This week the investigation moved into court. My firm filed the first lawsuit of the outbreak in federal court in Ohio — Ayyad v. Pacific Bells, LLC — on behalf of a North Olmsted man who ate at his local Taco Bell twice in mid-June, fell ill days later, and tested positive for Cyclospora. The suit names the company that operates that restaurant, along with the still-unnamed growers and suppliers behind the produce, and it won’t be the last.

The likely source of the summer’s Cyclospora outbreak now has a supplier and a restaurant attached to it. Investigators have traced the suspected vehicle to shredded iceberg lettuce grown by Taylor Farms and served at Taco Bell, the Washington Post reported on July 16, citing two people familiar with the investigation. The thread runs through all four of the hardest-hit states: a high share of the sick had eaten at Taco Bell, lettuce was the menu item they had in common, and when the FDA asked the chain where that lettuce came from, the answer — in Michigan, Ohio, West Virginia and Kentucky alike — was Taylor Farms. No agency has confirmed it on the record, the source is still called “potential,” and neither company had responded as of publication. But neither name is new to this kind of story.

Taco Bell has been the name in a lettuce outbreak before — more than once. In 2006, the CDC publicly named the chain in an E. coli O157:H7 outbreak across the Northeast that sickened 71 people, hospitalized 53 and left eight with hemolytic uremic syndrome, a form of kidney failure; investigators first blamed green onions, then settled on shredded lettuce. In 2010 and again in 2011, multistate Salmonella outbreaks were traced to the chain the CDC would only call “Restaurant Chain A” — later confirmed to be Taco Bell — in which roughly nine in ten of the sick had eaten lettuce. Further back sit a 1999 E. coli outbreak tied to its beef tacos and a 1995 hepatitis A outbreak traced to an infected worker. The produce that keeps surfacing in Taco Bell’s outbreaks is lettuce.

Taylor Farms’s record is heavier still. In 2013, the FDA traced a multistate cyclosporiasis outbreak — the illness clusters at Olive Garden and Red Lobster restaurants in Iowa and Nebraska — to salad mix from Taylor Farms de Mexico. In 2015 the company recalled a celery-and-onion mix tied to an E. coli outbreak in Costco chicken salads that sickened nineteen. And in 2024, Taylor Farms slivered onions were the likely source of the E. coli O157:H7 outbreak linked to McDonald’s Quarter Pounders — more than a hundred people sick across fourteen states, dozens hospitalized, one dead — after which FDA inspectors documented poor handwashing and dirty equipment at the company’s Colorado processing plant. A Cyclospora outbreak traced to Taylor Farms lettuce would not be the first time the parasite, or a supplier by that name, sat at the center of one.

All of this is unfolding while the official count sleepwalks. Every new case added to the national tally this week came from the same handful of states that still bother to update; Michigan alone drove the running total past 7,500. The rest have gone quiet — Ohio’s dashboard hasn’t moved in a week, sixteen states appear only as a range of “one to ten,” and the CDC’s own confirmed count froze at 1,645 three days ago. The outbreak isn’t slowing down. The counting is.

Look at how lopsided it has become. Michigan’s 4,312 cases — more than half the country’s total — dwarf the CDC’s entire four-state “outbreak,” which the agency still pegs at just over 400. In a normal year the state sees forty or fifty. Set one state’s dashboard against the federal cluster and the official number stops looking like the outbreak and starts looking like the slice of it investigators have so far been able to stitch together.

Ask all fifty health departments whether they are part of this and you get three different answers. Ohio, West Virginia and Kentucky say yes and are working the case alongside Michigan. Indiana, sitting on 206 cases, calls itself part of a nationwide outbreak. And then a run of states carrying real, elevated counts says, in effect, nothing unusual here: New York, at roughly 470 cases, calls it “not a major deviation from the norm”; Illinois, at 216, says there is “no evidence of a large outbreak”; New Jersey, at 46, is “not experiencing … outbreaks”; Massachusetts, at 18, calls it a “normal seasonal amount”; and Virginia, at 10, says its cases are not a pattern that would “constitute an outbreak”. Two of the largest states never entered the outbreak at all: California is seeing fewer cases than last year, and Washington says flatly that it “is not seeing an outbreak.” Whether a sick person is counted as part of an outbreak depends heavily on which state line they were standing behind when they got sick.

This is what last summer’s paperwork looks like once it reaches the produce aisle. When the CDC quietly made Cyclospora reporting optional under its FoodNet program in July 2025, it weakened the very machinery that turns a scatter of state counts into one fast national picture. The parasite is nationally notifiable in only 47 states, it is invisible on a routine stool test unless a physician asks for it by name, and every figure in the chart below is a floor, not a ceiling. When Oklahoma and Kansas finally opened their own books this week, each found roughly five times the cases the CDC had listed for it. The gap widens: a parasite moving through the national salad supply on one side, and on the other an official tally that trails the real one by thousands of cases.

For the person on the third week of watery diarrhea, whether their state is “officially” part of the outbreak is a distinction without a difference. They are sick, the source is still on the shelf, and the system built to find it is being asked to do the job with fewer people and a smaller net.

Here is where all fifty state health departments stood as of July 15, 2026 — the number each reports for 2026, and whether it calls those cases part of the outbreak.

The CDC’s official multistate outbreak of cyclosporiasis is four midwestern states — Michigan, Ohio, West Virginia and Kentucky — and just over 400 cases it has managed to tie to a  yet unnamed, common, source. That is the number the word “outbreak” gets attached to. On its own surveillance page, the CDC counts 1,645 lab-confirmed domestic cases across 34 states and admits it is aware of more than 5,100 more it has not finished confirming. Add up what the state health departments are actually reporting and the total sits closer to 7,000. The official outbreak and the real one are not the same size.

Start with Michigan, which has passed 3,700 cases on its own dashboard in a state that in a normal year sees forty or fifty, and whose investigators have named lettuce and salad greens as the likely vehicle. Set that one state against the CDC’s 400-case, four-state cluster and the federal figure stops looking like the outbreak and starts looking like the slice of it investigators have so far been able to stitch together.

Ask all fifty health departments whether they are part of this and you get three different answers. Ohio, West Virginia and Kentucky say yes and are working the case alongside Michigan. Indiana, sitting on 206 cases, calls itself part of a nationwide outbreak. And then a run of states carrying real, elevated counts says, in effect, nothing unusual here: New York, at roughly 394 cases, calls it “not a major deviation from the norm”; Illinois, at 216, says there is “no evidence of a large outbreak”; New Jersey, at 46, is “not experiencing … outbreaks”; Massachusetts, at 18, calls it a “normal seasonal amount”; and Virginia, at 10, says its cases are not a pattern that would “constitute an outbreak”. Two of the largest states never entered the outbreak at all: California is seeing fewer cases than last year, and Washington says flatly that it “is not seeing an outbreak.” Whether a sick person is counted as part of an outbreak depends heavily on which state line they were standing behind when they got sick.

This is what last summer’s paperwork looks like once it reaches the produce aisle. When the CDC quietly made Cyclospora reporting optional under its FoodNet program in July 2025, it weakened the very machinery that turns a scatter of state counts into one fast national picture. The parasite is nationally notifiable in only 47 states, it is invisible on a routine stool test unless a physician asks for it by name, and every figure in the chart below is a floor, not a ceiling. When Oklahoma and Kansas finally opened their own books this week, each found roughly five times the cases the CDC had listed for it. So, the gap widens: a parasite moving through the national salad supply on one side, and on the other an official tally that trails the real one by thousands of cases.

The four-state outbreak isn’t the outbreak. It’s the part of the outbreak we can still see.

For the person on the third week of watery diarrhea, whether their state is “officially” part of the outbreak is a distinction without a difference. They are sick, the source is still on the shelf, and the system built to find it is being asked to do the job with fewer people and a smaller net.

Here is where all fifty state health departments stood as of July 15, 2026 — the number each reports for 2026, and whether it calls those cases part of the outbreak.

Notes: State figures mix confirmed and probable cases and use different start dates (some “since May 1,” some “year-to-date”), so they are not strictly comparable. “In CDC count” means the state appears in the CDC’s 34-state domestic tally but has not published its own 2026 number; “no domestic cases” means the state is not in that tally (it may still have travel-related cases). Every count is a floor. Figures current to July 15, 2026.

On July 14, Senator Amy Klobuchar sent a letter to the acting heads of the CDC and FDA asking them to restore the food-safety programs this administration cut and start tracking foodborne illness again. She named them — FoodNet, the Food Emergency Response Network, the Food Safety Inspection Service, the Public Health Infrastructure Grants, and the Preventive Services Block Grants. I’ve spent thirty years representing the people those programs protect, and I’ll say it plainly: she’s right, and everyone who eats should be behind her.

Here’s what she’s pointing at. Last year the administration cut FoodNet’s tracking from ten foodborne pathogens down to two — and Cyclospora, the parasite now tearing through the country, was one of the eight it dropped. Reporting it became optional. The FDA’s Food Emergency Response Network was suspended after its scientists were let go, and the grants that pay for state lab capacity and disease tracking were terminated. Her letter lands the consequence in one sentence: there is no longer a central place to report and compare this data across state lines. 

A word about the numbers, because they make her case stronger than her letter does. She cites nearly 2,800 cases and 86 hospitalizations as of July 9. Those figures are already old and low. As I write this, the CDC has 1,645 confirmed cases and admits it is aware of more than 5,100 more under investigation — nearly 7,000 people. Michigan alone has counted 3,309. Add up just the eleven states that still publish their own tallies, and you are past 4,900.

That is the whole argument. The Senator could only cite “nearly 2,800” because the system meant to count the rest was defunded. The distance between her number and the real one isn’t a weakness in her letter — it is the proof of it. When you stop tracking a pathogen, you don’t make a single sick person well. You just lose the ability to see them. The people are still out there — children and grandparents living through weeks of dehydrating illness from a bag of salad greens — now invisible to the one agency we count on to sound the alarm.

I’ve watched this before. I started after the Jack in the Box E. coli outbreak in 1993, and the biggest improvement since has been our ability to catch outbreaks early, link cases across state lines, and pull tainted food before it kills someone. Every one of those abilities runs on the programs in Klobuchar’s letter. Strip them out and we are worse off than in 1993: the food supply is bigger and more national than ever, and a single lot of contaminated greens can reach dozens of states before anyone connects the first two cases.

So yes — I support her request without a caveat. Restore FoodNet, the Food Emergency Response Network, the Food Safety Inspection Service, the Public Health Infrastructure Grants, and the Preventive Services Block Grants. Put Cyclospora back on the list of things this country bothers to count. And do it now, while the outbreak is still climbing — because the alternative is what we are living: an agency reporting 1,645, a reality closer to 7,000, and a widening gap where real people get sick and no one in Washington is required to notice.

Senator Klobuchar is asking the right question. The rest of us should make sure it does not go unanswered.

Depending on where you look, each state in this outbreak has as many as three different case counts, and they rarely match. The CDC’s public map reports a range of lab-confirmed, domestically acquired cases. Each state health department keeps its own tally, which usually mixes confirmed and probable cases and runs higher. And the press, polling those departments day to day, often carries a third figure. The chart and maps below put all three side by side, state by state.

The national totals tell the same story in one line. As of July 13–14, the CDC’s map shows 1,645 lab-confirmed cases. The CDC says it is aware of nearly 7,000 once the more than 5,100 cases still under investigation are counted — the figure the media is now reporting nationally. And the eleven states that publish their own numbers already sum to roughly 4,900 on their own, triple what the CDC’s map shows.

The counts also disagree for structural reasons, not because anyone is lying. They use different cutoff dates, the CDC reports only what it has confirmed as domestic while states report what they see, and there is roughly a six-week lag between illness and the federal tally. Ohio shows the split at its widest: the CDC bins it at 1 to 10, the state’s last official release counted 177, and reporters now put it near 400. Washington shows the opposite — a state where all three roughly agree it is low: King County reports 15 cases but only about one acquired locally, and the state reports no surge.

Each state’s three numbers, with national totals and the suspected source boxed at right. The band is the CDC bin; the filled dot (●) is the state health department’s count; the open ring (○) is the current media figure.

The same comparison as three maps, each with its running total: what the CDC counts (1,645), what the states say (~4,900 across 11 states), and what the media reports (~7,000 nationally).

Every state with a published number, where each number comes from, and the national totals:

Totals: “1,645 confirmed” is the CDC’s lab-confirmed domestic count (the map). “~4,900” is the sum of the eleven outbreak-state health-department counts above; Washington is shown for reference but excluded, as its cases are mostly travel-related. “~7,000” is the CDC’s confirmed-plus-under-investigation figure as reported nationally by the press. † Kentucky: 100 reported statewide, 61 lab-confirmed. ‡ Washington: 15 cases in King County (Seattle), only about one acquired locally — the rest travel-related; the state reports no local surge, so unlike the outbreak states its cases are mostly imported. Florida earlier appeared in the press as 11, a stale CDC-derived figure, while the state’s Merlin system reports 50. A separate 440 travel-associated cases nationwide are not included here. Figures as reported roughly July 2–14, 2026, and still moving.

It is hard to keep up with the fast moving story.

The headlines this week put Taco Bell at the center of the Cyclospora story, and there is a real thread there. Credit to the Washington Post’s Lena H. Sun, who broke the news that federal and state health officials are investigating whether Taco Bell restaurants played a role in the outbreak. What set it off wasn’t an agency announcement: Detroit-area Taco Bell locations quietly posted signs that they couldn’t sell lettuce, cilantro-onion, pico de gallo, and guacamole “due to a nationwide recall” — a recall that doesn’t publicly exist. Sun is careful about the limits, and so am I: some of the sick reported eating at Taco Bell, and plenty didn’t, which tells you the chain is not the sole source of this thing. No grower, no supplier, and no Taco Bell ingredient has been named by the CDC or FDA.

Here is the part getting lost in the noise: this appears to not be one outbreak. The CDC now says as much in its own words, describing a large multistate outbreak across Michigan, Ohio, West Virginia, and Kentucky — the lettuce-and-salad-greens track Michigan has been pointing at — while separately tracking multiple additional clusters. And those other clusters point somewhere else entirely. Back on July 7, CNN obtained a CDC email showing the FDA’s Coordinated Outbreak Response and Evaluation network running traceback on cilantro, white and green onions, and cucumbers for a different set of cases — in Illinois, New York City, New York State, Pennsylvania, and Texas — that appear to be linked to Mexican-style restaurants, a grocery chain, and a catered event. Different states, different foods, different points of sale. The FDA’s own outbreak table now lists several distinct investigations running at once, each with the product not yet identified and traceback initiated. 

And that single number moved this week — hard. I spend a lot of time on this blog holding the CDC’s feet to the fire, so let me do the opposite for once. On Tuesday the agency issued a formal Health Alert Network advisory and revised its count from the 843 confirmed cases it was reporting late last week to 1,645 lab-confirmed, plus more than 5,100 under investigation — nearly 7,000 people nationwide, across 34 states, with 141 hospitalized and no deaths. The CDC put the scale in plain language: 1,645 confirmed is substantially higher than the 249 cases reported nationally over the same stretch last year. Michigan alone has now confirmed 3,309. For weeks the distance between what the states were counting and what Atlanta was publishing was the whole story; this week the agency finally revised upward, pushed the alert to every clinician and lab in the country, and named the outbreak. That is what a functioning surveillance system is supposed to do. Credit where it’s due.

And credit to the reporters who wouldn’t let it go. Beyond Sun’s work at the Post, NBC News’ Erika Edwards has been on this outbreak for weeks — chasing the state-by-state counts the federal dashboard wasn’t capturing and putting the surveillance gap in front of a national audience while the CDC’s number still read 843. Her colleague Aria Bendix reported last year that the CDC had quietly dropped Cyclospora from FoodNet — the decision that set this whole counting mess in motion. Good beat reporting isn’t glamorous; it’s showing up, day after day, on a story about a parasite most people would rather not read about over breakfast, and making sure the families getting sick actually get counted. Sun and Edwards did that.

None of this means the problem is solved. The only reason the states had to lead is that the CDC made Cyclosporareporting optional on July 1, 2025, and the federal count spent this outbreak playing catch-up from behind. The CDC is updating its state map as cases come in, and for once it is moving faster than it was a week ago. But the honest headline isn’t “Taco Bell did it.” It’s that a fractured food-safety system is now chasing several Cyclospora outbreaks at once — and the people getting sick deserve to have every one of them counted, traced, and named.