CDC’s August 20 update puts the iceberg lettuce Cyclospora outbreak at 10,930 illnesses, 454 hospitalizations and two deaths across seventeen states, with no new states added. Federal regulators have announced twenty-one foodborne outbreak investigations so far in 2026, against twenty-eight in all of 2025. The number of outbreaks is unremarkable. What is remarkable is that this one outbreak hospitalized nearly twice as many people as every outbreak announced last year combined, and that the tool for shortening it sits on the books unfunded until July 2028.
The number of outbreaks this year is ordinary. The number of sick people is not.
The U.S. PIRG Education Fund asked the question directly this week and answered it in one word: no. Regulators announce between twenty and thirty a year. There were twenty-eight last year and twenty-one so far this year. NBC News reported Saturday that none of the five researchers it consulted found evidence of more outbreaks now than in years past.
Then look at what those outbreaks did to people. PIRG’s Food for Thought 2026 counted 1,003 illnesses, 235 hospitalizations and 22 deaths across all twenty-eight of last year’s outbreaks. Set that against the one running now.
Measure | All 28 outbreaks announced in 2025 | The iceberg lettuce outbreak alone, as of August 20, 2026 |
Illnesses | 1,003 | 10,930 |
Hospitalizations | 235 | 454 |
Deaths | 22 | 2 |
Sources: PIRG, Food for Thought 2026, for 2025; CDC’s outbreak page as of August 20, 2026, for the Cyclospora figures. Check: 10,930 divided by 1,003 is 10.9; 454 divided by 235 is 1.9. Deaths are the one row where last year is higher.
Counting outbreaks measures how often regulators announce something. Counting recalls measures the same thing with worse resolution. Neither measures how many people ended up in an emergency room. Track only the announcements and this summer looks average. It was not average for the people in that right-hand column.
CDC tells you to wash the lettuce, and then tells you washing will not work.
Read the prevention bullets on CDC’s alert page in order. Wash your hands and any fresh produce thoroughly under running water. Then, immediately beneath it: since washing alone cannot guarantee removal, cook produce to at least 158 degrees Fahrenheit.
That is advice to boil your salad. I do not write that to mock the people who drafted it. It is the honest end of a road with no other exit. There is no validated kill step for this parasite on ready-to-eat produce. FDA says on its own fact sheet that chlorine and the other common antimicrobial treatments are not effective against it, and that the fecal-indicator testing the industry runs on water and product will not find it. Heat that kills it destroys the product.
Everything that actually prevents this happens upstream, in the water, months before anybody shreds a head of iceberg. When that fails, the consumer bullets are all that is left, and a shopper standing in front of a bag of lettuce has no move available that would have protected her.
This one outbreak filled more hospital beds than every outbreak announced last year combined.
Four hundred fifty-four hospitalizations in this outbreak. Two hundred thirty-five across all twenty-eight outbreaks announced in 2025. Cyclospora is the mild one in the deck, and by attack rate and case fatality it is. It still filled nearly twice the hospital beds of an ordinary year.
CDC says routine stool tests do not always screen for this parasite.
That sentence is on the alert page, addressed to patients, telling them to request the test by name. It is a concession about the denominator. The 10,930 counts people sick enough to seek care, whose clinician thought of this organism, and whose laboratory ran the right panel. Nobody knows how many failed one of those three tests, and CDC says so on the page where it publishes the number.
There is a fourth barrier, and it is not lag or behavior. In Lee County, Florida, a clinic serving mostly uninsured patients has stopped testing for this parasite because the test runs its patients hundreds of dollars. It has seen cases anyway, in patients and staff, and none are in Florida’s count. That is one clinic deciding about cost, not a state refusing anyone. It still changes the category of the undercount: lag and behavior describe people the system missed, cost describes people it priced out. If you do not have insurance you may not be tested at all, and an illness never tested is not an illness that is slowing.
Set that next to what happened to the surveillance. On July 1, 2025, reporting through FoodNet became optional for six pathogens, including this one. The network had required it since the nineteen-nineties; CDC said funding had not kept pace. Twelve months later this country recorded the largest Cyclospora outbreak in its history. FoodNet was never built to detect outbreaks and I will not claim it would have caught this one. It was built to tell us what normal looks like, which is what you need to know when something stops being normal.
The rule that would have drawn a line around one facility cannot be enforced until July 2028.
Congress ordered the Food Traceability Rule in January 2011 and set a deadline of January 2013. FDA missed it, a nonprofit sued, a judge set a schedule, and the rule issued in November 2022 with a compliance date of January 2026. In 2025 that date moved thirty months. Then Congress directed that no appropriated funds be spent enforcing the rule before July 20, 2028. FDA has said it intends to comply with the directive. I have laid the whole chronology out separately.
The distinction that gets lost matters most. A delay is an agency deciding industry needs more time. A funding prohibition is Congress removing the agency’s ability to act even if it changes its mind. The rule is law. FDA may not spend a dollar enforcing it until the summer of 2028.
Traceability draws a line. It says what everything else is. Without it, the advice to consumers has to be categorical, because categorical is all anyone can support, and categorical advice prices a category. Growers who did nothing wrong watched lettuce discounted or plowed under this summer. That loss belongs in the same ledger as the illnesses, and it is an argument for the rule the produce industry should be making itself.
FDA has not met its domestic inspection targets since 2018, and reaches about five percent of its foreign target.
The Government Accountability Office reported that FDA has not met the inspection frequencies Congress wrote into law since fiscal year 2018. On the foreign side the gap is not close: against a statutory target of 19,200 inspections a year, FDA averaged 917 from fiscal 2018 through fiscal 2023, about five percent. I wrote about that report when it came out. It belongs here because of geography. The lettuce came from central Mexico. Nobody was assigned to look at the great majority of foreign facilities shipping food into this country, because there was nobody to assign. That is architecture and budget, not a failing of the investigators who did go.
Endemic describes where an organism lives. It does not explain how it got into the lettuce.
You will hear, in nearly every account of this outbreak, that Cyclospora is endemic to Mexico. It is. That tells you the parasite is present in the region. It does not tell you how it reached a leaf shredded, bagged and eaten in Michigan. This organism comes out of human sewage, and it reaches produce when sewage reaches water that touches the crop. That is not an ambient condition of a country. It is a sanitation failure with a location and decisions behind it. Endemicity is a real fact doing rhetorical work it was not built for, turning something somebody could have prevented into weather nobody owns.
No new states is the honest version of the outbreak slowing.
The seventeen-state footprint has stopped moving. The counts inside it keep climbing, which is what a long reporting tail looks like on a parasite with a two-week incubation and a test most doctors have to be prompted to order. FDA’s investigation page says the recalled product is long past its best-by dates and should be gone from stores and restaurants. It has been gone for weeks. The number went up anyway.
What none of this is, at any point, is a consumer behavior problem. The field for next summer gets planted in a few months. Nobody chooses what is living in the water when it does, and washing will not help then either.
