Credit Where It Is Due on Formula Nutrients — and a Reminder That Fifty-Two Babies Were Paralyzed by Something Else

By Bill Marler on October 29, 2026

Today HHS Secretary Robert F. Kennedy, Jr., standing at a children’s center in Des Moines, announced the first major update to infant formula nutrient standards in nearly thirty years. The FDA issued the first of three planned proposed rules. It would update the requirements for iodine, manganese, vitamin A and vitamin D, reaffirm the requirements for chloride, potassium, selenium and sodium, and ask whether nutrients most manufacturers already add, such as DHA, ARA and L-carnitine, should be required. Comments are due February 26, 2027, in docket FDA-2026-56789. The agency also released another round of chemical contaminant testing, launched an Infant Formula Testing Webpage and a Chemical Contaminants Dashboard, and said it is working toward action levels for certain contaminants under its Closer to Zero initiative.

That is real work, and I want to say so plainly. The nutrient requirements have not had a comprehensive review since 1998. Putting contaminant results on a public dashboard is exactly the kind of transparency I have spent years asking agencies for. Operation Stork Speed and the Make America Healthy Again push forced the first serious federal look at infant formula in a generation. I said that to the Senate HELP Committee in June, and I will keep saying it.

Secretary Kennedy said today that “Parents deserve confidence in every bottle they feed their baby.” I agree. Which is why I read the release twice.

What the release does not say

The subheading says the new testing data “reaffirms safety of infant formula.” The testing it describes is for chemical contaminants. The words Clostridium botulinum do not appear in it. Neither do Cronobacter, Salmonella, nor the word pathogen. Nothing in today’s announcement touches the hazard that actually put American babies in intensive care this year and last.

Since November 2025, fifty-two infants have been hospitalized with botulism after being fed powdered infant formula: forty-eight in the ByHeart outbreak across seventeen states, and four in the Nara Organics outbreak in California, Pennsylvania and Washington. Many were on ventilators. Everyone was treated with BabyBIG, which the State of California prices at $69,300 a dose. Not one of them was hurt by a heavy metal or an outdated vitamin D requirement. They were paralyzed by spores of a soil organism that came in with the milk.

On August 27, CDC closed the Nara investigation and said, without a hedge, that the formula was contaminated with C. botulinum and made infants sick. Whole genome sequencing tied a Nara infant to a ByHeart infant, ByHeart base powder and four sealed cans of ByHeart formula. Both brands were made with milk supplied by Organic West Milk and spray dried by Dairy Farmers of America. One contamination, two brands, two outbreaks.

And the rule has not moved. Under 21 C.F.R. § 106.55, powdered infant formula makers must test finished product for two organisms, Cronobacter and Salmonella. C. botulinum is not on the list. It was not on the list when the FDA named the organism by genus and species in its March 8, 2023, Call-to-Action letter to the entire powdered formula industry. It was not on the list when the FDA wrote to industry again on July 13, 2026, after both outbreaks, telling manufacturers that their responsibility reaches every ingredient and every supplier. It is not on the list today.

The agency’s own record says testing is needed

The FDA added Cronobacter to the testing list in 2014 on a record of roughly one reported illness a year, because voluntary testing might not be enough for a food that is an infant’s sole source of nutrition. The botulism record is fifty-two hospitalized infants in two outbreaks.

The FDA’s own post-outbreak response page reports that when it sampled 48 powdered milk samples from four suppliers to infant formula makers in February 2026, two came back positive for a Clostridium carrying a gene associated with botulinum toxin. The Reagan-Udall Foundation’s September report, from a roundtable funded in part by ByHeart and the International Dairy Foods Association, found that there is no standardized method for detecting these spores in formula and that the indicator test the industry relies on can miss spores at levels high enough to make babies sick. The FAO/WHO expert group that met in Rome in June added botulinum neurotoxin-producing clostridia to the priority hazards for powdered formula and said dairy ingredients have been shown to carry them.

Letters, roundtables and referrals to international bodies are not a rule. Guidance has been tried twice now, and it has a body count.

No reply to the letters, no reply to the t-shirts

I wrote the House Energy and Commerce Committee in support of H.R. 7867, the Infant Formula Safety Modernization Act, on June 23, with fourteen families’ statements attached, and again on August 28 after CDC confirmed the genetic link. I wrote the Senate HELP Committee on June 29 and again on August 28. Four letters. No reply to any of them. In September, two “For Babies sake, Test for Botulism in Infant Formula” t-shirts, compressed into the shape of a baby bottle, were hand-delivered to every one of the 77 offices on those two committees, and one went to every other member of Congress. H.R. 7867 has sat in Energy and Commerce since its April 29 hearing without a markup.

So, I am going where the rule is actually written.

The petition

I posted a draft citizen petition earlier this month asking the FDA to add botulinum neurotoxin-producing clostridia to the finished-product testing table at § 106.55(e), and, because the organism arrived in a dairy ingredient, to require that each lot of whole milk powder, nonfat dry milk, whey and similar ingredients be tested before it is used. It also asks for one-business-day notice to the FDA of any confirmed positive, a validated test method within 180 days, lot-level records tracing ingredients back to the dryer and the farm, and environmental monitoring as a secondary control. Each request stands on its own so the Commissioner can grant in part what he will not grant in whole. Forty-six parents of twenty-four of the infants are Co-Petitioners.

The comments have made it better. A pediatrician and neonatologist wrote to endorse every request and to add that cans should tell parents the product was tested and found negative. A veteran food microbiologist answered the question I asked about testing raw milk at the dryer: test the fluid milk at the gate, where a low-level contaminant is more evenly mixed and has not yet been diluted across lots, and hold powder from any presumptive-positive tanker until the result is confirmed. Comments close tomorrow, October 30. I will file shortly after.

What I am asking

Of the FDA: you just showed you can open Part 106 and 107 and do the work in phases. Make botulism the next phase. If the agency can review thirty nutrients in three rounds, it can add one organism to a table that has two. And when you update the testing webpage and dashboard you launched today, put the pathogen results there too.

Of readers: the nutrient docket is open until February 26. If you comment, and you should, tell the FDA that a modern infant formula rule has to cover what is in the can as well as what is supposed to be in it. And when my petition is docketed, I will post the number so you can comment there too.

Nutrition matters. So does a baby who can still swallow, hold up his head and breathe on his own. Parents deserve confidence in every bottle, and confidence comes from testing for what has actually hurt babies.

Perhaps I need to order more t-shirts.

ByHeart Is “Discontinued” – But Its Founders Are Buying Perrigo’s Infant Formula Business

By Bill Marler on October 29, 2026

Posted in Case News

A hat tip to Anna Edney at Bloomberg News, who has done as much as any reporter in the country to keep infant formula safety in front of the public, for breaking the news today that the people who ran ByHeart are about to become major players in the store-brand formula business.

According to Edney’s story, a company led by ByHeart co-founders Ron Belldegrun and Mia Funt – a brother and sister who are now CEO and president of MeridianMade, a contract manufacturer – has agreed to buy Perrigo’s infant formula business for up to $175 million. Dublin-based Perrigo makes the store-brand formula sold under retailer labels across the country. Perrigo says the sale, expected to close in the first half of 2027, will let it focus on its consumer self-care products.

Most important for the families: a MeridianMade spokesperson told Bloomberg that ByHeart has been “discontinued” and will not relaunch. ByHeart formula is not coming back.

So, ByHeart is gone for good. The babies it sickened, and their families, are not. Before everyone moves on, here is a reminder of what happened and why.

What Happened

In October 2025, California’s Infant Botulism Treatment and Prevention Program noticed that three infants with suspected infant botulism had all been fed the same powdered formula – ByHeart Whole Nutrition. The signal was hard to miss. ByHeart had roughly one percent of U.S. powdered formula sales, yet of the 36 infants treated with BabyBIG between August 1 and November 10, 2025, who had any powdered-formula exposure, 15 – more than 40 percent – had been fed ByHeart. The FDA’s own internal outbreak record called that “highly unusual.”

It was the FDA, not ByHeart, that asked for the recall: two lots on November 8, 2025, then every ByHeart product on November 11. ByHeart’s own testing soon found Clostridium botulinum in five of 36 unopened cans spanning three lots.

When the CDC declared the outbreak over on February 26, 2026, the final count was 48 infants – 28 confirmed and 20 probable – in 17 states. Every one of them was hospitalized and treated with BabyBIG. Thankfully, none died. But illness onsets ran from December 24, 2023, to November 29, 2025, and ByHeart reported that one positive lot had a production date of November 15, 2024 – meaning contaminated formula was being made and sold nearly a year before the recall. Preliminary CDC data found that about two-thirds of the infants evaluated needed physical therapy or feeding support after they left the hospital. According to the FDA, this was the first botulism outbreak tied to infant formula anywhere in the world since infant botulism was first described about fifty years ago.

Why It Happened

Edney accurately reports that federal inspectors found no evidence ByHeart’s production process caused the outbreak. That is where the story starts, not where it ends. C. botulinum was found in finished formula, in the base mix, and in the organic whole-milk powder ByHeart bought from Organic West Milk, which was spray dried by Dairy Farmers of America in Fallon, Nevada. The FDA sequenced 37 isolates and sorted them into 17 strains. One of those strains tied together a milk-powder retain at Dairy Farmers of America, ByHeart’s own whole-milk powder, its finished formula, an unopened can from an Arizona family, and a sick baby in New Jersey.

Powdered infant formula is not sterile, and heat that kills Salmonella and Cronobacter does not reliably kill botulism spores. Nor did the indicator tests the industry relies on see the hazard – ByHeart itself reported that every sample in which its lab presumptively detected C. botulinum had sulfite-reducing clostridia below the level those tests flag.

None of that lets a formula maker off the hook. The FDA has said for years, and repeated in its July 13, 2026, letter to the entire infant formula industry, that suppliers of ingredients that will not receive a lethal treatment at the formula plant are an extension of the manufacturing process. If you put it in the can and sell it to parents, you own it. And ByHeart’s compliance history was hardly spotless: an FDA Warning Letter at its former Reading, Pennsylvania plant in 2023, and Form 483 inspection findings at its Allerton, Iowa base-powder plant in February 2025 and again in January 2026.

Then there is Nara Organics. When ByHeart recalled its formula, it steered its customers to Nara with a 20 percent discount code in a since-deleted Instagram post. Nara’s formula was made with milk from the same Organic West Milk, spray dried by the same Dairy Farmers of America. During the ByHeart investigation, Organic West Milk gave the FDA a customer list that left out Nara’s manufacturer. In June 2026, Nara recalled all of its whole-milk formula; the CDC closed that outbreak on August 27 with four confirmed infant botulism cases in California, Pennsylvania and Washington, all hospitalized. Whole genome sequencing matched a Nara baby’s isolate to a ByHeart baby, to ByHeart’s base powder, and to four sealed, unopened ByHeart cans. The FDA concluded the matching sequences, and shared suppliers indicate a common source. It still has not said whether the milk was contaminated before or during drying, and its root-cause investigation remains open.

What Comes Next

The outbreaks forced a real response. The FDA’s July 2026 industry letter, a joint FAO/WHO expert meeting that added botulinum-producing clostridia to the priority hazard list for powdered formula, and a Reagan-Udall Foundation roundtable all reached the same conclusion – there is no standardized method to find C. botulinum in formula, and the industry cannot rely on indicator testing to catch it.

Which brings me back to today’s news. Perrigo makes the store-brand formula that a great many families on tight budgets depend on – and Perrigo’s own Wisconsin formula plant received an FDA Warning Letter on the same day in August 2023 that ByHeart did. That business is about to be run by the people who ran ByHeart. Before it was discontinued, ByHeart said it would not resume production without C. botulinum-specific testing of every dairy ingredient and every finished batch. ByHeart will never resume, so that promise was never tested. MeridianMade should make that same commitment, publicly, for every can that leaves the Perrigo plants.

Better yet, safety should not depend on any one company’s promise. That is why I am pushing for required C. botulinum testing of powdered infant formula and its dairy ingredients – both through H.R. 7867 and through the FDA citizen petition I am filing with ByHeart and Nara parents as Co-Petitioners. Public comments on the draft petition are due tomorrow, October 30.

Marler Clark represents families whose babies were hospitalized in both the ByHeart and Nara Organics outbreaks. Executives can sell, buy and rebrand. These families do not get to start over.

FDA Says the Trader Joe’s Chicken Wrap E. coli Outbreak Is Over – and That Telling You More “Would Not Have Provided Consumers with an Action.” Section 705(b) Has a Name for That: Gross Deception of the Consumer

By Bill Marler on October 29, 2026

Posted in Case News

On October 28, the FDA declared over the E. coli O157:H7 outbreak linked to “multi-ingredient wraps with chicken” – FDA reference #1421, USDA 2026-04. Fifty people sick. Here is the agency’s explanation, in full:

“The wrap identified through the investigation is past shelf life and no longer available for purchase. The implicated wraps were proactively removed from the market before the link to the outbreak was established. At the time of this action, illnesses occurred during one week in August. Since these products were removed from sale and the wraps people ate were out of shelf life, more detailed product information would not have provided consumers with an action they could take to prevent illness. FDA’s outbreak investigation remains active.”

The product column on the FDA table still reads “Multi-ingredient wrap with chicken (suspect).” Suspect. That is five days after USDA’s Food Safety and Inspection Service confirmed to USA Today that the wraps came from Trader Joe’s, eight days after Trader Joe’s told the same paper that FDA had notified it in September and that it had pulled its Tarragon Chicken Salad Wraps from shelves in nine states, and thirteen days after we filed Gottlieb v. Trader Joe’s Co. (I represent the plaintiffs). Food Safety News has the update.

I wrote last week that naming the product cost the government nothing on October 23 and would have cost nothing on September 23. It would cost nothing today. So, let’s take the explanation apart.

“Illnesses occurred during one week in August.” That is the first epidemiological fact FDA has disclosed about this outbreak – not the states, not the hospitalizations, not the HUS cases, just a week. FDA first posted the outbreak on September 23, weeks after the last of those illnesses. It named the food, as a category, on October 7. Of course, the wraps were past shelf life by the time the public heard about them. The delay produced the shelf-life excuse, and now the shelf-life excuse is offered to justify the delay.

“More detailed product information would not have provided consumers with an action they could take.” This assumes the only reason to name a food is so you can throw it out. Fifty people are already sick. Their doctors need to know what they ate. The health departments interviewing them need to know. The people who ate a Trader Joe’s wrap in August, got sick, and were never tested need to know. And anyone who bought another product made with the same contaminated ingredient – which, by FDA’s own account, it is still trying to identify – needs to know.

“No ongoing risk” and “investigation remains active.” Both agencies have said they are looking for a contaminated ingredient and whether it went into other products. If you have not found the ingredient, how do you know there is no ongoing risk?

Frankly, withholding this information is the definition of gross deception of the consumer.

Two days before FDA closed this outbreak, the Department of Justice’s Office of Legal Counsel issued an opinion to HHS on exactly this question. It concluded that FDA’s current regulations largely block disclosure of supply chain information, but that FDA can rewrite them under section 705(b) of the Food, Drug, and Cosmetic Act, 21 U.S.C. 375(b), which lets the Commissioner disseminate information about food “in situations involving, in the opinion of the [Commissioner], imminent danger to health or gross deception of the consumer.” A regulation like that, OLC said, would be “authorized by law” under the Trade Secrets Act.

OLC also said a regulation authorizing disclosure “after a foodborne illness outbreak has already ended would likely be inconsistent with section 375(b).” That is imminent-danger logic. An outbreak’s danger ends. And two days later, FDA declared this one ended.

But the statute has two triggers, and OLC treated them as independent – FDA may authorize disclosure when “either of the statutory conditions” is met. Deception does not end when people stop getting sick. And deception is not only a false statement. It is also an omission.

Think about who the consumer is. It is a parent deciding what to put in a child’s lunch. It is the buyer at Walmart, or for a school district or a hospital, deciding which supplier to keep and which to drop. Every one of them makes that decision on the information available. When FDA knows which company’s food made people sick and keeps it to itself, every one of those decisions is made on a record the government knows is materially incomplete. The parent cannot avoid the brand, and the buyer cannot avoid the supplier, that FDA has already identified. That is deception of the consumer, and when it is about the source of an outbreak, it is gross. Section 705(b) does not say deception by the seller. It says situations involving gross deception of the consumer.

It is not just the chicken wraps. The same week FDA declared #1421 over, it posted an Executive Incident Summary Abstract for an 18-person Cyclospora outbreak traced to a catered lunch. The unredacted text says the traceback ran through a single caterer to its sole produce distributor, that two people were hospitalized, that no samples were collected, and that no product or firm actions were taken. Everything else is blacked out – the health department that investigated, the caterer, the foods of interest, the distributor, where the produce came from, and even the states where people got sick. Every fact a school, a hospital, an employer planning a lunch, or a produce buyer could use to decide whether to keep buying from that caterer or that distributor is redacted.

That is the same pattern as the 2024 romaine outbreak, where the traceback converged on one processor and one grower and the public learned their names more than a year later, through records requests and lawsuits. The government had the answer the whole time.

What I changed in the petitions. I have revised my two pending FDA citizen petitions to put this outbreak, and this argument, on the record:

Outbreak firm disclosure petition. The requested 705(b) rule now rests on both grounds. It asks FDA to define when it will find gross deception of the consumer: when an implicated firm or its customer makes a public statement about the food that the agency’s own investigative record contradicts; when the public record names only a commodity, so consumers blame firms that were not implicated; and when the agency’s own record still calls a food “suspect” after it or a partner agency has confirmed the link. It adds the case where FDA has identified the firm whose food caused an outbreak and withholds it, so that consumers and commercial buyers choose among products and suppliers on a record the agency knows is incomplete. The limiting principle is FDA’s own knowledge: the deception ground reaches what FDA has confirmed, not its working hypotheses. The petition also puts the Cyclospora abstract on the record and asks that FDA’s declaring an outbreak “ended” not, by itself, close the window for disclosure while the contaminated ingredient remains unidentified – which is exactly where #1421 sits today.

Retail consignee list petition. The removal that made the wraps “unavailable” is the same removal that, because no one called it a recall, produced no consignee list. The petition asks FDA to treat the removal of an outbreak-implicated food as what it is – a recall – whatever the firm calls it.

I plan to file both petitions together in November. If you would like a copy of either draft, or have comments, email me at bmarler@marlerclark.com – before October 30, if you can.

Fifty people got sick. The government knows the retailer, the product, and the manufacturer. FDA’s answer is that you did not need to know. The same goes for the people who ate lunch at that catered event. That is not “radical transparency.” It is the definition of gross deception of the consumer. (See also: Fifty Sick in an E. coli Outbreak, Two Federal Agencies, and Still No Names.)

Brookshire Farms E. coli Outbreak: Five Sick, Three Children in Intensive Care, and the County’s First Concrete Advice Is About the Pumpkins, Not the Animals

By Bill Marler on October 28, 2026

Posted in Case News

On Tuesday, October 27, the County of San Luis Obispo Department of Public Health confirmed five cases of Shiga toxin-producing E. coli and said it is investigating a possible link to Brookshire Farms, the pumpkin patch and corn maze at 4747 Los Osos Valley Road on the edge of San Luis Obispo. The suspected exposure window is October 16 through 26. Two of the sick are adults and three are children, and all five visited the farm, county spokesman Tom Cuddy told ABC News. The three children are in serious condition in a pediatric intensive care unit, according to the county. The farm closed voluntarily on Tuesday. No source has been identified.

This one does not look like a petting zoo outbreak, at least not yet.

Last week in Illinois, the first thing Bengtson’s Pumpkin Farm closed was its petting zoo. Nothing San Luis Obispo County has said so far points at animals. Brookshire’s promotional listings, as compiled by Food Safety News, describe hayrides, grain-cart rides, pedal carts, a jumping pillow, duck races, a playground, a hay climb, a hay maze and the corn maze. A U-pick directory does list farm animals, and that deserves a straight answer, but neither of the county’s statements mentions animal contact, an animal area or animal sampling. What the county did add, at 2:30 p.m. on Wednesday, was advice about the pumpkins themselves: throw away carved pumpkins from Brookshire, scrub intact pumpkins that will be handled a lot with hot, soapy water, and wash your hands after handling them.

Why the pumpkin advice matters.

The county calls the guidance precautionary, and I take it at its word. But health departments do not tell people to throw things away for no reason, and this tells me investigators are at least considering that the pumpkins, or what was on them, carried the bacteria off the farm. A pumpkin sits on the ground for months. Its rind carries whatever the field put on it: soil, manure or compost if the field was amended, irrigation water, wildlife droppings, runoff from livestock on or near the property. Carving opens the pumpkin up, and children scoop out the seeds and pulp with their bare hands. That is a plausible route. It is not the only one. A hay bale, a dusty maze path, a pedal-cart handle, a hand-wash station that was out of soap, a concession stand: any of them can move Shiga toxin-producing E. coli from the environment to a child’s mouth. I do not know which it was. Neither does the county, yet.

Halloween is Saturday. Pumpkins from Brookshire are sitting on porches and kitchen tables across the Central Coast right now, and many of them were carved this week. That is why the guidance should have come out with the first announcement on Tuesday, not a day later, and why it should be pushed out through schools, pediatricians and the farm’s own social media, not just appended to a county web page.

The exposure window moved, and that tells you something.

The first version of the county’s release gave an exposure window of October 16 through 18, and that is what KSBY and others first reported. At 4 p.m. the same day, the county widened it to October 16 through 26, and for a time the Spanish-language version of the release still carried the old three-day window. Going from three days to eleven can mean the county has cases with later visit dates, or that it could not rule out ongoing exposure. Either way, it suggests whatever made these people sick may have been on the farm until the day it closed. That also makes the Spanish translation more than a clerical point. Families who read the release in Spanish needed the same dates as everyone else.

Date	What happened	Source
Oct. 16–26	Suspected exposure period at Brookshire Farms	SLO County Public Health
Oct. 27, 2:29 p.m.	County confirms five STEC cases and a possible link to Brookshire; first release gives an Oct. 16–18 window	SLO County Public Health; KSBY
Oct. 27	Farm closes voluntarily	SLO County Public Health; Paso Robles Daily News
Oct. 27, 4:00 p.m.	Exposure window widened to Oct. 16–26	SLO County Public Health
Oct. 27	Cases are two adults and three children, all farm visitors; three children in serious condition in pediatric ICU	ABC News; SFGATE / SLO Tribune
Oct. 28, 2:30 p.m.	Precautionary pumpkin guidance: discard carved pumpkins, scrub intact ones, wash hands	SLO County Public Health

Sources: County of San Luis Obispo Department of Public Health release and updates (October 27–28, 2026); ABC News; SFGATE and The San Luis Obispo Tribune as reported via Yahoo News; KSBY; Paso Robles Daily News. Read October 28.

Every pumpkin patch outbreak I can find ran through the animals.

I went back through OutbreakDatabase.com, Fair-Safety.com and the health department reports looking for E. coli outbreaks at pumpkin patches. The pattern is consistent. In October 2012, Cowlitz County reported at least ten people sick with E. coli after visiting the petting zoo at Willow Grove Gardens Pumpkin Patch in Longview, Washington; both O157:H7 and O121 were isolated from patients. In 2013, three children between 15 months and seven years old were infected with O157:H7 after contact with goats or cattle at Dehn’s Pumpkins in Dayton, Minnesota, and one developed hemolytic uremic syndrome. The Minnesota Department of Health’s annual summary adds that goats and cattle were on site and that multiple animal fecal and environmental samples there tested positive for Shiga toxin 2. The Compendium cites a British Columbia pumpkin patch outbreak where there was no running water and the sanitizer was mounted too high for some children to reach. And this month, the first thing Bengtson’s in Illinois closed was its petting zoo.

I have not found a U.S. outbreak where the pumpkins themselves were the vehicle. That does not mean it cannot happen. Produce picks up Shiga toxin-producing E. coli in the field all the time; in the 2006 spinach outbreak, investigators found the outbreak strain in cattle manure and water on a San Benito County ranch within a mile of the implicated fields, a couple of hours up the road from San Luis Obispo. A pumpkin is not eaten raw the way spinach is, but it is handled, carved and scooped out by children with their bare hands. If the Brookshire investigation lands on the pumpkins, it would be something new, and the county should say so plainly and publish how it got there.

In yesterday’s post I grouped Brookshire with Bengtson’s and the State Fair of Texas as investigations tied to animals at fall attractions, and in Tuesday’s I put the two pumpkin farms side by side. The county’s pumpkin guidance suggests Brookshire may not belong in that group, and I would rather say so now than wait. What the three share is the season: tens of thousands of small children in close contact with soil, hay, produce and sometimes animals at the tail end of the peak months for these infections.

What the county should tell the public now.

Is this E. coli O157:H7 or a non-O157 strain, and have the five patient isolates been sequenced and found to match? Has any patient developed hemolytic uremic syndrome? What has been sampled: pumpkins, soil, irrigation water, hay, compost, surfaces, and any animals on or next to the property? Were the pumpkins grown on the farm or brought in, and if brought in, from where? That last question matters, because if the pumpkins came from somewhere else, the problem may not stop at Brookshire’s gate. The county should also ask the farm to reach visitors through its ticketing and card records, post a weekly update with case counts, ages and hospitalizations, and, when the investigation is done, put the whole file online so the next farm operator can learn from it.

What parents should do.

If your family visited Brookshire Farms between October 16 and 26, or has been handling a pumpkin bought there, watch for diarrhea, especially bloody diarrhea, and severe stomach cramps. Symptoms usually appear three to five days after exposure but can take up to ten. Hemolytic uremic syndrome can follow a week or more after the diarrhea starts; decreased urination, unusual paleness and lethargy are the warning signs. Tell the doctor about the farm visit and ask for a stool culture. Do not give anti-diarrheal medicine, and ask before using antibiotics. The consumer version of this post is on Food Poison Journal; prevention material for families and farm operators is at Fair-Safety.com; and what HUS is, and why it follows E. coli in young children, is at about-hus.com.

How many people are sick in the Brookshire Farms E. coli outbreak? Five confirmed as of October 27, 2026: two adults and three children, all of whom visited the farm. The three children are in serious condition in pediatric intensive care.

Has the source been found? No. On October 28 the county issued precautionary guidance on Brookshire pumpkins, but it has not named a source.

Is it a petting zoo outbreak? Nothing the county has released points to animals. The farm’s listed attractions are hayrides, rides, play areas and the corn maze, though one directory lists farm animals.

Have pumpkins caused an E. coli outbreak before? Not that I can find. Past pumpkin patch outbreaks, including Willow Grove in Washington in 2012 and Dehn’s in Minnesota in 2013, were tied to petting zoo animals, not the pumpkins.

Is Brookshire Farms open? No. The farm closed voluntarily on October 27 while the investigation continues.

What should I do with a pumpkin from Brookshire Farms? The county says to throw away carved pumpkins, scrub intact pumpkins that will be handled often with hot, soapy water, and wash your hands after handling them.

Tip of the Pen to the Healthy People Food Safety Workgroup: Forty Years of Scorekeeping Show Salmonella Has Not Budged Since 1997, Three of Four Pathogens Are Going the Wrong Way, and We Stopped Measuring the Trend. Here Are Ten Things That Would Move the Needle.

By Bill Marler on October 28, 2026

Posted in Case News

The Journal of Food Protection just posted an open-access article that every food safety regulator, industry executive and member of Congress should read, and then reread Table 1. “Tracking Illnesses Caused by Four Foodborne Pathogens Over Four Decades of the Healthy People Initiative,” by Robin Pendley Louis, Daniel Weller, Hazel Shah, Kristal Southern, Doug Noveroske and Priya Kadam, walks through how the federal government has set and measured national targets for Campylobacter, Listeria monocytogenes, Shiga toxin-producing E. coli (STEC) and Salmonella from Healthy People 2000 through the most recent Healthy People 2030 data. Tip of the pen to the authors. It is careful, honest bookkeeping. It is also a report card, and the grades are not good.

What the report card says

In Healthy People 2000, the government set modest targets and met all four. In Healthy People 2010, it set out to cut each infection rate in half and met only one, E. coli O157:H7. In Healthy People 2020, it met none. And in Healthy People 2030, as of the most recent data:

Campylobacter went from 16.2 infections per 100,000 people to 20.6, against a target of 10.9. Getting worse.

STEC went from 4.6 to 5.7, against a target of 3.7. Getting worse.

Listeria went from 0.27 to 0.36, against a target of 0.22. Getting worse.

Salmonella went from 15.3 to 15.2, against a target of 11.5. No detectable change.

Now look at Salmonella across the whole table. The Healthy People 2010 baseline, set in 1997, was 13.6. The target was 6.8. By 2009 the rate was 15.0. The Healthy People 2020 baseline was 15.0, and the final number was 14.5. Today it is 15.2. We have gone from a goal of cutting Salmonella in half, to a goal of cutting it by a quarter, to a goal pegged to FSIS performance standards that simply carried the old target forward – and we are still at roughly 15 per 100,000. Thirty years. Different surveillance, different definitions, different targets. Same number.

The authors are right that comparisons across decades require caution. Baselines moved from single years to three-year averages. Healthy People 2020 began counting all STEC, not just O157. Healthy People 2030 added infections diagnosed by culture-independent diagnostic tests (CIDTs) and dropped travel-related cases. And the authors point to research showing that much of the recent increase tracks the growing use of CIDTs, which find infections that used to go undiagnosed. I take that seriously. But it cuts both ways. If better tests are finding illness that was always there, then the burden was always bigger than we were counting – which is a reason to work harder, not to set softer targets for 2040. And no change in lab methods explains thirty flat years of Salmonella.

We stopped measuring the trend

Here is the part of the article that should alarm everyone. FoodNet, the active surveillance network that has measured these targets since 1996 and covers about 16 percent of the country, went from eight pathogens to two on July 1, 2025, for lack of money. Only STEC and Salmonella are still required. The article says plainly that 2023 is the last year Campylobacter was measured for Healthy People.

To be precise about it, because people who know this system well have corrected me before: sick people are still reported. Cases of the dropped pathogens still flow through the national notifiable disease system, and Listeriathrough the Listeria Initiative. What ended is the measured trend – the ability to say whether anything we do is working. Campylobacter is the pathogen moving fastest in the wrong direction, and it is the one we stopped measuring. And Cyclospora was dropped from FoodNet a year before the country ran the largest parasitic foodborne outbreak in its history this summer, more than 18,000 laboratory-confirmed cases. You cannot hit a target you have stopped measuring. You can only stop being embarrassed by it.

What is at stake

The article cites the numbers. CDC’s most recent estimate is that seven major pathogens cause 9.9 million domestically acquired foodborne illnesses, 53,300 hospitalizations and 931 deaths a year. The Government Accountability Office found that between 2011 and 2022 more than 9,000 foodborne outbreaks were reported in all 50 states, at a cost of $75 billion a year, and that CDC, FDA and FSIS have made limited progress toward their own targets. And those numbers undercount. The case count stops the day the fever breaks. A ten-year-old on dialysis for nine days with hemolytic uremic syndrome does not show up in the death count, and the kidney damage that follows her for the next twenty years does not show up anywhere. These are not abstractions to me. They are my clients.

“The workgroup does not identify strategies.”

That sentence, from the article’s introduction, is the problem in seven words. Healthy People sets the targets. It leaves it to “federal agencies, communities, states, food industry, and organizations across the country” to decide what to do about them. A target without a strategy is a wish.

The article’s own history makes the point. The wins it credits – FSIS declaring E. coli O157:H7 an adulterant in ground beef in 1994, Pathogen Reduction/HACCP in 1996, juice HACCP in 2001, the Listeria rule for ready-to-eat meat and poultry in 2003, six more STEC serogroups declared adulterants in 2011, and the Food Safety Modernization Act – almost all came after outbreaks forced someone in power to act. And the only pathogen in four decades to meet a “cut it in half” target was E. coli O157:H7, the one the government made illegal to sell in ground beef and the industry was forced to test for. That is the lesson of Table 1: binding rules move the numbers. Aspirations do not.

The workgroup is now setting targets for Healthy People 2040. Here are ten things that would actually move the needle. Most of them come out of something I have been working on with friends – and a few people who are not so friendly – from consumer and victims’ groups, public health, state regulators, academia, labor and the food industry itself. We are calling it Project 2027: a working agenda of known problems with known solutions, things that would make American food safer and can be done by people who already have the jobs. The not-so-friendly readers have made it better. They have told me where I am wrong, and a few times they have been right.

1. Enforce the one Salmonella standard FSIS already has, then write the one it needs. In May 2024, FSIS declared raw breaded stuffed chicken with Salmonella at 1 CFU per gram or more adulterated. Testing was to start May 1, 2025. FSIS pushed it to November 3, then on December 2, 2025 delayed it indefinitely with no new date. The standard is law, and nobody is checking for it. Meanwhile, in April 2025 FSIS withdrew its Salmonella framework for the rest of raw poultry after three years of work. That matters to Healthy People directly: the 2030 Salmonella and Campylobacter targets were set to “align” with FSIS performance standards. Step one is to make FSIS publish its method data and set a testing date. Step two is my citizen petition asking FSIS to declare Salmonella at 1 CFU per gram an adulterant in raw ground chicken and turkey and raw chicken parts. I used to be in the beef business with E. coli, and what FSIS did in 1994 took that practice away from me. It is the best thing that has happened to food safety in my lifetime, and every argument being made against doing it for poultry was made and lost by the beef industry thirty years ago.

2. Measure again – the laboratories first. Protect PulseNet and the state public health laboratories that do the sequencing behind every multistate outbreak count; without them an adulterant standard nobody can sample for is a press release. Then restore or rebuild FoodNet for all eight pathogens, with modernization written into the ask rather than used as the excuse to let it go. Consumer and academic groups asked Congress last November for a 50 percent increase, about $36 million, for CDC’s food safety program. Ask again, and this time attach the summer to the request. And Healthy People 2040 should not lower a target because better tests find more of the illness that was always there.

3. Sample the food, not just the sick. From 2001 to 2012, USDA’s Microbiological Data Program tested about 15,000 produce samples a year for Salmonella, Listeria and E. coli, sent its positives to PulseNet and triggered 38 recalls in its last four years. It cost about $4.5 million a year, and it ended in 2012 when nobody asked for the money and Congress did not appropriate it. The produce industry lobbied against it because its findings caused recalls – which is another way of saying it worked. FDA’s own sampling assignments since then have been valuable, but sixteen hundred samples of one commodity every year or two is not a baseline. Two asks: publish, sample by sample, the testing FDA already pays the states to do through its laboratory cooperative agreements, which costs nothing; and fund dedicated produce sampling again. Surveillance of people tells you who is already sick. Sampling of product tells you what is coming.

4. Kill the traceability rider. Section 780 of the law that ended last November’s shutdown bars FDA from spending a dollar to enforce its own food traceability rule before July 2028 and nudges the agency to loosen lot-level tracking, the heart of the rule. I wrote in Food Safety News how that happened. Every day a traceback runs on invoices, phone calls and memory is a day contaminated food is still moving and people are still eating it. The fiscal 2027 appropriations bill should carry no version of that language.

5. Send an inspector when an outbreak points at a plant – and set a floor for the foods that keep making people sick. FSMA sets inspection frequency by risk category, every three years for high-risk facilities, and says nothing about the one event that most reliably tells you something went wrong at a particular plant: people got sick from its product. FDA routinely closes outbreaks without ever walking the facility, which means it never learns the root cause. Require a for-cause inspection within a set period once an outbreak is linked to a firm, with the findings made public. Then do for leafy greens, sprouts, cantaloupe, cucumbers, onions and frozen berries what Congress did for infant formula in 2022: an annual inspection floor, reaching foreign facilities too. That last part matters. FSMA set a goal of 19,200 foreign food inspections a year; last year FDA did 1,024, and as The Washington Post just showed, it has inspected about 2 percent of the Mexican facilities that ship us produce. The Listeria line in Table 1 needs the same discipline on the meat side: when a plant racks up documented noncompliance, as the Boar’s Head plant did before ten people died in 2024, the next step should be suspension, not another report. (In the interest of full disclosure, I represent people sickened in that outbreak.)

6. Let FDA sample the feedlot next door. STEC is rising, and leafy greens keep telling us why. In the 2018 Yuma romaine outbreak, the outbreak strain was found in a canal running past a large cattle operation, and FDA had no authority to set foot on it. The investigation stopped at a fence. The Expanded Food Safety Investigation Act, S. 376 and H.R. 852, would let FDA sample animal feeding operations during an outbreak investigation; it has never had a hearing. My citizen petition asks FDA to do the same thing by rule under section 361 of the Public Health Service Act, the authority it used in 1975 to reach pet turtles. A root cause nobody is allowed to find is a root cause nobody fixes. And this fall’s run of petting zoo outbreaks is the same bacteria without the lettuce.

7. Name the food, the firm and the stores – while it matters. Every day an outbreak food sits in kitchens unnamed adds cases to the Healthy People numerator. On Monday, the Justice Department’s Office of Legal Counsel told HHS that FDA’s own regulations, not the law, keep outbreak supply chain information secret, and that FDA can change them. My two petitions – one on naming firms implicated in outbreaks, one on publishing retail consignee lists in Class I recalls, as USDA has done since 2008 – will go in together.

8. Count what happens afterward. Build a national registry of long-term outcomes from foodborne illness – kidney damage after HUS, reactive arthritis, Guillain-Barré, the rest – seeded through the FoodNet sites and the pediatric HUS surveillance that already runs. USDA’s own cost model admits the long-term data are thin. If Healthy People 2040 measured severity and sequelae and not just case counts, the targets would point at what actually costs this country money and children their kidneys. This is research money, the kind an appropriator can put in report language.

9. Put a name next to every target. Federal food safety oversight has been on GAO’s High-Risk List since 2007. In 2017, GAO recommended a national food safety strategy to deal with the fragmentation among the fifteen agencies and thirty statutes involved. In January 2025, OMB told GAO it had no plans to write one. Healthy People is the closest thing we have to a national scorecard, and by design nobody owns the score. For each pathogen target in 2040, FDA, FSIS and CDC should publish what they are doing to hit it, report against it every year and bring it to their appropriations hearings – and the White House should finally write the strategy GAO asked for nine years ago.

10. Tell people the truth where they buy the food. Some risks will not be regulated away in the next decade, so consumers deserve to know them at the point of sale. My petitions asking FDA to require warning statements on raw milk cheese and raw sprouts – foods that keep turning up in STEC and Salmonellaoutbreaks – ban nothing. They put one sentence on the label.

And an eleventh, for the workgroup itself. Healthy People tracks four bacteria. The biggest number in foodborne illness is not one of them. CDC puts norovirus at 5.5 million domestically acquired foodborne illnesses a year, more than any other pathogen, and the mechanism most of the time is a food worker who came in sick because staying home cost a shift. Put norovirus in Healthy People 2040, and the Food Code’s employee health rules and paid sick leave for food workers will finally have a number attached to them.

None of this works without money.

Every item above assumes someone is there to do it. An adulterant standard nobody can sample for is a press release. A traceability rule with no investigators behind it is a filing cabinet. Congress has to fund the people who do this work, and fund them adequately: FDA’s food program, so it can inspect the plants and foreign facilities it is already supposed to inspect; FSIS and its roughly eight thousand frontline inspectors; CDC’s food safety program, PulseNet and the surveillance that tells us whether any of this is working; and the state and local health departments, public health laboratories and epidemiologists who take the first call, interview the sick, sequence the isolates and find most outbreaks before Washington hears about them. The government spends a small fraction on prevention of what foodborne illness costs this country every year. Cutting it is not saving money. It is moving the bill to hospitals, families and the businesses whose products end up on a recall list.

The petitions

Seven of the asks above do not have to wait for Congress, an appropriation or a new administration, so I am putting them in front of the agencies myself as citizen petitions. To FSIS: declare Salmonella at 1 CFU per gram an adulterant in raw ground chicken and turkey and raw chicken parts, and start the breaded stuffed chicken testing it already promised. To FDA: write a rule letting investigators sample the animal feeding operation implicated in an outbreak; name the firms implicated in outbreaks; publish the stores, restaurants and schools that received food in a Class I recall; put a warning statement on raw milk cheese; put one on raw sprouts; and, with forty-six parents of infants hospitalized in the ByHeart and Nara Organics outbreaks, add Clostridium botulinum to the required testing of powdered infant formula and the dairy ingredients it is made from. That last one is not one of Healthy People’s four pathogens. It should not have to be. If you would like to see the drafts and tell me what I got wrong, email me.

The bottom line

The Healthy People Food Safety Workgroup has done what it was asked to do. It set targets, measured honestly, and published the results, even when the results are unflattering. That is more than can be said for a lot of government. But forty years of data now say the same thing: the targets we met were the ones backed by binding rules, and the ones we keep missing are the ones we left to hope. Healthy People 2040 is being written now. If we want a different report card in 2040, we need different rules in 2027 – and we need to keep measuring.

Read the full article. It is short, and Table 1 is worth the price of admission, which is free. If you want to help with Project 2027, friendly or not, or want to tell me what I left off this list, email me at bmarler@marlerclark.com.

Outbreaks are preventable. But you have to decide to prevent them.

Pumpkin Patches, Petting Zoos and State Fairs: Three New E. coli Investigations, Twenty-Five Years of Outbreaks Just Like Them, and the Prevention Playbook Everyone Already Has

By Bill Marler on October 28, 2026

Posted in Case News

In the space of five days this October, three health departments opened investigations into Shiga toxin-producing E. coli (STEC) infections tied to animals at fall attractions: five confirmed cases, four of them hospitalized, at Bengtson’s Pumpkin Farm in Homer Glen, Illinois; five confirmed cases at Brookshire Farmsin San Luis Obispo, California, which has closed; and a cluster of sick children Dallas County is investigating in connection with the petting zoo at the State Fair of Texas. None of this is new. What follows is a short history of how we got here, and the prevention measures that have been written down, revised and re-published for twenty years.

This happens every year, and the numbers are not small.

The Compendium of Measures to Prevent Disease Associated with Animals in Public Settings has been published by the National Association of State Public Health Veterinarians (NASPHV) since 2005. The 2023 revision is summarized in the December 2023 issue of JAVMA, and the full version counts 156 outbreaks involving animals in public settings in the ten years ending in 2020. Those outbreaks caused 6,388 illnesses, 1,233 hospitalizations and 10 deaths. The full version also notes that outbreaks are only part of it: in a Minnesota study of non-outbreak illnesses linked to animal agriculture, 28 percent of the STEC patients had visited a fair, petting zoo, agritourism farm or similar venue, and six developed HUS.

I started tracking these outbreaks on Fair-Safety.com years ago. The outbreak list there runs back to the early 1990s, but it stops in 2015 and still points readers to the 2017 Compendium. It is overdue for an update, and this post is the start of it. Here is a selection of the U.S. E. coli outbreaks, old and new:

Year	Venue	State	What happened
2001	Wyandot County Fair	Ohio	92 E. coli O157 cases tied to the fair.
2002	Lane County Fair	Oregon	82 ill; the largest E. coli O157:H7 outbreak in Oregon history.
2003	Fort Bend County Fair	Texas	25 visitors and animal exhibitors sickened.
2004	North Carolina State Fair petting zoo	N.C.	108 cases; 15 children developed HUS. The outbreak strain was found in shavings on a stroller and on visitors’ shoes.
2005	Florida State Fair, Central Florida Fair, Strawberry Festival	Florida	63 sick after goat contact at petting zoos run by one vendor; 7 developed HUS.
2005	Big Fresno Fair petting zoo	Calif.	At least six children; a two-year-old suffered HUS, kidney failure and strokes. The petting zoo relied on hand sanitizer.
2007	Day camp petting zoo	Florida	Seven children infected; a goat was implicated.
2011	North Carolina State Fair	N.C.	At least 27 sickened from eight counties.
2012	Cleveland County Fair	N.C.	106 people reported ill.
2012	Willow Grove pumpkin patch	Wash.	Four children sickened after a pumpkin patch and corn maze visit.
2013	Dehn’s Pumpkins	Minn.	Three children after contact with goats and cattle; one developed HUS.
2014	Traveling petting zoo, county fairs	Minn.	More than a dozen sickened by one traveling petting zoo that worked three county fairs and a festival.
2015	Oxford County Fair	Maine	Two toddlers with the same rare E. coli O111 strain; 20-month-old Colton Guay died of HUS.
2016	Goat dairy farm, kidding season	Conn.	50 confirmed cases, including secondary spread; three developed HUS. Goats and the farm environment were widely positive.
2019	San Diego County Fair	Calif.	11 confirmed and 2 probable cases; three hospitalized; a two-year-old boy died. Animal and environmental tests were all negative.
2019	Minnesota State Fair, Miracle of Birth Center	Minn.	At least 11 sick, ages 2 to 43; six hospitalized, one HUS. Some never touched an animal.
2022	Farming day camp	Tenn.	Campers sickened; farm isolates matched patients by sequencing, and the farm euthanized two STEC-positive goat kids.
2023	Appalachian Fairgrounds school field trips	Tenn.	64 cases; seven hospitalized; four with HUS. Five of the nine lab-confirmed cases were household members of children who went.
2025	Arizona State Fair petting zoo	Ariz.	18 sick, median age 10; 11 hospitalized; six with HUS. No soap-and-water handwashing near the exhibit.
2026	Bengtson’s Pumpkin Farm	Ill.	Five confirmed, four hospitalized; petting zoo closed; investigation open.
2026	Brookshire Farms	Calif.	Five confirmed; whole farm closed voluntarily; investigation open.
2026	State Fair of Texas petting zoo	Texas	A “cluster” of children under investigation; at least two hospitalized.

A selection, not a complete list. Sources are linked in each row: Fair-Safety.com outbreak summaries, CDC MMWR, state and county health departments, OutbreakDatabase.com and news reports. Counts are as last reported by the investigating agency.

The list could be much longer. Fair-Safety.com also documents outbreaks of Cryptosporidium, Campylobacterand Salmonella at farms, fairs and zoos here and abroad, and Food Poison Journal has a longer history of petting zoo and fair outbreaks.

What the outbreaks have in common.

Read twenty-five years of these investigations and the same facts keep showing up. The Compendium pulls them together, and they explain this October as well as they explain 2004.

Healthy-looking animals carry it. Cattle, goats and sheep are the usual sources. They carry STEC without getting sick and shed it intermittently, and young animals – the calves, kids and lambs that draw children – shed more than adults. Stress from transport, crowding and handling increases shedding.

You do not have to touch an animal. Fur, saliva, railings, bedding, sawdust, hay bales, manure on a fence, dust, a stroller wheel, a shoe: all have carried the bacteria in past outbreaks. STEC can survive in soil for months. Minnesota’s 2019 State Fair cases included people who never touched an animal, and Connecticut’s 2016 goat farm outbreak spread to people who had only been near someone who visited.

Testing the animals afterward rarely settles it. Because shedding comes and goes, the Compendium warns that a negative test does not mean the animal was not shedding earlier. In San Diego in 2019, every petting zoo animal, every pony and 32 environmental samples tested negative – after a child had died. The Compendium does not recommend routine screening for this reason. Prevention has to assume the animals are positive.

Temporary exhibits are the riskiest. The Compendium says temporary and seasonal exhibits are particularly vulnerable to design flaws, and so are working farms that were never designed for public events. The full version gives examples that read like this month’s news – animal displays added to festivals, farm stands, pick-your-own farms and Christmas tree lots – and lists the usual problems: inadequate barriers, surfaces that cannot be cleaned, insufficient plumbing, missing signs and too few handwashing stations. The traveling vendor that sets up for a weekend is exactly the operation least likely to have running water at the exit.

Hand sanitizer is not handwashing. In outbreak after outbreak, the Compendium says, handwashing has been protective: lathering with soap and water lowered the risk, while feeding animals without washing afterward and drying hands on clothing raised it. The full version adds that soap and water is still preferred over sanitizer, which does not work equally well on all pathogens or on dirty hands. The 2005 Fresno petting zoo had alcohol sanitizer. So, twenty years later, did the 2025 Arizona State Fair.

Food and animals do not mix. Contamination of food and drinks by people who have just touched animals has caused outbreaks, and the full Compendium adds that buying food at a farm visit and eating sticky foods like ice cream and cotton candy have been associated with STEC illness. Feeding animals and then eating without washing raises the risk. Some petting zoos still sell feed in ice cream cones, which the Compendium specifically says not to do.

Young children are the most at risk. Children under five put their hands in their mouths, sit on the ground and get close to the animals, and they are the most likely to develop HUS. The 2023 Compendium says preweaned calves, lambs, goat kids, reptiles and live poultry are not appropriate where groups at high risk are expected – and on a fall weekend at a pumpkin patch, or a school field-trip day at a fair, that is a lot of the crowd. Add that young animals shed STEC more than adults, and it is no surprise that so many of the rows in that table end with a toddler on dialysis. More on HUS is at about-hus.com.

How to prevent them – the Compendium’s playbook.

None of this is a secret. The Compendium is free, it has been around in some form since 2005, and the 2023 edition is posted by CDC and published by the AVMA. Washington’s Department of Health has turned it into a practical guide for fairs. The core recommendations for operators:

Tell visitors the risk before they go in. The Compendium calls educating visitors “of utmost importance”: awareness of the risk was protective in the 2004 North Carolina State Fair outbreak. That means signs, staff at the entrance, and information sent home before school field trips.
Divide the venue into three zones. Animal areas, transition areas at every entrance and exit, and non-animal areas where food is sold and eaten. No animals except service animals in non-animal areas.
Put soap, running water and paper towels at every exit. Enough sinks for peak crowds, at a height children can reach, and as close to the exit as possible – not across the midway.
Staff the exit. A staff member at the exit reminding visitors to wash is one of the interventions the Compendium cites as improving handwashing.
Keep food, drinks, bottles, pacifiers, sippy cups and toys out of animal areas. Provide stroller parking at the entrance.
Control feeding. Allow it only where contact is controlled, never sell feed in edible containers, and keep feed and water bins away from the public.
Keep the youngest children away from the riskiest animals. Children under five should not have direct contact with preweaned calves and similar high-risk animals, and should be supervised at all times around the rest.
Design for one-way traffic and avoid overcrowding. Use double barriers where contact is not intended, remove manure promptly and keep manure tools and piles out of public reach.
Use a veterinarian and keep records. Exhibit only healthy animals, keep sick animals and herds with recent diarrhea out, and keep records of where animals came from and where they have been, so investigators can trace them.
Do not sell raw milk or unpasteurized cider. The full Compendium says their sale or distribution should be prohibited.

For parents and visitors, the short version: wash with soap and water as soon as you leave the animal area and before eating, even if you think you did not touch anything; do not eat, drink or let children use pacifiers or bottles in animal areas; do not let small children sit or play on the ground there; wash after handling strollers and shoes that went into the pens; and if a child develops bloody diarrhea within ten days of a farm or fair visit, see a doctor and tell them about the visit. Antibiotics and anti-diarrheal drugs are generally not recommended for suspected STEC in children.

Recommendations are not requirements.

Here is the uncomfortable part. As the full Compendium acknowledges, no U.S. federal law addresses the risk of disease transmission where animals and the public mix, and only some states regulate things like handwashing stations. The Compendium tells agencies to work to ensure compliance and to build its recommendations into regulations, and the full version notes that studies keep finding the recommendations are not fully followed. After Arizona’s outbreak, a bill requiring handwashing, signage and supervision at petting zoos died in the legislature; the Arizona State Fair and the Maricopa County Fair dropped direct animal contact on their own. Voluntary decisions by a few fairs are not a standard.

Every state should turn the Compendium into a rule: soap and running water at every animal-area exit, staff at the exit, no food in animal areas, no feed in cones, records of animal sources, and a public notice requirement when a cluster is identified. Every health department that opens one of these investigations should publish the case count, ages, hospitalizations and HUS cases on a schedule, sequence every isolate, and put the final report online, so the next operator can learn from it. Twenty-five years of outbreaks is enough data.

Frequently asked questions

Can you get E. coli at a petting zoo without touching an animal? Yes. Outbreaks have been traced to railings, bedding, sawdust, hay bales, dust and contaminated shoes and strollers, and STEC can survive in soil for months.

Is hand sanitizer enough after a petting zoo? No. The full 2023 Compendium says soap and running water is preferred, because sanitizers do not work well on dirty hands or equally well against all pathogens.

Who is most at risk? Children under five, adults over 65, pregnant people and people with weakened immune systems. Young children are the most likely to develop hemolytic uremic syndrome.

Are petting zoos regulated? There is no federal law addressing disease risk at animal exhibits. Some states require handwashing stations; most rely on voluntary guidance like the Compendium.

Two Pumpkin Farms and the State Fair of Texas: Three E. coli Investigations Tied to Animal Attractions in Five Days, in Three States. Here We Go Again.

By Bill Marler on October 27, 2026

Posted in Case News

On Tuesday afternoon, October 27, the County of San Luis Obispo Department of Public Health announcedfive confirmed cases of Shiga toxin-producing E. coli (STEC) and an investigation into a possible link to Brookshire Farms, the pumpkin patch, corn maze and Christmas tree farm on Los Osos Valley Road. The farm has closed voluntarily. Tuesday night, FOX 4 in Dallas reported that Dallas County Health and Human Services is lab-testing a cluster of STEC infections in children and investigating whether they came from the petting zoo at the State Fair of Texas. Add the five confirmed cases at Bengtson’s Pumpkin Farm in Homer Glen, Illinois, four of them hospitalized, and that is three investigations, in three states, tied to animal attractions at fall events – all made public in the five days since October 23. Every October I write some version of this post. This year I am writing it three times over.

What San Luis Obispo County has said.

Five confirmed STEC infections. A suspected exposure period at Brookshire Farms that was first announced as October 16 through 18 and then, in a 4:00 p.m. update, widened to October 16 through 26 – the entire last ten days the farm was open. The source of the infections is under investigation. The county is telling anyone who visited during that window and develops cramping and bloody diarrhea, typically three to ten days after the visit, to call their doctor, and it notes that children under 10 face the greatest risk of complications. County Health Officer Dr. Christy Mulkerin urged anyone exposed and symptomatic to seek medical attention immediately. The farm is cooperating, and, according to KSBY, its owner has said a direct link to the farm has not been confirmed.

One housekeeping note for the county: the updated release widens the window to October 26, but the subheadline and the Spanish-language version still say October 16 through 18. A family that reads the Spanish release, visited on October 22 and has a child with bloody diarrhea today may conclude they are outside the window. Fix it.

What the county has not said.

Ages. Hospitalizations. Whether anyone has hemolytic uremic syndrome (HUS). The serotype – O157 or a non-O157 STEC. Whether the five isolates have been sequenced and match. Most importantly, what at the farm is suspected. A pumpkin patch is not a single exposure. According to the farm’s promotional listings, Brookshire offers hayrides, grain-cart rides, pedal carts, a jumping pillow, duck races, a playground, a hay climb, a hay maze and a corn maze, and a U-pick directory lists farm animals, school tours, birthday parties, fresh eggs and a picnic area where families can eat their own food. Was there animal contact? Where were the handwashing stations relative to the animals and the kettle corn? Were school groups on the property between October 16 and 26? Those answers tell parents and physicians where to look, and they cost the county nothing to release.

The State Fair of Texas: a “cluster,” children in the hospital, and a fair that says it has not been notified.

According to FOX 4, Dallas County Health and Human Services confirmed it is lab-testing a cluster of Shiga toxin-producing E. coli cases affecting children of a range of ages, is still interviewing families, and is investigating whether the State Fair petting zoo is the source. The county is not yet calling it an outbreak. At least two families told the station their children are hospitalized in Dallas, and Children’s Medical Center Dallas says it has seen an uptick in cases. The county has not released a case count, ages, hospitalizations, whether any child has HUS, or the serotype.

The fair’s own website describes the petting zoo in Little Lone Stars Land as a place to get close to emus, zebras, pigs and goats, with a nursery of sows, ewes, nanny goats and their young – and it advertises that fairgoers can feed the animals. The fair’s statement to FOX 4 says it has not received official notification, that it follows CDC-recommended precautions at its animal exhibits, and that no precaution can eliminate every risk. That last line is true. It is also why the precautions have to be the right ones – soap and running water at the exit, no food or bottles in the pen, someone watching the children – and why the fair should be asked now, in writing, what exactly was in place.

This is not the State Fair of Texas’s first brush with this. In the mid-2000s, after a run of petting zoo outbreaks around the country, the fair scrapped its Children’s Barnyard petting zoo when insurers raised rates; at the time, that petting zoo relied on hand-sanitizing gel and signs. Two decades later, the Arizona State Fair’s petting zoo outbreak taught the same lesson about sanitizer, and the Arizona State Fair has dropped direct animal contact for 2026. The State Fair of Texas did not.

Dallas County should do what I am asking of San Luis Obispo and Will counties: publish a case count, ages, hospitalizations and any HUS; issue a health alert to North Texas pediatricians and emergency departments; open a survey for fairgoers; sample the animals and the environment before the animals are gone; and sequence everything. The fair runs for weeks and draws enormous crowds, and once it closes the petting zoo and livestock animals scatter to their home farms. Every day of delay is a day of lost samples.

Brookshire closed the whole farm. Bengtson’s closed only the petting zoo. The State Fair has closed nothing.

The responses are worth comparing. In Illinois, Bengtson’s closed its petting zoo on October 23 and kept the rides, food stands and corn maze open with the health department’s blessing. Animal and environmental samples were taken on October 26, results are expected around November 5, and, as I wrote this morning, more than 200 visitors have filled out the county’s symptom survey. In California, Brookshire shut down entirely, in the last week of pumpkin season, while the county figures out what happened. That is a costly decision for a family farm, and it is the right one. It does not tell us whether the source is an animal, the environment or something sold on site. It does mean no one else is being exposed while the county finds out. In Texas, nothing has closed, and the fair says it has not even been officially notified.

	Brookshire Farms (San Luis Obispo, CA)	Bengtson’s Pumpkin Farm (Homer Glen, IL)	State Fair of Texas (Dallas, TX)
Made public	October 27 – SLO County Public Health	October 23 – Will County Health Department	October 27 – Dallas County HHS, via FOX 4
Confirmed STEC cases	Five	Five (as of October 26)	A “cluster”; no count released
Ages / hospitalized / HUS	Not released	Ages 4 to 38; four hospitalized; HUS not reported	Children of various ages; at least two hospitalized per families; HUS not reported
Exposure window	October 16–26 (revised from October 16–18)	On or after September 29	Not released
What closed	The whole farm, voluntarily	Petting zoo only	Nothing reported; fair says it has not been officially notified
Suspected vehicle	Not stated	Petting zoo under investigation; food-service inspection clean	Petting zoo under investigation
Animal / environmental testing	Not stated	Sampled October 26; results expected about November 5	Not stated

Sources: SLO County Public Health release (October 27, 2026, as updated 4:00 p.m.); KSBY (October 27); FOX 4 Dallas-Fort Worth (October 27); Will County Health Department statements as reported by The Herald-News (October 26). Read October 27.

These are not food outbreaks, and food inspections will not solve them.

When STEC follows a farm visit, the vehicle is rarely a menu item. It is manure – on a goat’s coat, a pen rail, a hay bale, a stroller wheel, a shoe, the hand that then holds a cup of cider. Ruminants carry these bacteria without looking sick and shed them intermittently, which is why the Compendium of Measures to Prevent Disease Associated with Animals in Public Settings warns that a negative animal sample taken days after the fact does not clear the pen. The record goes back decades: the 2004 North Carolina State Fair, a Tennessee farming camp in 2022, the Willow Grove pumpkin patch in Longview, Washington, and last fall’s Arizona State Fair petting zoo outbreak, which sickened 18 people, hospitalized 11 and left six with HUS. The fixes have not changed either: soap and running water, not hand sanitizer, at every exit from an animal area; animal areas kept away from where people eat; no food, bottles or pacifiers near the animals; and someone watching to make sure small children actually wash.

Are the three connected?

Nothing released so far suggests they are. Three venues in California, Illinois and Texas, in the same few weeks of the busiest season of the year, is what the calendar looks like when hundreds of thousands of families with young children get close to livestock in October. Whole genome sequencing will answer the question, and all three counties should say whether their isolates have been uploaded to PulseNet.

What the counties should do now.

San Luis Obispo specifically: correct the exposure dates in every version of the release. Publish ages, hospitalizations and any HUS, and update them on a set schedule. Say whether the farm had animal contact and whether animals and surfaces have been sampled. Open a visitor survey the way Will County did, and push it to the schools that brought classes between October 16 and 26. Alert local pediatricians and emergency departments that bloody diarrhea in a child who visited the farm is STEC until proven otherwise – and that antibiotics and anti-diarrheals are a bad idea in that child. Sequence the isolates and publish the results, whatever they show.

For families: if your child visited Brookshire Farms between October 16 and 26, Bengtson’s on or after September 29, or the State Fair of Texas this fall, and develops bloody diarrhea, severe cramps, decreased urination, unusual paleness or lethargy, see a doctor and tell them about the farm visit. HUS can develop a week or more after the diarrhea starts. Background on E. coli, HUS and staying safe around farm animals is available.

Frequently asked questions

How many people are sick in the Brookshire Farms E. coli outbreak? Five confirmed STEC cases as of October 27, 2026. The county has not released ages, hospitalizations or HUS cases.

What is the exposure period? October 16 through 26, 2026, according to the county’s 4:00 p.m. update. Some versions of the release still show October 16 through 18.

Is Brookshire Farms open? No. The farm closed voluntarily while San Luis Obispo County Public Health investigates.

Has the source been found? No. The county says the source is under investigation and has not said what at the farm is suspected.

Is the State Fair of Texas petting zoo linked to E. coli? Dallas County Health and Human Services is investigating whether a cluster of STEC cases in children came from the petting zoo. No case count or confirmation of a source has been released, and the county is not yet calling it an outbreak.

Are the Brookshire, Bengtson’s and State Fair of Texas cases connected? There is no indication they are. Whole genome sequencing will show whether any of the isolates match.

Tip of the Pen to The Washington Post: FDA Has Inspected Only About 2 Percent of Mexican Produce Facilities Since 2009. Congress Sent Letters. The Post Did the Oversight.

By Bill Marler on October 27, 2026

Posted in Case News

Some reporting tells you something you suspected. Some reporting proves it with numbers. “The FDA rarely inspects much of the produce Americans eat,” published this morning by The Washington Post, does the second. Tip of the pen to reporters Caitlin Gilbert and Rachel Roubein (with Diana Baptista contributing), who, for the first time, combined FDA fresh-produce import data with FDA inspection data to see who is actually being checked.

The answer: almost no one.

What the Post found

Of the foreign facilities that account for 90 percent of fresh produce shipments to the U.S., only 7.6 percent have been inspected by FDA since 2009.
In Mexico, the top supplier by far, FDA has inspected only about 2 percent of the facilities that ship produce here.
Mexico supplies 96 percent of imported berries. There are 3,321 major berry suppliers, and FDA has conducted 28 inspections of them since 2009. Peppers: 398 major suppliers, 57 inspections. Fresh herbs: 198 suppliers, 14 inspections.
The Food Safety Modernization Act, signed in 2011, set a goal of 19,200 foreign food inspections a year. Last year FDA completed 1,024, about 5 percent of the target.
Imports now make up 34 percent of fresh vegetables and 59 percent of fresh fruit Americans eat, up from roughly 17 and 46 percent two decades ago. Produce imports from Mexico have grown more than fourfold, to $16.7 billion last year.

Why it matters: look at this summer

The Post ties the data to the outbreaks we have been living with. Iceberg lettuce from Taylor Farms de Mexico was linked to the nation’s largest known cyclosporiasis outbreak: 12,883 people sick in 21 states, at least 570 hospitalized, and two deaths in Michigan. FDA had last inspected Taylor Farms’ Mexico operations seven years earlier. When investigators did go, they found Cyclospora in a wastewater tank and a drainage ditch.

In the Salmonella outbreak linked to jalapeños, which sickened nearly 500 people, FDA never sent a team to the grower it named as the “likely source.” The farm was in a region known for cartel activity. Security risk is a real problem. It is also a real gap in our food safety system.

FDA’s acting deputy commissioner for food, Donald Prater, told the Post that “inspections are not a good proxy for our overall oversight activities,” and pointed to border testing, importer record reviews and artificial intelligence. But the Post notes that only a sliver of imported food is ever physically examined at the border. Each foreign inspection costs about $38,700. The agency is asking for $9 million more for foreign inspections, plus authority to collect fees from foreign facilities that could bring in about $71 million.

The Post did the oversight work Congress has not

As cyclosporiasis cases ballooned this summer, I told the Post, I expected a congressional uproar, with hearings and lawmakers asking tough questions. This is a Jack in the Box–level event. Instead, outside of some letters to federal agencies and to Taylor Farms, “it’s been pretty silent.” That is still bizarrely shocking to me.

Letters are not oversight. Oversight is asking how many foreign produce facilities FDA has actually inspected, matching that against where our food comes from, and putting the answer in front of the public. That is what Gilbert and Roubein just did. Two reporters with FDA’s own import and inspection data produced the analysis that a congressional committee, with staff, subpoena power and the purse strings, should have demanded months ago. The Washington Post is doing Congress’s job.

Now Congress has no excuse. Senate Appropriations Chair Susan Collins’ office calls the outbreaks “troubling.” Rep. Rosa DeLauro calls food safety one of her “highest priorities.” Good. The data is sitting in the Post. Jack in the Box in 1993 led to E. coli O157:H7 being declared an adulterant in ground beef. Spinach in 2006 and peanut butter in 2008 helped push FSMA through. Outbreaks have moved policy before, but only when someone in power picked up the ball. Here is where Congress should start:

Fund FDA to actually meet the foreign inspection mandate Congress set in 2011, starting with the highest-risk produce and countries.
Give FDA the foreign facility fee authority it is asking for.
Hold importers accountable under the Foreign Supplier Verification Program, and stop treating paper third-party audits as a substitute for boots on the ground.
Hold real oversight hearings, not just send letters, with FDA and the companies tied to this summer’s outbreaks at the witness table.

Read the Post’s full story, charts and methodology included. It is the most important produce safety reporting of the year.

Outbreaks are preventable. But not if no one is looking, and right now the people looking hardest work for a newspaper.

Bengtson’s Pumpkin Farm E. coli Outbreak: Four of Five Sick Were Hospitalized, More Than 200 Visitors Have Reported Symptoms, and the Animal Samples Went to the Lab Monday. We Are Investigating for One of the Families.

By Bill Marler on October 27, 2026

Posted in Case News

As of October 26, 2026, the Will County Health Department counts five laboratory-confirmed Shiga toxin-producing E. coli infections possibly linked to visits to Bengtson’s Pumpkin Farm in Homer Glen, Illinois, on or after September 29. Four of the five were hospitalized, the patients are between four and 38 years old, no deaths have been reported, and more than 200 visitors have filled out the county’s symptom survey. The petting zoo is closed; the rest of the farm is open. Marler Clark has spoken with the family of one of the children who was sickened and is investigating.

We are investigating on behalf of one of the families.

I will say this once and briefly. I have spoken with the family of one of the children sickened in this outbreak, and the firm is investigating. I am not going to describe the child, the family or the illness, now or later, beyond what the family chooses to make public. I have represented children sickened at pumpkin patches, petting zoos and fairs for a long time, including at a pumpkin patch in Longview, Washington, and the pattern of these outbreaks is familiar enough that the questions worth asking are already known. I listed them on Friday. What follows is what has changed since.

Four of the five people sickened were hospitalized, and that was not in Friday’s announcement.

When the county announced the investigation on October 23, it gave five lab-confirmed cases and nothing about ages or hospital care. The Herald-News reported Monday that four of the five were hospitalized, that at least three have been discharged, and that the county does not know the status of the fourth. The ages run from four to 38, and the patients do not all live in Will County, which communications manager Kevin Juday said is part of why the cases took time to surface and to be linked. Four hospitalizations out of five confirmed cases is a severe ratio for Shiga toxin-producing E. coli. It can mean a virulent strain, or it can mean that only the sickest people have been tested so far. Either way it belongs in the public notice, because it tells parents how seriously to take a child’s bloody diarrhea this week.

More than 200 survey responses means five is a floor, not a count.

The county asked anyone who visited on or after September 29 and became ill to complete a confidential online survey. More than 200 people have. Most of them will not turn out to be E. coli, and a survey is not a stool culture. But a petting zoo at a popular pumpkin farm in the Chicago suburbs during the busiest four weeks of its year sees tens of thousands of visitors, and the people who were sick two or three weeks ago and did not see a doctor will never be counted. Five confirmed cases spread across more than one county, four of them in the hospital, is the top of an iceberg whose size depends on how many of those 200 the county follows up with a call and a test request.

Date	What happened	Source
Sept. 29 onward	Exposure window for visitors asked to report symptoms	Will County Health Department
Oct. 21	County contacts the farm; food service areas and vendors inspected, no problems found	Farm statement; WGN
Oct. 23	Investigation announced: five lab-confirmed STEC cases; petting zoo voluntarily closed; survey opened	Will County Health Department
Oct. 26	Still five cases, ages 4 to 38, four hospitalized, at least three discharged; 200-plus surveys; railings, hand-wash stations and tables swabbed; veterinarian samples the animals	The Herald-News
About Nov. 5	Animal and surface laboratory results expected	Will County Health Department, via The Herald-News

Sources: Will County Health Department statements as reported by The Herald-News (October 26, 2026), WGN and NBC Chicago (October 23), read October 27. Check: four hospitalized of five confirmed is 80 percent; the confirmed count has not moved between October 23 and October 26 while surveys passed 200.

The samples taken Monday are the ones that matter, and a negative animal does not clear the pen.

On Monday, October 26, county environmental health staff swabbed railings, hand-washing stations and tables near the food service area and the petting zoo, and a veterinarian sampled the animals. Results are expected in about ten days. Two things about those results. First, ruminants shed E. coli O157 and the other Shiga toxin producers intermittently, so a goat that is negative on October 26 can have been shedding on October 10; the Compendium of Measures to Prevent Disease Associated with Animals in Public Settings says as much and warns against reading a negative as exoneration. Second, the question is not only whether an animal is positive but whether the animal isolate, the railing isolate and the five patient isolates are the same strain by whole genome sequencing. Illinois has the lab to do that. The county should say now that sequencing is planned and that the results, positive or negative, will be published.

A clean food-service inspection does not clear a petting zoo.

The farm’s statement leans on the October 21 inspection of its food service areas and vendors, which found no problems, and the county’s food-preparation tests have come back negative so far. Good. That is also beside the point. In the petting zoo outbreaks on the public record, from the 2004 North Carolina State Fair through the pumpkin patches in Washington and Minnesota, the vehicle was rarely a food item. It was manure on a rail, on a shoe, on a stroller wheel, on a hand that held a sippy cup, in sawdust that blew across a walkway. The county itself says it does not normally regulate petting zoos and is involved because the farm holds food permits. That is the regulatory architecture, not the county’s fault, and it is why the Monday swabs of railings and hand-wash stations are more important than any food test.

What the farm’s statement says, and what it leaves out.

In its statement to NBC Chicago, Bengtson’s says testing is under way, no source has been confirmed, its first thoughts are with the sick and their families, and the petting zoo is staffed, has hand-washing sinks at the entrance and exit and cleans its pens daily. Rides, attractions, food stands and vendors remain open. The statement does not mention that four people were hospitalized. It does not say how many visitors came through the petting zoo between September 29 and the day it closed, whether the farm will contact ticket holders, whether the animals have ever been tested before, or where the animals came from and under what contract. The county’s environmental health director, Sean Conners, says the farm can safely keep operating. I do not quarrel with that call for the rides and the corn maze. I would want the petting zoo to stay closed until the animal and environmental results are back and sequenced, not merely until the season ends.

What should happen in the next ten days.

Post a case definition and a weekly update, with ages, hospitalizations and any hemolytic uremic syndrome, so parents are not learning the hospitalization count from a newspaper. Call every one of the 200 survey respondents who reported bloody diarrhea and get them tested. Sequence every patient, animal and environmental isolate and publish the comparison. Ask the farm to notify ticket holders directly. And when the results are in, whatever they show, put the whole file on the county’s website, because the next pumpkin farm’s operator should be able to read it. The consumer version of this post, with symptoms, incubation and when to see a doctor, is on Food Poison Journal; the prevention material for families and operators is at Fair-Safety.com; and what HUS is and why it follows E. coli in young children is at about-hus.com.

How many people are sick from Bengtson’s Pumpkin Farm E. coli? Five laboratory-confirmed cases as of October 26, 2026, ages four to 38, four of them hospitalized and at least three discharged. More than 200 visitors have submitted symptom surveys that the county is still reviewing.

Has the source of the Bengtson’s E. coli outbreak been found? No. Food-preparation tests are negative so far. Animals, railings, hand-wash stations and tables were sampled October 26, with results expected in about ten days.

Is Bengtson’s Pumpkin Farm open? Yes, except for the petting zoo, which the farm closed voluntarily on October 23.

The Justice Department just told FDA what I have been saying: its own regulations are what keep outbreak names secret, and it has the authority to change them. Here is how I am changing my two petitions.

By Bill Marler on October 26, 2026

Posted in Case News

On October 26, 2026, the Department of Justice’s Office of Legal Counsel issued an opinion to the Acting General Counsel of the Department of Health and Human Services titled Public Disclosure of Supply Chain Information in Foodborne Illness Outbreaks and Recalls. HHS asked a simple question: is FDA authorized by law to tell the public who is in the supply chain of a food that is making people sick?

The answer, in eight pages, is the argument I have been making for years. FDA’s current regulations mostly say no. The statute says FDA can change that. And FDA has not.

That matters right now. Fifty people are sick in the chicken wrap E. coli O157:H7 outbreak, reference #1421, that FDA and FSIS are investigating together. The public learned the wraps were Trader Joe’s Tarragon Chicken Salad Wraps from a lawsuit my firm filed and from a newspaper, not from either agency. In July, FDA told Stop Foodborne Illness that its pending petition on this exact question raised “complex issues.” The government’s own lawyers have now answered the hardest of them.

What the opinion says

First, the problem is FDA’s regulations, not the law. The Trade Secrets Act bars federal employees from disclosing confidential business information only when the disclosure is not “authorized by law,” and a properly issued regulation is enough authorization. FDA’s regulations treat the names and locations of firms in a supply chain as confidential commercial information under 21 CFR 20.61, and they wall that information off. The opinion concludes that, given those regulations, the Trade Secrets Act “largely prevents” FDA from disclosing supply chain information.

Second, the exception FDA leans on barely works. 21 CFR 20.91 lets FDA disclose information to the extent necessary to carry out an enforcement action like a recall. The opinion calls that authority “exceedingly narrow in practice.” It applies only when there is an actual enforcement action. In the opinion’s words, “FDA’s hands are tied before a recall is officially initiated.” And by the time FDA collects the information, it is often no longer “necessary.” The opinion also says the statute “nowhere requires” the case-by-case analysis in FDA’s 2020 consignee guidance, which it calls “arduous.” That is the chicken wrap outbreak in a sentence. The wraps were “proactively removed,” no recall was ever announced, and so nothing was ever disclosed.

Third, FDA can fix it. Section 705(b) of the Food, Drug, and Cosmetic Act lets FDA disseminate information regarding food in situations involving, in the Commissioner’s opinion, “imminent danger to health or gross deception of the consumer.” Combined with FDA’s general rulemaking authority, the opinion concludes, that provision lets FDA amend its regulations to authorize disclosure of supply chain information. Disclosure under such a rule would be “authorized by law” and would not violate the Trade Secrets Act. The opinion adds that this would hardly be a “transformative expansion” of FDA’s power, and that neither the FOIA exemption for commercial information nor the Act’s own trade secret provision stands in the way.

Fourth, in a footnote, the opinion questions whether the identity of a supply chain partner is even the kind of information the Trade Secrets Act protects. Every item the Act lists concerns a company’s internal processes, and who a company buys from or sells to is different in kind. The opinion assumes coverage for purposes of its analysis and notes, fairly, that some appellate courts would likely see it the other way.

Fifth, there is a limit, and I want to be honest about it. The opinion says a regulation “purporting to authorize disclosure after a foodborne illness outbreak has already ended would likely be inconsistent” with section 705(b). The imminent-danger authority is a tool for an open outbreak.

Where disclosure should have happened, and why

The opinion is not an abstraction. Run the outbreaks in my petitions through it and you see exactly where the public was left in the dark, and on what basis it should not have been. FDA, CDC and their partners have named farms, processors and brands at least eight times since 2006 after the product was gone. Here is what happens when they choose not to.

Romaine, fall 2024 (reference #1280). Eighty-nine people sick in fifteen states, thirty-six hospitalized, seven with kidney failure, one dead. FDA closed the investigation with one word: romaine. When the traceback summary was first released, the processor, the grower and the ranch were blacked out. It came out unredacted more than a year later, in stages, and only after three firms in the supply chain signed consents to release their own information. The document showed traceback had converged on a single processor and a single grower. Why it should have been disclosed: while people were getting sick that fall, this was imminent danger by any measure, and a section 705(b) rule would have authorized naming the firms then. Instead, as the opinion confirms, the only route left was the firms’ own consent. The public’s right to know depended on the permission of the companies being named.

Romaine, summer 2026 (reference #1382). Fourteen people sick. For ten weeks the only public record was a single row reading “Not Yet Identified.” On September 10, FDA closed the file, said romaine for the first time, and named no grower and no processor because the lettuce was past shelf life. It never said how many states had cases, how many people were hospitalized, whether anyone developed HUS, or whether traceback converged. Why it should have been disclosed: case counts are not anyone’s trade secret, and FDA published exactly those numbers for a closed romaine outbreak in 2019. Naming the firm after closure is where the opinion’s limit bites, which is precisely why the petition asks FDA to define by rule that an implicated firm’s name is not confidential to begin with.

Salmonella Newport (reference #1403). Posted August 26 with eighty-four cases and the food “not yet identified.” Traceback and sampling began. The count rose to ninety-seven, then 111. On October 21, FDA closed it at 119 cases without ever naming a food. No inspection, no advisory, no CDC notice. Why it should have been disclosed: the name of a food is not confidential commercial information at all. While the count was climbing for two months, this was the textbook open outbreak the opinion says section 705(b) covers. The petition asks FDA to name the suspect food once traceback begins. One hundred nineteen people deserved at least that.

Cantaloupe (reference #1366). Seventy people sick in twenty-five states. FDA said the fruit was likely past shelf life. Yet the importer, Ayco Farms of Pompano Beach, recalled a lot on March 24, 2026, FDA put two Guatemalan growers on import alert, and all three were named. Why this one is the tell: the same year, past-shelf-life romaine got no name. The difference was not the danger. It was whether a recall or an enforcement action happened to attach, which is exactly the narrow 20.91 trigger the opinion describes. And even here, FDA’s executive summary still redacts the states with cases and the other firms in the traceback.

Packaged organic salads (reference #1339). FDA’s own executive summary says traceback converged on a supplier and processor of packaged organic salads, then redacts the name and even the number of traceback legs. Why it should have been disclosed: the work was done and the conclusion reached. All that is missing is a name, and the rule that would require it.

Happy Hens eggs, San Diego County. California named the forty-one stores that received the implicated eggs this September. The federal record named none, and the recalled date codes were packed in August, after nearly every illness. Why FDA could have said the same thing: once a state lawfully publishes the stores, that information has “previously been disclosed in a lawful manner to any member of the public,” one of the exceptions the opinion confirms is authorization by law. FDA did not need a new rule to repeat it. It needed to want to.

Beef kofta, May 2026, and the chicken wraps now. When product is gone, FSIS’s routine is to issue a public health alert naming the establishment. In May it named The Kebab Shop and the producer of beef made four months earlier. In the joint chicken wrap investigation, with fifty sick and the product identified, it has issued nothing. One government, two defaults, and the only difference is who happens to be making the call.

What I am changing in the Outbreak Firm Disclosure petition

A new disclosure rule, exactly as the opinion describes. The petition already asked FDA to declare by rule that an implicated firm’s name is not confidential commercial information, and to make distribution information disclosable whenever an outbreak food is pulled, whatever the company calls the removal. It now also asks for a new regulation under sections 705(b) and 701(a). That regulation would authorize disclosure of the firm’s name, location and distribution information whenever the Commissioner finds an outbreak presents imminent danger. It would also require conforming changes to the three regulations the opinion identifies as the current bars. Each of these three changes is an independent ground. If a court set one aside, the others would still authorize disclosure.

Drawing the line where the opinion draws it, at the outbreak and not the package. The new 705(b) rule applies while an outbreak is open. That means while people are still getting sick or cases are still being found, or while the food or an ingredient may still be in commerce, in homes or in other products. Past its shelf-life date or not. When an investigation closes, naming the firms rests instead on the rule defining what counts as confidential, on the case law holding that a business’s name is not commercial information in itself, and on FDA’s existing exceptions. I took out language that said section 705(b) carries no requirement that an investigation still be open. The petition now cites the opinion’s limit against itself, because a petition that hides the hard part is not worth filing.

“Use the authority you already have” now means the authority the opinion confirms. Until a rule is final, the petition asks FDA to name firms whenever one of four existing authorizations applies. The opinion confirms each counts as authorization by law: information already lawfully disclosed to the public, a court order, the firm’s consent, and an enforcement action under 20.91. The first one is bigger than it sounds. When a retailer tells a national newspaper which product it pulled and in how many states, that information is public. Once Trader Joe’s did exactly that, FDA’s own regulation let it post the same facts on its outbreak table. It still has not.

Two smaller changes. Because the opinion reads 20.91 as requiring an actual enforcement action, the petition no longer asks FDA to treat a whole genome sequencing match as one by interpretation. That request moves into the rule. And because the legal question is now answered, the petition asks FDA to respond with a proposed rule, not another interim letter.

What I am changing in the Recall Consignee List petition

The petition still asks FDA to do by rule what USDA has done since 2008: publish the stores, restaurants, schools, hospitals and other places that received food in every Class I recall, when the recall is announced. It now rests that rule on two authorities instead of one: 20.91, and section 705(b) as the opinion reads it. A Class I recall already means there is a reasonable probability the food will cause serious harm or death. So the petition asks that, when FDA classifies a food recall as Class I, it also state whether the Commissioner finds imminent danger. That way publishing the store list no longer turns on a case-by-case finding that it is “necessary.”

It also uses the opinion to support two requests already in the petition. One is dropping the guidance’s packaging test, the analysis the opinion called “arduous.” The other is confirming that pulling an outbreak food is a recall no matter what the company calls it, because the opinion confirms 20.91 does nothing until a recall is “officially initiated.” And it explains why the rule fits within the opinion’s limit. The list is published when the recall is announced and updated while it is underway, which is when consumers can still use it and health departments are still finding cases.

What is not changing

Both petitions still ask for the same core things. They ask for a written standard with the exemption closed and the burden on the firm. They ask for one business day for a firm to correct errors and a prominent correction if a named firm is cleared. They reach the whole supply chain, from grower to retailer. They ask for case counts and the suspect food while an investigation is open, and an honest account of the traceback when it closes. They still go in together, on the same day. Both now also copy the HHS Acting General Counsel, who asked for the opinion. I still support the Stop Foodborne Illness petition. And S. 4974, passed by the Senate in September, is now less a precondition than a supplement. FDA does not need to wait for Congress.

Read them. Tell me what you think.

If you would like the revised drafts, email me at bmarler@marlerclark.com and ask for the petitions. Please get comments back to me before October 30, 2026.

For years FDA has said its hands were tied. The Justice Department has now said who tied them. FDA did, in its own regulations, and it can untie them by rule. Fifty people are sick from a wrap whose maker the government will not name. The law no longer gives FDA an excuse. Let’s see whether it will act.

Marler Blog

Marler Blog

Credit Where It Is Due on Formula Nutrients — and a Reminder That Fifty-Two Babies Were Paralyzed by Something Else

ByHeart Is “Discontinued” – But Its Founders Are Buying Perrigo’s Infant Formula Business

FDA Says the Trader Joe’s Chicken Wrap E. coli Outbreak Is Over – and That Telling You More “Would Not Have Provided Consumers with an Action.” Section 705(b) Has a Name for That: Gross Deception of the Consumer

Brookshire Farms E. coli Outbreak: Five Sick, Three Children in Intensive Care, and the County’s First Concrete Advice Is About the Pumpkins, Not the Animals

Tip of the Pen to the Healthy People Food Safety Workgroup: Forty Years of Scorekeeping Show Salmonella Has Not Budged Since 1997, Three of Four Pathogens Are Going the Wrong Way, and We Stopped Measuring the Trend. Here Are Ten Things That Would Move the Needle.

Pumpkin Patches, Petting Zoos and State Fairs: Three New E. coli Investigations, Twenty-Five Years of Outbreaks Just Like Them, and the Prevention Playbook Everyone Already Has

Two Pumpkin Farms and the State Fair of Texas: Three E. coli Investigations Tied to Animal Attractions in Five Days, in Three States. Here We Go Again.

Tip of the Pen to The Washington Post: FDA Has Inspected Only About 2 Percent of Mexican Produce Facilities Since 2009. Congress Sent Letters. The Post Did the Oversight.

Bengtson’s Pumpkin Farm E. coli Outbreak: Four of Five Sick Were Hospitalized, More Than 200 Visitors Have Reported Symptoms, and the Animal Samples Went to the Lab Monday. We Are Investigating for One of the Families.

The Justice Department just told FDA what I have been saying: its own regulations are what keep outbreak names secret, and it has the authority to change them. Here is how I am changing my two petitions.

Blogs

Disease

Insight & Litigation