traceability Rule back on schedule and take Section 780 off. Lot-level records, producible in twenty-four hours, leafy greens on the covered list, compliance January 20, 2026. A coalition of consumer, public health and labor organizations asked Congress to rescind the rider on August 12, pointing directly at this outbreak, and has had no answer. December 11 is the next date on which it can be undone. After that the calendar runs to July 2028 by itself.
7. Write it into the purchase order. Nothing stopped any large buyer from making parasite-specific water treatment and verified sourcing a condition of purchase. Private contracts move faster than FDA ever will, and the biggest customers in this supply chain have the leverage to write the specification in an afternoon.
8. Name the food faster. The first illness began June 14. FDA announced its investigation July 16. A month of exposure ran in between, and the records that would have closed that gap are the ones the rule requires and the rider forbids paying for.
What would have let us see it while it was happening.
9. Get the test ordered and publish one national number. Routine stool panels do not always screen for Cyclospora, so a clinician has to ask for it specifically. CDC advised clinicians in July. Forty-one states carried cases, a shrinking handful still maintain a public page, and CDC has now come off a weekly schedule on the largest outbreak of the year. The public cannot measure a thing it cannot see, and neither can Congress.
10. Make the test affordable. People without insurance do not pay out of pocket for a stool panel to diagnose a diarrheal illness they expect to outlast. Every untested case is a case that never counts, and the count is the only thing that moves an agency.
Four dates describe this summer better than any argument I can make. June 14, the first person got sick. June 15, FDA convened a public meeting on making lot-level traceability easier. July 15, the comment period on that meeting closed. July 16, FDA announced it was investigating. The meeting about relieving the industry of the requirement opened the day after the first person fell ill and closed the day before the government said out loud that there was an outbreak. The lobbying record explains how the dates got that way.
None of the ten is a lawsuit and none of them is mine to accomplish. Thirteen members of Congress have written about this outbreak and not one has a published answer. The trade associations that spent the summer explaining what else it might have been could spend the fall asking for the rider to come off, and the companies that say they already comply have the strongest standing of anyone to ask. December 11 is the date. After that, everybody gets to be surprised again next June.