This afternoon I fly from Seattle to Oslo. On Wednesday and Thursday, September 23 and 24, I give two talks at the Aquatiq Food Forum 2026 at the Clarion Hotel Oslo: “Why It Is a Bad Idea to Poison Your Customers” and “The Criminalization of Food Safety Failures.” I wrote about both talks last week. This is a shorter note about the one thing that changed after I wrote it.

What happened on September 14

Last Monday the Justice Department announced that Abbott Laboratories agreed to pay $384,999,040 to resolve allegations that it caused false claims to be submitted to federal and state programs for powdered infant formula and nutritional therapy products made at its Sturgis, Michigan, and Casa Grande, Arizona, plants between January 1, 2018, and December 31, 2022. Of that, $348,700,868 goes to the United States and $36,298,172 to states for Medicaid and WIC claims. Three Abbott employees who brought the whistleblower case will share $69 million.

The government’s complaint, filed in November 2025, described a roof at Sturgis that leaked water over equipment, temporary measures to divert the leaks, and what the Justice Department called “a culture of concealment.” Those are allegations. The settlement is not a finding of liability, and Abbott denies wrongdoing. Abbott has also said, according to Reuters, that the related criminal investigation has been closed.

When the settlement came down, I said it bought $385 million worth of accountability for how formula was made between 2018 and 2022, and that every dollar of it was backward-looking. Not one dollar requires anyone to run a single additional test tomorrow. I still think that.

Why it belongs in a talk about criminal law

My second talk in Oslo asks who goes to prison when food makes people sick. The honest American answer, this month, is: a company pays, a whistleblower is rewarded, and no individual is charged. That is not an accident. Executive Order 14294, signed in May 2025, calls criminal enforcement of strict liability regulatory offenses generally disfavored and tells prosecutors to treat it as a last resort. The Park doctrine, which lets a responsible corporate officer be convicted without proof of intent, is still good law. It is simply not being used.

Europe is moving the other way. In France, Nestlé France and its subsidiary were placed under formal investigation in July 2024 for involuntary manslaughter over Buitoni frozen pizzas that left 48 people, almost all of them young children, with hemolytic uremic syndrome and killed two children. Lactalis and its Craon subsidiary were placed under formal investigation in 2023, in part for failing to carry out withdrawal and recall measures. Neither case has been decided. The direction is still clear.

Why Norway should care

Norway sits on the European side of that line. Article 19 of Regulation (EC) No 178/2002 requires a food business that has reason to believe its product is unsafe to withdraw it immediately and inform the authorities. Norway took that regulation into the EEA Agreement in 2007, and Mattilsynet enforces it under the Food Act. A duty framed as immediate is a duty a prosecutor can measure in hours.

That is the message I am carrying to a room that will include Mattilsynet, producers, retailers and some of the largest food companies in the world. It is the same message I have carried for thirty-three years: it is cheaper to be safe. The difference now is that on one side of the Atlantic, the price of not being safe is starting to include a courtroom with a defendant’s name on the docket, not just a company’s.

I will write up what the audience asks, and what I learn from it, when I get home. If you are in Oslo on the 23rd and 24th, come find me.

We need a lawyer who can find the defendant and a paralegal who can find the one page that matters.

Marler Clark has been suing the companies that poison people since 1993 — since a nine-year-old girl named Brianne Kiner nearly died from a hamburger. Sixteen years later it was Stephanie Smith, a children’s dance instructor paralyzed by a hamburger, whose case put the grinding of American beef on the front page of The New York Times and won Michael Moss a Pulitzer Prize. More than thirty years on, we are still the only law firm in the country that does nothing but foodborne illness — E. coli, Salmonella, Listeria, Cyclospora, Vibrioand botulism, tens of thousands of clients and over $850 million recovered. The names on the other side are ones you already know: Jack in the Box, Cargill, Dole, Nestlé, Walmart, Wendy’s, McDonald’s, Yum Brands, and this year Taylor Farms and Boar’s Head. There is more work than there are people to do it, and we are hiring two of them.

Where the work stands. As I write this, on September 19, 2026, CDC has closed the iceberg lettuce Cyclospora outbreak tied to Taylor Farms at 12,883 sick in 21 states. Its own national count is 19,883 confirmed cases. That leaves 7,000 confirmed cases nobody has tied to a food, and this fall we will be working with epidemiologists to find where they came from. Salmonella Javiana in jalapeños grown in Sinaloa has sickened 431 people in 32 states; the grower still has not been named, the recall has spread to more than 160 grocery products, and we filed the first lawsuit in August. Salmonella from a Ramona, California deli and the egg farm that supplied it, with three complaints on file. Two sprout outbreaks at once. Infant botulism against ByHeart and Nara Organics. Listeria from Boar’s Head. E. coli from raw goat milk, raw milk cheese, romaine and frozen blueberries. Every one of those numbers is a person with a chart, a bill, a lien and a phone number. Clients in need find us; that has never been the hard part.

Finding the defendant is. Nobody hands you one in a foodborne case. The lawyers here read FDA inspection reports and environmental assessments, USDA establishment records, whole genome sequencing and PulseNet clusters, epidemic curves built by state health departments and, this summer, the sanitation and rainfall records of the Mexican municipality where the lettuce was grown. The paralegals get those records in the first place: public records requests to state and county health departments, FOIA to FDA, CDC and USDA, and some of it in Spanish, a real plus. You will also have the Outbreak Database, which this office built and still maintains — decades of outbreaks, searchable, and no other firm in the country has it.

The other half of the office is not litigation at all. E. coli O157:H7 became an adulterant in American beef in 1994; six more strains followed because we petitioned USDA to make them so. That file had to be built by somebody. This month: citizen petitions to FDA, now being drafted, for a warning label on raw sprouts and for botulism testing in powdered infant formula, the second with twenty-four parents of hospitalized babies; comment letters on the traceability rule Congress delayed to 2028; and a standing demand that CDC keep counting Cyclospora, which it stopped updating weekly on September 15. I publish Food Safety News, have written over 6,000 posts here since 2004, and teach at the Harvard T.H. Chan School of Public Health. People here get their names on that work. Lawyers write here under their own names, where reporters read it, and anyone who wants to speak or write about the work is pushed forward, not held back. Cases fix one company; the rest fixes the industry.

What you will not do here. No billable hours, no timesheets and no origination credit to chase — the clients find us, and the firm carries the cost of every case out of its own pocket until it resolves. No committee, and no layer between you and the person who decides; when you need an answer you walk down the hall and get one. And no volume. We turn down cases every week that we could not do right, because a file nobody has time for is worse than no file at all.

Where you will do it. Our office is on Bainbridge Island, and we want you living here, not commuting here. We are a few minutes’ walk up from the ferry terminal, thirty-five minutes across the water from downtown Seattle, in a town where you can walk to work, walk to lunch and walk home. But the right person matters more to us than the right zip code: if you are elsewhere in Puget Sound, or willing to relocate, and can be here most days, we should talk, and we will help you get here.

The associate attorney: at least five years of litigation experience — personal injury, product liability, med-mal or complex tort — and ready to run cases, not just paper. We are a handful of lawyers, not a floor of them; you carry your own docket from the start. You will file in federal and state courts across the country; write the settlement demand letters that resolve most of our cases, with the medical chronology, the liability analysis and the damages model all yours; take and defend depositions of plant managers and epidemiologists; work up hemolytic uremic syndrome, Listeria meningitis and Salmonella reactive arthritis with the best experts in the field; sit with families in hospital rooms; and try cases. You will be on airplanes. You need a license in good standing somewhere and admissibility pro hac vice everywhere else (Washington admission is not required), real deposition experience, the writing to draft a complaint reporters quote, and the stomach for what an outbreak does to a child. A science background is a bonus; we will teach you the microbiology.

The paralegal: at least ten years in plaintiffs’ personal injury or complex litigation. Ten years, because you already know what a lien is, what a hospital billing department will and will not send you, and how to keep sixty clients straight without losing one — and because nobody here will look over your shoulder. You will run your own files: intakes with families who are frightened and exhausted, medical records chased and chronologized, the damages and lien files that become the settlement demand, e-discovery and document productions, calendaring and cite-checking, coordination with experts and health departments, FDA and USDA dockets tracked and comments filed on deadline, and travel with us when a case goes to trial. You need mastery of medical records and chronologies, e-discovery platform experience, absurd organizational instincts, and genuine kindness on the phone: the person calling you may have just buried someone.

The children’s cases belong to both jobs. Many of our clients are children, and the last step on their cases is the hardest. A child’s settlement has to be approved by a judge, cleared of Medicaid, Medicare, ERISA and sometimes federal military liens, and structured — now and then into a special needs trust — so the money is still there decades later, when the kidney an outbreak took finally has to be replaced. If you have never done a minor’s compromise, you will.

What we offer. Competitive salary depending on experience, plus bonus; medical, dental and vision; 401(k) with match; paid vacation and holidays; bar dues and CLE for the associate; and cases that end up in The New York Times, in front of Congress, and in Poisoned, Jeff Benedict’s book about this work and the Netflix documentary that followed. Read the book and watch the documentary before you apply.

One last thing, and it is the whole thing. Nobody here treats this as a practice area or a file-management job. A client wrote to us that I came to the hospital while their son was there, and that somebody from this office was in touch every step after. Being that somebody is the job, for the lawyer and the paralegal both. The people who last here can still name the child a year after the case settles, and they are still angry that any of it happened. I am not looking for someone who wants a job. I am looking for someone who cannot look away from it.

How to apply. Both jobs are open now. Send a resume and a short note to bmarler@marlerclark.com. Associates, add a writing sample, and in the note pick any case or outbreak off marlerblog.com and tell me what you would do with it first. Paralegals, tell me about the hardest lien you ever cleared or the record you had to fight hardest to get. marlerclark.com and billmarler.com if you want to know more before you write. No recruiters. Equal opportunity employer.

Saturday, September 19, 2026

This is the eleventh Saturday column. Last Saturday’s edition ran under the headline CDC Closed the Largest Cyclospora Outbreak in American History at 12,883 Sick; its point was that closed is not the same as answered. This week the answers turned up in places nobody would think to look: FDA confirmed the sprout seed recall and left the supplier’s name in the Enforcement Report, Washington counted 27 sick in the broccoli sprout outbreak while CDC’s page says 22, FDA posted five closure summaries with a name blacked out, CDC said it will stop counting Cyclospora weekly, Abbott agreed to pay $385 million while admitting nothing, and a fourteenth congressional letter went out with no hearing behind it. The record exists. It is just not on the page where people look.

On Tuesday FDA said the seed supplier agreed to recall the alfalfa seed lot, and the advisory still does not say who the supplier is. FDA’s sprouts advisory, stamped September 15, says the seed supplier of Everything Sprouts, LLC agreed to recall the seed lot, sold to sprout growers and wholesale distributors in Puerto Rico and the United States and shipped to Canada, Mexico, Tahiti and Venezuela. The count is unchanged since September 1 at 55 sick, four hospitalized and no deaths in fifteen states, 46 E. coli, seven Salmonella Agona and two with both. The page names no supplier and no lot. Both have been public since September 10, when Phyllis Entis found them in FDA’s Enforcement Report: International Sprout Holdings Inc. of Cookeville, Tennessee, lot SAL2-24JB, 43,799 pounds, recall initiated August 23. I wrote about it that day; Food Safety News reported Wednesday that FDA has identified at least one more sprout grower that received the lot and has not named it either. On Tuesday Todd’s Seeds, an online sprouting-seed seller, recalled its alfalfa seed and mixes packed between January and May, named International Specialty Supply as the source, and cited a positive E. coli result on spent irrigation water tied to the lot. The name went to the record, not to the page the public reads.

Washington counted 27 people in the broccoli sprout outbreak on Thursday, and the federal pages still say 22. Washington’s Department of Health updated its page September 17 to 21 Washington cases, two hospitalized, in nine counties (King and Whatcom five each, Spokane three, Clark and Skagit two each, one each in Island, Snohomish, Walla Walla and Yakima), up from 16 on September 9, and says the outbreak now includes 27 people from four states. CDC’s page was last updated September 11 at 22 sick in Idaho, Montana, Utah and Washington; FDA’s advisory has not moved since September 9. Evergreen Fresh Sprouts of Moyie Springs, Idaho, recalled the sprouts September 4, the notice reached FDA’s page September 10, and the last use-by date was Wednesday. The two sprout outbreaks together stand at 82 sick by the states’ counts, and CDC says they do not appear to be related. I wrote about the broccoli outbreak the day it went federal and about both together two days later, when the number was 77.

Utah’s Bear River Health Department ruled the sprouts out and still has no source for 21 cases.The department’s September 1 release has not been updated; the Box Elder News Journal reported Wednesday that whole genome sequencing shows the Cache and Box Elder strain is not the Evergreen strain. Twenty-one cases in a district of about 200,000 people, three weeks without a date range, a serotype or a suspected food, and on Thursday I listed the twelve questions the department should answer in public.

The Happy Hens recall is eleven days old, the egg test results are five days overdue, and the count is 23. The California Department of Public Health’s September 11 release put the recall date at September 8 and the count at 23 sick and seven hospitalized, unchanged since September 4; nothing has been published since by the County, the state or FDA. The farm said August 31 that 4,000 eggs would be tested over two weeks, which ran out Monday; as of this morning the farm’s website carries no recall notice and no results, and the shop is taking egg orders for pickup in Ramona. I published the state’s list of 41 San Diego County locations that received the eggs. Vermont’s Leaping Bear Farm outbreak sits at 14. The UK Health Security Agency counted 586 confirmed Salmonella Enteritidis cases as of September 15 in three clusters likely linked to imported eggs, per Phyllis Entis’s Thursday ledger; I added up the American and European egg outbreaks on Thursday at more than 750 sick and two dead.

CDC will stop counting weekly at 19,883, and the fourteenth letter went to USDA. CDC’s case-data page, as of September 15, lists 19,883 confirmed domestic cases, 1,064 hospitalizations and two deaths in 49 states and the District of Columbia, with 9,765 still being sorted between domestic and travel and CDC aware of at least 6,594 more, and says CDC will stop updating it weekly because recent infections have significantly declined. Against the outbreak CDC closed at 12,883, that leaves exactly 7,000 confirmed cases with no food named. I laid out what each number counts last Saturday, from the states’ 33,548 and CDC’s 19,595 that day down to the one percent of infections any count catches. On Monday Representative Debbie Dingell and nine House colleagues wrote to Secretary Rollins asking why two of the Agricultural Research Service’s three Cyclospora research projects were shelved and the third is moving from Beltsville, Maryland to Iowa, with a two-week deadline. That is the fourteenth congressional letter of this outbreak; none has a published answer, and there is still no hearing. Helena Bottemiller Evich wrote Friday that Congress has stopped doing its job on food safety; the letters are the proof.

The produce industry spent the week talking about the outbreak, and I answered each piece separately. The California Leafy Greens Marketing Agreement marked twenty years since the 2006 spinach outbreak on September 15; the International Fresh Produce Association closed its outbreak page September 14 saying the findings rested on epidemiology and traceback rather than a positive laboratory test, which Taylor Farms and its 2,000 negative samples should read next to FDA’s 2020 warning letter to Fresh Express; Donna Garren of FDA’s Office of Produce Safety said in Portland, Maine on Tuesday never to let a good crisis go to waste; and De Ann Davis at Western Growers wants FDA to say exactly which food is unsafe, where I agree and wrote out where we differ.

FDA’s files with no food named grew by six, and its five new closure summaries include one with the name blacked out. FDA’s CORE table, stamped September 16, moved the Salmonella Newport file from 109 to 111, the largest anonymous Cyclospora file from 231 to 234, and the Listeria file from nine to ten; the eight rows with no food named now carry 575 sick, up from the 569 I counted Wednesday, and CDC’s dashboard lists 38 active multistate investigations, up from 37. On Thursday FDA posted executive summaries for five closed 2026 files: cantaloupe names Ayco Farms of Pompano Beach, Florida, and the Guatemalan farms behind 70 cases in 25 states; requesón names Clover Hill Dairy of Mechanicsville, Maryland, behind 15 cases and one death, and a suspended Maryland permit; moringa says a supplier refused FDA access and redacts its name. I went through all five on Food Poison Journal; Food Safety News has the cantaloupe and requesón detail.

The recall ledger, kept as every week by Phyllis Entis at eFoodAlert. No FSIS recall or public health alert could be found for the week. Gias Foods recalled bettergoods Lemon Alfredo Fettuccine sold nationwide at Walmart on Tuesday after Washington and Florida state sampling found Listeria monocytogenes in two lots; no illnesses are reported. A Markys Caviar recall of Mon Sire brie for Listeria, distributed in Florida, appears only in the Enforcement Report. Danone recalled So Delicious salted caramel frozen dessert for small stones, Saratoga Potato Chips recalled J. Higgs bacon and cheddar chips for undeclared soy at Save A Lot stores in five states, GF Blends recalled gluten-free flour and mixes over 20 parts per million of gluten, and in Canada Alasko frozen raspberries were recalled for norovirus with illnesses reported across seven provinces. CDC touched two pages Thursday: hepatitis A in frozen black clams, still 37 sick and 30 hospitalized in four states, and pet turtles, now 20 sick in twelve states, half under five; I wrote about the turtles this morning. The frozen blueberry E. coli O145 file holds at 17 sick in Florida and Georgia while Publix’s recall page says there are no Publix recalls at this time.

Two agencies published rules, one company paid $385 million, and 77 congressional offices got a t-shirt. FDA released the 2026 Food Code Thursday, the biennial model behind more than a million restaurants and stores; written employee illness policies are now required. On Wednesday FDA published its survey of 438 grocery delis: 84 percent slipped on holding food safely and 62 percent on handwashing. USDA opened a $50 million Stand-Up program Wednesday for state meat and poultry inspection programs; thirty states run one now. On Monday the Justice Department announced that Abbott will pay $384,999,040 to settle False Claims Act allegations about its Sturgis, Michigan and Casa Grande, Arizona formula plants from 2018 through 2022, $348.7 million federal, $36.3 million to the states and $69 million to three whistleblowers, with the claims called allegations only. The rule still does not require the test. On Wednesday and Thursday t-shirts rolled into baby bottles went to 77 House and Senate committee offices with four unanswered letters asking for Clostridium botulinum testing in powdered infant formula; 52 babies were hospitalized in the ByHeart and Nara outbreaks. Firm business, briefly: three jobs are still open, an associate attorney with five or more years, a paralegal with ten or more years and a law student or LL.M. candidate, paid and remote, to cite-check four citizen petitions, resume and a short note to bmarler@marlerclark.com, no recruiters; in the Caruso case in Ohio we moved September 10 to drop Taco Bell and proceed against Taylor Farms alone; and Wednesday and Thursday I will be in Oslo giving two talks at the Aquatiq Food Forum.

That’s the week — a seed supplier named everywhere except on the advisory, a broccoli sprout count that is 27 in Olympia and 22 in Atlanta, an egg farm with a recall on the state’s page and none on its own, seven thousand confirmed Cyclospora cases with no food, and a fourteenth letter waiting on an answer. The record is there. Check back next Saturday.

About Bill Marler

William “Bill” Marler has spent more than thirty years as a food safety lawyer and advocate—work that began with the 1993 Jack in the Box E. coli outbreak and has never really stopped since. In the years that followed, he has represented victims in nearly every major foodborne illness outbreak in the United States. That case, and the movement it launched, is the subject of the book “Poisoned” and the Emmy Award–winning Netflix documentary of the same name. Bill’s work has been profiled in The New Yorker (“A Bug in the System”), the Seattle Times (“30 years after the deadly E. coli outbreak, a Seattle attorney still fights for food safety”), the Washington Post (“He helped make burgers safer. Now he’s fighting food poisoning again”), and many others.

Dozens of times a year, Bill speaks to industry, regulators, and universities across the United States, Canada, Europe, Africa, China, and Australia about a simple idea: outbreaks are preventable. He has testified before Congress on the Food Safety Modernization Act and teaches food safety at institutions including the Harvard T.H. Chan School of Public Health. He writes regularly about food litigation and food safety at Marler Blog, and in 2009 he founded Food Safety News, which he continues to publish.

A Salmonella outbreak linked to pet turtles has more than doubled in six weeks. In its September 18, 2026 update, the Centers for Disease Control and Prevention reported 20 sick people in 12 states, up from 8 children in 6 states when the outbreak was first announced on August 5. Five people have been hospitalized. No deaths have been reported. Half of the people who got sick are children younger than 5.

The outbreak strain is Salmonella Paratyphi B variant L(+) tartrate+. CDC says any turtle can carry Salmonella, but turtles with shells less than 4 inches long are a known source of illness. Selling those turtles as pets has been against federal law since 1975.

The numbers as of September 18

CDC’s updated outbreak page lists 20 cases (12 new), 5 hospitalizations (2 new), no deaths, and 12 states (6 new). CDC notes that the true number of sick people is likely higher, because many people recover without seeing a doctor and are never tested, and because it usually takes three to four weeks to confirm that a new illness is part of an outbreak.

CDC’s case data, current as of September 17, place the sick people in Florida (6), Texas (4), and one each in Iowa, Kentucky, New Mexico, New York, North Carolina, Oklahoma, Pennsylvania, Rhode Island, Virginia and Wisconsin. Florida and Texas together account for half the cases. Illnesses began between June 6 and August 22, 2026, with the two most recent both starting on August 22. The outbreak did not end with the July cases CDC first reported; twelve of the twenty people got sick after July 4.

What investigators found in August

CDC’s first investigation notice described 8 children, counted as of July 27, who got sick between June 6 and July 4, 2026. They ranged in age from under 1 to 7 years, with a median age of 4, and a quarter of them were a year old or younger. Three of the six children with information available were hospitalized.

Parents of six of the children were interviewed, and five of them (83%) reported that their child had contact with a pet turtle before getting sick. Four parents knew the size of the turtle, and three of those four said its shell was less than 4 inches long. Parents of three children said the turtle came from a pet store, and one child received the turtle as a gift. CDC had not identified a common turtle supplier. As of this writing, CDC’s investigation-detail page still carries those August figures, so it does not yet say how many of the new patients reported turtle contact, how old they are, or where their turtles came from.

A 1975 rule that worked

On May 23, 1975, the Food and Drug Administration banned the sale and distribution of viable turtle eggs and live turtles with a shell length of less than 4 inches. The rule sits at 21 CFR 1240.62 and says those turtles “shall not be sold, held for sale, or offered” for commercial or public distribution.

The problem it addressed was enormous. According to a 2016 paper in CDC’s journal Emerging Infectious Diseases, by the early 1970s about 15 million turtle hatchlings were sold in the United States every year, 4% of households owned a pet turtle, and 14% of human salmonellosis cases were attributed to small pet turtles. The Council of State and Territorial Epidemiologists has noted that voluntary education efforts failed to fix the problem before FDA acted.

The rule worked. CDC has estimated that the ban prevents about 100,000 cases of salmonellosis among children each year.

I have seen the same thing in my own practice. I used to be in the beef business with E. coli O157:H7. Once the government set a clear rule for ground beef and enforced it, those cases largely disappeared from my firm. Clear rules, enforced, protect children. The turtle ban is proof of that from fifty years ago.

Why children are still getting sick

The rule has exceptions. It does not apply to sales for bona fide scientific, educational or exhibition purposes other than use as pets, to distribution “not in connection with a business,” or to turtles labeled for export only. In 2013, FDA also removed the procedures in the rule that called for routine destruction of turtles found held for sale, saying routine destruction was not necessary to achieve the rule’s purpose.

The larger gap is enforcement. CDC says on its outbreak page that small turtles are still sold illegally online, in stores, at flea markets and at roadside stands. The 2016 Emerging Infectious Diseases paper reached the same conclusion: despite the long-standing ban, small turtles can be readily acquired through multiple venues and remain the main source of turtle-associated salmonellosis in children. CDC keeps a menu of state turtle laws, which shows how much of the job has been left to the states.

What should happen now

The rule does not need rewriting. It needs to be enforced where the turtles are actually sold today, which increasingly means online listings, marketplaces and shipping. FDA and CDC should also tell the public where these turtles came from once a supplier is found. Parents cannot avoid a seller no one will name.

There is a larger point here too. The turtle rule was issued under Section 361 of the Public Health Service Act, the communicable-disease authority codified in 21 CFR Part 1240. That same authority is what I rely on in a citizen petition I have drafted asking FDA to sample at animal feeding operations implicated in produce outbreaks. FDA has used Part 1240 to regulate an animal at its source when that animal was making children sick. The turtle ban shows it can work.

What parents should do

CDC advises families not to buy turtles with shells less than 4 inches long, and to buy larger turtles only from a reputable pet store. Pet turtles are not recommended for children younger than 5, adults 65 and older, or people with weakened immune systems. Wash hands with soap and water right after touching a turtle or anything in its habitat, do not kiss or snuggle turtles, keep turtles and their supplies out of the kitchen, and do not let turtles roam where young children crawl and play. If a turtle is not right for your family, do not release it outdoors; call a reptile rescue, an animal shelter or a pet store about rehoming it.

Questions parents ask about turtles and Salmonella

Can a healthy-looking turtle make my child sick? Yes. CDC says turtles of any size can carry Salmonellain their droppings even when they look healthy and clean, and the germs spread easily to their bodies, their tank water and anything they touch. State epidemiologists have noted that turtles can shed the bacteria intermittently and that antibiotic treatment of turtles has not eliminated it.

Is it legal to buy a turtle with a shell under 4 inches? Not as a pet from a business. Federal law has barred that since 1975, with narrow exceptions for bona fide scientific, educational and exhibition uses, non-business distribution and export. FDA administers the rule.

When should I call a doctor? CDC says to call a health care provider right away for diarrhea with a fever higher than 102°F, diarrhea for more than two days that is not improving, bloody diarrhea, vomiting so severe that liquids cannot be kept down, or signs of dehydration. Symptoms usually start 6 hours to 6 days after exposure, and young children are among those most likely to become seriously ill.

Helena Bottemiller Evich’s Friday edition of Food Fix made an argument I have been making in pieces all summer, and she made it in one clean sweep: Congress has stopped doing its job. When she had me on her American Dish podcast on Sept. 2 to talk about the Cyclospora outbreak, I said much the same thing: the blame for our food safety troubles rests less with any one administration than with Congress’s decades-long unwillingness to stay on top of food safety policy and to pay for it. She called what we are watching “the slow rot of what was supposed to be our most important branch.” I agree. After more than thirty years of watching Congress respond, or fail to respond, to outbreaks, I want to add what that rot looks like from where I sit.

A 12-to-11 farm bill

On Wednesday the Senate Agriculture Committee advanced the farm bill 12 to 11 on a party-line vote, and it took the return of Sen. Mitch McConnell of Kentucky, back after more than three months away following a fall, to get there. Most of the coverage focused on McConnell. Helena focused on the more important point: that his vote was needed at all.

The contrast with 2018 is stark. That year the Senate committee cleared its farm bill 20 to 1, and the full Senate passed it 86 to 11, the most votes a Senate farm bill had ever received. The farm bill is supposed to be rewritten every five years. The 2018 law expired in 2023 and has been kept alive by extensions ever since, and the current extension runs out Sept. 30. Child nutrition programs, which cover school meals, have not had a full reauthorization since the Healthy, Hunger-Free Kids Act of 2010.

The largest outbreak in three decades, and no hearing

This summer’s Cyclospora outbreak is the one I know best. As of Sept. 15, CDC counted 19,883 lab-confirmed, domestically acquired cases with onsets from May 1 through Aug. 31, including 1,064 hospitalizations and two deaths, in 49 states and the District of Columbia. CDC says it is aware of more than 36,000 cases in all. The same months in 2025 produced 1,180. When CDC declared the outbreak over on Sept. 11, it tied 12,883 illnesses to iceberg lettuce from Taylor Farms de Mexico, which leaves exactly 7,000 confirmed cases with no food named.

Members of Congress noticed. Thirteen letters had gone to FDA, CDC, HHS and Taylor Farms by the end of August. Rep. Debbie Dingell and nine House colleagues sent a fourteenth, to USDA, on Sept. 14. On Sept. 17, Rep. Rosa DeLauro, Senate Democratic Leader Chuck Schumer and 23 other members sent a fifteenth, the first from both chambers, asking USDA to stop closing the Beltsville lab where its Cyclospora research is done. Eight of the fifteen set dated deadlines, and all eight passed without a published answer, as I wrote this week. No committee has held a hearing.

There is a structural reason for that, and it is not a partisan complaint. Every one of those letters came from members of the minority party; the newest carries 25 signatures, and none of them belongs to someone who holds a gavel. A member in the minority can write a letter. Only the majority, through its committee chairs, can call a hearing, issue a subpoena and put a witness under oath. When the gavel stays on the table, the letters are all there is, and agencies and companies have learned that letters can go unanswered.

Fifty-two babies, and no oversight hearing

Helena also pointed to the infant formula botulism outbreaks. By CDC’s count, 48 infants in 17 states were hospitalized in the ByHeart outbreak, the largest infant botulism outbreak in American history, and four more in the Nara Organics outbreak. That is 52 babies. I represent families in both. In late August, whole genome sequencing linked the two outbreaks through a shared dairy ingredient. There has not been a single oversight hearing on either one.

Compare 2022. When the Abbott plant in Sturgis, Michigan, shut down and the formula shortage hit, the House Energy and Commerce oversight subcommittee brought formula company executives in to testify in May, and the FDA commissioner went before the Senate HELP Committee the next day. Compare 2009, when the same House oversight panel called Peanut Corporation of America’s Stewart Parnell to testify within weeks of the recall, and he took the Fifth. Compare July 1998, when the Senate Permanent Subcommittee on Investigations held a hearing on Cyclospora in imported raspberries. That was Congress doing its job.

There is a bill. H.R. 7867 would put C. botulinum on the list of organisms formula makers must test for. The House Health Subcommittee held a legislative hearing on it April 29, and it has waited since for a markup nobody has scheduled. I have written Energy and Commerce and Senate HELP twice each, in June and again in August, and none of the four letters has drawn a reply. HELP Chairman Bill Cassidy did demand answers from Nara and ByHeart, and he deserves credit for it, but a letter is not a markup. This week t-shirts are being hand-delivered to all 77 members of those two committees. I have also turned to a citizen petition asking FDA to require the testing by regulation, with 24 parents of hospitalized infants as co-petitioners. When Congress will not act, the agencies are the only door left open.

When Congress does act, it acts quietly

The most consequential thing Congress did on food safety in the past year it did by rider. Section 780 of the continuing resolution that ended last fall’s shutdown, P.L. 119-37, bars FDA from spending money to administer or enforce the food traceability rule before July 20, 2028. That is the rule built to speed tracing of high-risk foods, leafy greens among them. The Senate Appropriations Committee had written the opposite, language telling FDA not to delay enforcement past July 2028. The House version won. There was no hearing on that choice either, and the lobbying behind it became public only because the Guardian went and read the disclosure filings.

The agencies are paying for it, too

Helena’s reading list makes the rest of the case. STAT reports that the FDA is trying to hire more than 2,000 people after losing about 3,500 in the April 2025 cuts, and that the hiring is stuck in the same administrative bottlenecks the cuts created. Civil Eats reports a GAO finding that the government spent about $9.5 billion in 2025 paying employees on administrative leave, most of it through the deferred resignation program, and that USDA cut about 20 percent of its staff and HHS about 18 percent. Whatever anyone thinks of those decisions, this is exactly what oversight hearings exist for: asking whether the agencies that inspect our food have the people to do it.

What would help

I have no interest in stirring the political pot, and Helena was careful to say the decline has held no matter which party is in power. The fixes are not partisan either.

First, hold the hearings. The House Energy and Commerce Committee, the Senate HELP Committee and the agriculture appropriations subcommittees in both chambers have jurisdiction over this summer’s outbreak and the formula outbreaks. Call CDC, FDA and the companies. Ask where the other 7,000 Cyclospora cases came from, why fifteen letters have no published answer, and why powdered infant formula is still not tested for C. botulinum. If witnesses will not come, subpoena them.

Second, do the routine work. Pass a farm bill instead of a fourth extension. Reauthorize child nutrition programs for the first time since 2010. Fund the FDA on time, so the agency is not planning around the next shutdown.

Third, give the next Congress a to-do list it can act on. I have been circulating a food safety agenda, Project 2027, among consumer and public health groups for comment, and a signed version goes to the new Congress in January. It names specific problems and specific fixes.

Helena asked what disaster it would take to get Congress’s attention. More than 36,000 Cyclospora cases, fifteen letters without a published answer and 52 babies hospitalized with botulism should have been enough. If they were not, the problem is not the size of the disaster. It is an institution that has forgotten one of its core jobs, and the rest of us need to remind it.

On September 15, the California Leafy Greens Marketing Agreement (LGMA) posted a white paper marking twenty years since the 2006 spinach outbreak that led to its creation. It is called “Twenty Years of Progress,” and its executive summary ends on a line I agree with: food safety is never finished.

Four days earlier, CDC declared the 2026 Cyclospora outbreak linked to Taylor Farms de Mexico iceberg lettuce over. CDC’s final count was 12,883 illnesses in 21 states, at least 570 hospitalizations and two deaths in Michigan. FDA says it is still investigating how the lettuce was contaminated. The white paper does not mention Cyclospora, this outbreak, Mexico or iceberg lettuce.

I have sued lettuce companies for nearly thirty years. I also think the LGMA is the organization best placed to keep the next one of these outbreaks from happening. This post is about both: what the LGMA has said this summer, what it has not, and work we could do together.

What the LGMA said this summer

The LGMA was not silent in July and August. Its public record, in order:

  • July 15 and 24. LGMA’s first outbreak statement called lettuce a leading suspect and noted that CDC had discontinued routine national reporting of cyclosporiasis in 2025 after budget cuts. By July 24 a different statement sat at the same web address: sympathy for those sickened, credit to health officials, and acceptance of Mexican-grown lettuce as the likely source. I wrote about the swap in July, and said then that the deleted paragraph was the truest thing any trade group had said about the outbreak.
  • July 20. The California and Arizona LGMAs held a public webinar with food safety scientist Dr. Jennifer McEntire on what is known and unknown about Cyclospora, including why testing produce and water for the parasite is so hard. California LGMA CEO Tim York said the goal was to cut through confusion and misinformation.
  • July 27. York told CBS News that LGMA membership is voluntary, but “once they join, they must follow our rules.”
  • August 9. California LGMA said California lettuce and leafy greens were in full production and not associated with the outbreak, and published a fact sheet on why its members’ greens are low risk for Cyclospora. Arizona LGMA posted its own.
  • August 20. LGMA reported hosting five FDA produce safety staff on an August 12 farm tour in the Salinas Valley.

The webinar and the FDA tour were good work. The paragraph that got deleted was good work too.

What the LGMA has not said

Since August 9, I cannot find anything either LGMA has said publicly about the outbreak, on its website or in the press. There was no statement when CDC closed the outbreak, and none on FDA’s continuing investigation. There has been nothing on the roughly 7,000 laboratory-confirmed, U.S.-acquired cases this season that are not tied to any food. CDC’s case page now counts 19,883 confirmed domestic cases from May 1 through August 31, against the outbreak’s 12,883. The two counts carry as-of dates four days apart.

The white paper names two outbreaks, spinach in 2006 and Yuma romaine in 2018. It never names a pathogen, not even E. coli O157:H7 in 2006. The 2019 and 2020 romaine E. coli outbreaks appear only as timeline entries about “new risks” and “food safety updates.” Yet one of the twenty changes the paper lists is root cause investigation after food safety events.

Why the LGMA is in a unique position to lead

Taylor Farms is an LGMA member. It is on the California LGMA’s 2025 certified member list, it holds a seat and an alternate seat on the Advisory Board, and its people sit on the Communications and Technical Committees. The lettuce that sickened 12,883 people came from Taylor Farms de Mexico, outside the program the LGMA audits. A family buying a bag of shredded iceberg sees a brand, not a farm. For them there is no line between a member’s California lettuce and the same member’s Mexican lettuce. The company’s outbreak and recall history is its own story; the point here is that the grower at the center of this outbreak helps govern the program.

That is the problem, and it is also the opportunity. The LGMA has things no regulator or advocacy group has:

  • Reach. The white paper says LGMA members grow 94 percent of U.S. leafy greens.
  • Speed. Membership is voluntary, compliance is not, and audits are done by the California Department of Food and Agriculture under USDA oversight. The white paper counts more than twenty updates to LGMA standards in twenty years. Federal rules take years each.
  • Data. Through Romaine Test & Learn, members already pool pathogen-testing results.
  • A shared stake. California farmers told Ag Alert in August that the loss of consumer confidence was the worst since the 2006 spinach outbreak. Growers who did nothing wrong paid for this one.

The white paper’s own 2026 strategic plan names public trust building, through accountability and transparency, as one of three priorities. This is the moment for it.

Ten things we can work on together

  1. Close out the outbreak in public. Acknowledge the final count, the people sickened and the two who died and commit to report what members learn from it.
  2. One standard for the label, wherever the lettuce is grown. Members’ leafy greens from outside California and Arizona should meet LGMA-equivalent standards, verified by audits comparable to CDFA’s. Where they do not, buyers should be told which product is outside the program.
  3. Write Cyclospora into the metrics. Cyclospora comes from human feces, so the controls are field sanitation and handwashing, keeping sick workers out of harvest, and protecting water sources. Workers need paid sick leave for staying home to be possible. LGMA’s own webinar stressed that testing is not the answer, which means prevention has to carry the load, and better detection methods deserve research money through the Center for Produce Safety.
  4. Publish the root cause. When FDA finishes its investigation, the LGMA should say publicly what it will change. The public should see root cause findings, not just be told they happen.
  5. Name the grower when the file closes. In November 2024, a romaine E. coli O157:H7 outbreak sickened 89 people in 15 states, hospitalized more than a third and killed one. FDA closed it in February 2025 without naming a grower or processor. On September 10, FDA closed a fourteen-person romaine outbreak, reference #1382, the same way, saying the lettuce was past its shelf life. FDA has named farms, processors and brands in that exact situation before. The LGMA should ask FDA to name growers and processors when traceback reaches them, and members whose product was in a traceback should say so themselves.
  6. Share the data with public health. Expand Romaine Test & Learn to other crops and to water and environmental results, and give FDA and CDC access as results come in.
  7. Bring back the Microbiological Data Program. From 2001 to 2012, USDA’s Microbiological Data Program worked with state agriculture departments to test fresh produce for pathogens. At about $4.5 million a year it was more than 80 percent of government produce testing when its funding ended. The produce industry lobbied to eliminate it. FDA’s Donna Garren has now called for a national approach to produce microbiological data like it. This is surveillance, not testing as prevention, and it tells FDA, CDC and growers where the risk is. LGMA support for its return would carry weight in Congress.
  8. Meet federal traceability rules now. The compliance date for FDA’s food traceability rule under FSMA Section 204 has slipped to July 2028. The white paper says members already keep extensive electronic records. They can meet the rule ahead of the deadline.
  9. Open the neighbor’s gate for E. coli. Leafy greens E. coli investigations keep pointing to nearby cattle, and investigators often cannot get onto adjacent animal operations to sample. I have drafted a citizen petition asking FDA to use its existing public health authority to get that access during outbreak investigations. The LGMA’s support would matter.
  10. Stand up for public health capacity. This industry needs outbreaks found fast and sources named fast. Say so to Congress when CDC, FDA and state laboratory budgets are on the table, and support affordable Cyclospora testing so sick people get counted.

An invitation

I have been on the other side of the table from LGMA members for a long time, and I expect to be again. I would rather not be. The LGMA was built after a tragedy by growers who decided the whole industry had to change because one company’s problem was everyone’s problem. A leafy greens executive says exactly that in the white paper. This summer taught the same lesson again, at more than sixty times the size of the 2006 outbreak.

The LGMA says food safety is never finished. I agree. Let’s finish the next piece of it together, and I would welcome the conversation.

De Ann Davis, senior vice president of science at the Western Growers Association, published a piece on August 31 arguing that FDA outbreak communication is itself a food safety tool and needs to work better. I missed it when it came out. It deserves a reply, because I agree with most of it, and the places where I do not are where the next outbreak will be won or lost.

One thing first. Davis told the Guardian in August that growers have collected shipment data voluntarily for more than a decade and that Western Growers supports the Food Traceability Rule. That put growers on the record on the other side from the retail and foodservice associations that lobbied for delay. Nothing below changes my respect for that.

Where we agree, and it is most of it

Her framework is close to what I have asked for all summer. Every advisory should say what is confirmed, what is reasonably suspected, what is unknown and what would change the agency’s view. As soon as the evidence permits, the FDA should name the commodity, product form, brand or supplier, country and growing region, distribution channel and dates, and when those details are unknown, say so. Corrections should be time-stamped and pushed through the same channels as the original. After the hazard passes, the agency should publish root-cause reports that tell growers, importers and retailers what to change.

She is right that the July 17 naming of Taylor Farms de Mexico was the moment the response started working. It gave restaurants and shoppers a target, and it told everyone else what was not in play. The FDA’s outbreak page still carries that history. She is also right that the FDA should be the hub that separates confirmed outbreaks from concurrent clusters and background surveillance. North Carolina said on July 17 that its increase did not appear connected to the Taco Bell outbreak. That was careful work by a state health department, and no federal page set it beside the national count and explained how the pieces fit.

Where I part company, gently

The first is Michigan. Davis faults the July 4 Michigan advisory for listing produce from past Cyclospora outbreaks, including salad mixes, cilantro, basil, raspberries, snow peas and green onions, when no source had been identified. She calls that a missed opportunity for the FDA, and on that point I agree. The fix is a federal voice, not a quieter state. A health department watching its case count climb with no named food owes the public something, and the history of the parasite was the evidence it had.

The second is what the FDA can honestly say is not implicated. Davis wanted the agency to say this summer that domestic leafy greens had not been implicated, and her September 15 follow-up goes further, arguing that the risk from domestically grown produce is meaningfully lower than from imports. On the history she has a point, and the FDA’s own Cyclospora action plan says most domestically acquired outbreaks have been tied to imported produce. Now look at this year’s arithmetic. The CDC’s case data page, updated September 15, counts 19,883 confirmed domestically acquired cases from May 1 through August 31. The federal outbreak linked to Taylor Farms de Mexico closed at 12,883. That leaves 7,000 confirmed cases with no food named by any agency.

Davis poses three good questions in that follow-up: where the illness was acquired, where the product was grown and where the contamination occurred. For those 7,000 people the second and third answers are blank, because no agency has said what they ate. An advisory that clears a category before the evidence does is its own kind of imprecision, and it is the kind that gets people sick.

The third is her proposed verification hold. Davis wants preliminary laboratory results with public and commercial consequences verified before release whenever consumer protection does not require immediate disclosure. After the FDA announced a positive lettuce sample on July 18 and called it a false positive on July 19, I understand why. The same correction said the epidemiological and traceback evidence still supported the recall. People were protected by the epidemiology, not by the sample. A verification step is reasonable if it has a clock measured in days and the agency says publicly that a result is pending. Without a clock, verify first becomes announce later.

The cost of imprecision is real

I hold an economics degree, and I do not wave these numbers away. Western Growers estimated $125 million in short-term wholesale losses to California strawberry growers after the 2022 hepatitis A outbreak tied to imported berries. An International Food Information Council survey in mid-August found that among adults who had heard about this summer’s outbreak, 30 percent said they had stopped eating lettuce entirely in the previous month. Fewer vegetables on American plates is a public health cost, and Davis is right to say so. Precision is the answer to it. Precision, though, means naming, and naming sometimes means naming somebody you would rather not.

What precision requires

Two things, and Western Growers is in a position to help with both.

The first is traceability. Supplier, growing region and dates, the second item in Davis’s framework, come from records. Section 780 of last November’s continuing resolution barred the FDA from spending money to administer or enforce the traceability rule before July 20, 2028. I made this argument in July when Joelle Mosso of Western Growers wrote about the Cyclospora response, and it has only gotten stronger since. Western Growers already supports the rule. If it said out loud that Section 780 should be unwound, that would carry weight with Congress that my saying it never will.

The second is naming. On September 10 the FDA closed CORE investigation #1382, an E. coli O157:H7 outbreak with 14 sick, naming romaine lettuce as the source and no grower or processor. The FDA’s investigations table is where that closure lives, in a single line. The 89-person romaine outbreak from late 2024 closed the same way. Under Davis’s framework, if traceback reached one grower, the advisory names the grower. I would like to know whether Western Growers will say so when the grower is a member. That is the hardest version of the question, and it decides whether consumers learn which lettuce or only that it was lettuce.

Davis closed her September piece by asking industry, academia and public health to work together on fact-based context before the next Cyclospora season. I am in. Traceability draws a line. It says what everything else is. Without the line, fear prices the whole category, and the whole category pays.

I have spent more than 30 years representing people sickened by contaminated food. I do not often read a speech by an FDA official and find myself agreeing with nearly every line. This week I did.

Dr. Donna Garren, director of the Food and Drug Administration’s Office of Produce Safety, spoke on Sept. 15 at the annual meeting of the National Association of State Departments of Agriculture in Portland, Maine. I could find only two outlets that covered it, Agri-Pulse and Lancaster Farming, and both are worth reading in full. Her message, in short: the 2026 Cyclospora outbreak exposed weaknesses in how we oversee fresh produce, and FDA needs a system that finds hazards before people get sick rather than after.

She has been on the job less than three months. She joined FDA effective June 29, after nearly 15 years at the American Frozen Food Institute and earlier leadership roles at Boskovich Farms, the United Fresh Fruit and Vegetable Association, the National Restaurant Association, the Consumer Goods Forum and NSF International. Her office writes FDA’s regulations, policy and guidance for fresh fruits and vegetables, including the minimum standards under the FSMA Produce Safety Rule. That background matters. When someone who has spent a career working alongside industry says the system has to change, growers and processors will hear it in a way they may not hear it from me.

Never let a good crisis go to waste

The numbers are hard to overstate. According to FDA, CDC reported 19,595 laboratory-confirmed cases of cyclosporiasis in the United States as of Sept. 8, including more than 12,000 from one multistate outbreak linked to iceberg lettuce from Taylor Farms de Mexico. CDC declared that outbreak over on Sept. 11. Two people died.

Garren told the state agriculture officials that the outbreak should be a moment of reckoning for FDA, because it showed a food safety system that does not always reflect the complexity of today’s global supply chain. Then, according to Lancaster Farming, she reached for the old saying: never let a good crisis go to waste. She is exactly right, and that saying is the history of food safety in this country.

Crises are how this system changes. After the 1993 Jack in the Box E. coli outbreak, USDA declared E. coliO157:H7 an adulterant in ground beef in 1994. Hamburger cases were the core of my practice for a decade, and today they are close to zero for my firm. That is the best thing to happen in food safety in my lifetime, and it came out of a crisis. After the 2006 spinach outbreak and the 2008 to 2009 peanut outbreak, Congress passed the Food Safety Modernization Act, signed in January 2011.

Crises also get wasted. Less than two years after FSMA was signed, USDA shut down its produce testing program. Fifteen years after FSMA, the food traceability rule Congress required, the tool that would let investigators trace lettuce faster, has been delayed again after pushback from industry. Attention fades when the case counts stop climbing. Garren is telling her own agency not to let that happen this time.

If we don’t act, shame on us

The sharpest line of the day did not come from Garren. It came from California Secretary of Food and Agriculture Karen Ross, who chairs NASDA’s Food Systems and Nutrition Committee, the committee Garren addressed. Speaking about how communication breaks down in every outbreak, Ross said: “If we don’t act on this, shame on us.”

She is right, and the “us” is broad. It includes FDA and USDA, the growers, processors and retailers, the Congress that writes the budget, the state agriculture departments in that room, and people like me who have been making these arguments for more than thirty years. More than 19,000 people were sickened in one season, more than 12,000 of them in a single outbreak, and two people died. If a year from now all we have is another round of statements, with no testing program, no rule on naming firms and no published root cause, then shame on all of us.

She wants produce testing back

This is the part of her speech I most want to highlight. Lancaster Farming reports that she wants FDA to reinstate a national approach to microbiological data on produce, similar to USDA’s Microbiological Data Program.

I know that program well. USDA ran it from 2001 to 2012, working with state agriculture departments to sample and test fresh fruits and vegetables for foodborne pathogens, and it ended in December 2012 when its funding was eliminated. In its final year it was a $4.5 million program that accounted for more than 80 percent of government produce testing. The Associated Press reported in 2012 that it had found more than two dozen contaminated samples that led to recalls of lettuce, tomatoes and other foods. I argued at the time that ending it was a mistake, and I have said so many times since. Hearing a senior FDA official say publicly that something like it should come back is welcome news.

She wants the data used before people get sick

Garren described a system that brings together surveillance data, inspection records, outbreak histories and even satellite imaging to flag hazards before illnesses occur. That is the right goal. FDA already holds much of this information. All summer I have been tracking FDA’s own outbreak investigations table, where many files list no food, no on-site inspection, or both. The information exists. What has been missing is putting it to work, and putting it in front of the public. A director who wants to do both has my support.

Same rules for everyone

She said domestic growers, foreign farms and importers must be held to the same standard, and that fairness and public health point in the same direction. That principle matters. The Associated Press found that FDA inspections of foreign food sites are down nearly 35 percent since 2019. When I went through FDA’s investigations table in August, eleven of seventeen open investigations showed no on-site inspection, and the six that did all involved firms on American soil. Parity is fair to American farmers and safer for American consumers.

Name them

Here is the line from her speech I would put on the wall. Garren called transparency both a value and a mechanism, and she said: “When bad actors know they will be named … behavior changes.” She added that this means publishing outcomes, sharing data, and showing that oversight is real and applied without exception.

I have been making that argument in outbreak after outbreak. Earlier this month FDA closed a summer E. coliO157:H7 file by saying romaine lettuce caused it, naming no grower and no processor. Stop Foodborne Illness has a citizen petition pending at FDA, Docket FDA-2026-P-1448, on naming the firms tied to outbreaks. Garren’s principle, applied, answers the question that petition asks.

States as partners

She also wants FDA to work more closely with state agriculture departments and to share information better, describing state staff as the boots on the ground who talk to producers every day. The old Microbiological Data Program ran through state agriculture departments, so a closer federal-state partnership is exactly how a revived testing program would work.

How to make it stick

A speech is a start. Three steps would turn what she said into policy: funding for a national produce testing program in the next federal budget, a rule requiring FDA to name the firms linked to outbreaks, and publication of the root cause findings in the iceberg lettuce outbreak, which FDA says it is committed to sharing. Each of those follows directly from her own words, and together they are what not wasting this crisis looks like. If she wants allies in getting them done, she has one here.

Garren told Lancaster Farming that friends say she picked the worst time to join FDA. She disagrees, and said she likes figuring out what happened and building plans that keep it from happening again. After this summer, that is exactly the attitude the job needs. Welcome, Dr. Garren. Keep saying it.

Woods Hill Table, the farm-to-table restaurant on Commonwealth Avenue in West Concord, Massachusetts, closed on September 16, 2026 while town and state health officials investigate a cluster of Salmonella infections. As of September 17, ten people had been reported ill, both customers and restaurant employees, according to Concord Public Health Director Melanie Dineen. She said she had not heard of anyone being hospitalized. The restaurant has not said when it will reopen.

Nearly everything the public knows comes from one place: reporting by Dakota Antelman of The Concord Bridge, which obtained Dineen’s September 15 memo to the Board of Health and interviewed her. The Bridge’s first version of the story said the memo did not name the illness; the updated version, after Dineen spoke with the paper, identifies it as Salmonella. As of this writing, neither the Town of Concord nor the Massachusetts Department of Public Health has posted anything about the cluster.

What the Concord investigation has found so far

According to Dineen, the investigation began on September 8. An initial inspection of the restaurant found no priority violations, and in those early days she did not believe there was a risk to the public. Testing then turned up more cases, including one person who was carrying Salmonella without symptoms. About a week into the investigation, Dineen advised Woods Hill Table to close, and the restaurant agreed.

Her memo says state epidemiologists individually contacted every affected patron and that there is currently no ongoing risk to the general public. It also says management and staff gave officials full access for environmental sampling and employee retraining. Each staff member must have two negative test results before returning to work. Local officials believe this is most likely an isolated incident, but Dineen wrote that more testing is needed to confirm it.

Owner Kristin Canty told the Bridge in an emailed statement that the restaurant takes food safety protocols very seriously and is cooperating with state authorities. Dineen praised the management’s response, saying they are doing everything they are supposed to.

What Concord got right

When employees are among the sick, closing the kitchen and testing everyone who works in it is the right call, and requiring two negative results before anyone returns is how you keep a cluster from restarting a week after the doors reopen. The asymptomatic carrier found here is the reason that rule exists: a food handler who feels fine can still shed the bacteria. As Dineen noted, Salmonella spreads through food and from person to person.

It also matters that the restaurant cooperated, and that the health director said so publicly. In more than thirty years of foodborne illness cases, I have seen the damage done when a business fights the health department instead of working with it. That did not happen here.

What the public still does not know

1. Which serotype? Salmonella has more than two thousand serotypes. Knowing whether this is Enteritidis, Typhimurium or something rarer says a lot about where to look.

2. What dates? Neither the memo nor the Bridge story gives the dates of illness or the meal dates involved. A diner who ate there in early September and was sick but never tested has no way of knowing whether they belong in this count.

3. Who got sick first? If employees fell ill before customers, the likely explanation is an infected worker. If customers and employees fell ill together, a contaminated ingredient served to both becomes the stronger possibility. The order matters.

4. Has the state lab sequenced the isolates? Whole genome sequencing will show whether the ten cases share one strain and whether that strain matches illnesses anywhere else in the country.

5. What did the environmental sampling and food testing find? The memo says sampling was done. The results have not been released.

6. How were the patrons identified? Contacting every affected patron is good work. The public should know whether that list came from reservations and card records, and whether anyone who walked in or paid cash could have been missed.

Why “isolated” has to be proven, not assumed

A restaurant is often the last stop in a supply chain, not the first. The same contaminated ingredient that sickens diners in one dining room can sicken people in other states who have no idea they shared a supplier. The only way to rule that out is laboratory data.

Massachusetts has that data. Salmonella infections are reportable to the state Department of Public Health, and clinical labs must send isolates to the State Public Health Laboratory. A CDC Emerging Infectious Diseases study of Massachusetts surveillance reported that the state averaged about 1,200 confirmed Salmonella cases a year from 2014 through 2018, and that the state lab has reported cluster patterns to CDC’s PulseNet network since 1996, now increasingly by whole genome sequencing. DPH also posts year-to-date foodborne illness counts, currently through September 1, which is too early to show this cluster.

Dineen and the Concord Board of Health have handled this responsibly. The next step is for DPH to say publicly what the sequencing shows. If the Woods Hill Table strain matches nothing else, officials can say so with confidence. If it matches cases elsewhere, the public needs to know which ingredient and which supplier, because the risk will not have ended at a closed door in West Concord. CDC estimates that Salmonella causes about 1.35 million infections, 26,500 hospitalizations and 420 deaths in the United States every year.

If you ate at Woods Hill Table and got sick

Symptoms of salmonellosis include stomach cramps, diarrhea, fever, nausea and vomiting, and some cases are severe enough to require hospitalization. If you ate at the restaurant in September and have been ill, see a health care provider, ask for a stool test, and tell them where you ate. A positive test is what puts your case into the investigation. You can also contact the Concord Public Health Division. Anyone with symptoms should not prepare food for others until they have recovered.

Dineen’s September 15 memo went to the Board of Health, and should become public through the Board’s meeting records or the Great Meadows Public Health Collaborative, where Concord posts agendas and minutes.

Three questions about the Woods Hill Table Salmonella cluster

Why did Woods Hill Table close? Concord’s public health director advised the West Concord restaurant to close about a week into a Salmonella investigation that began September 8, after testing found additional cases. The restaurant agreed and closed September 16.

How many people are sick? Ten as of September 17, including both customers and employees, with no reported hospitalizations.

Is there a risk to the public? Town officials say there is currently no ongoing risk, and state epidemiologists have contacted each affected patron. They describe the cluster as most likely isolated but say more testing is needed to confirm it.

I will update this post when Concord or the Massachusetts Department of Public Health releases the serotype, the dates of illness, or the sequencing results.

The multistate Cyclospora outbreak linked to iceberg lettuce from Taylor Farms de Mexico closed on September 11, 2026 with 12,883 illnesses, 570 hospitalizations and 2 deaths in 21 states. Investigators did not lose the supply chain. They followed it to the supplier within a day and lost it at the last link, the farm field and whatever water touched the lettuce there. That last link is what the federal Food Traceability Rule is built to record, and Congress has told the FDA it may not spend money enforcing that rule until July 20, 2028.

The Packer ran a thoughtful piece Thursday by Jill Dutton with Fabiola Negrón, director of food safety at Registrar Corp. Her diagnosis is sound. People remember meals poorly and weeks after the fact. Records run out a few steps back. A company can keep good internal records that never connect to its growers’ records. Produce with a short shelf life is gone before anyone can test it. On Cyclospora specifically, she said finding the product was the simpler part, and that once investigators knew the supplier, the gap was that the supplier did not know the source.

She is right. The headline, though, says the outbreak ended in a supply chain mystery, and the distinction matters for what gets fixed.

What investigators found, and how fast

The FDA and CDC announced the investigation on July 16. On July 17 the agencies said traceback had converged on a single supplier, Taylor Farms de Mexico, and the company recalled all iceberg lettuce sourced from central Mexico. On July 19 the FDA withdrew a positive lettuce sample as a false positive, and no product sample has since tested positive. The link rests on epidemiology and traceback. In August, CDC added genotyping showing that the parasites from sick people were genetically related, which ties patients to each other and to a single source, though not to a sample of lettuce.

For fairness: Taylor Farms says it raised scientific questions about the government’s assumptions and methodology in its first call with the FDA, CDC and Michigan health officials. Neither agency has changed its conclusion. CDC’s outbreak page still states that the investigation linked the outbreak to processed iceberg lettuce from Taylor Farms de Mexico.

Where the trail went cold

On August 20 the FDA’s inspection moved from the processing plant to iceberg lettuce growers in Mexico. At the close, Food Safety Magazine reported that the FDA had completed inspections and sampling at the growers and the processing facility, with results pending and root cause work continuing. The FDA’s own post-outbreak page says its investigation of the Taylor Farms de Mexico outbreak goes on, along with the investigations into the other clusters this summer.

Consider also the size of the recall: all iceberg lettuce from a region of Mexico. That is my reading, not an agency finding, and a company acting out of caution would recall the same way. Still, a recall drawn around a region instead of a lot or a field is what the gap Negrón describes looks like in practice.

The rule that records the field

Section 204 of the Food Safety Modernization Act, signed in January 2011, told the FDA to propose traceability requirements for high-risk foods by January 2013. The agency missed the deadline and wrote the rule only after the Center for Food Safety sued and a court set a schedule. The final rule came out in November 2022. For foods on the Food Traceability List, which includes leafy greens, it requires key records at each step, including a traceability lot code and where the food was harvested, and it requires companies to produce those records for the FDA in a sortable electronic spreadsheet within 24 hours.

The compliance date was January 20, 2026, nearly five months before the first illness in this outbreak. In March 2025 the FDA announced a 30-month delay. In November 2025 Congress went further. Section 780 of Division B of H.R. 5371, the bill that ended the shutdown, bars the FDA from using appropriated funds to administer or enforce the rule before July 20, 2028, according to a Congressional Research Service report. A delay is an agency deciding industry needs more time. A funding prohibition leaves the agency unable to act even if it changes its mind. The Packer notes the 2028 date without explaining where it came from. Fifteen and a half years separate the statutory deadline from the first day of enforcement.

Nobody can prove after the fact that an enforced rule would have taken investigators to one field this summer. Getting investigators to the field is what the rule is designed to do, and this outbreak is the case it was written for.

The fear she describes is a cost to industry

Negrón made a point the produce industry should take to heart. Frightened shoppers do not read lot numbers. They avoid the whole category until the news fades. The Packer’s own reporting found that this outbreak hurt grocery foot traffic well beyond the produce aisle. Traceability draws a line around the problem, and without that line fear prices the entire category. Growers, packers and retailers who did nothing wrong paid for that this summer.

Some of the associations speaking for those same retailers and distributors asked Congress to wait on the rule, as their own lobbying disclosures show. I wrote about those filings in August. Western Growers told the Guardian that it supports the rule, so industry is not of one mind, and that is worth remembering.

The bigger mystery

The mystery The Packer’s headline promises is real, but it sits elsewhere. CDC now counts 19,883 laboratory-confirmed, domestically acquired _Cyclospora_ cases from May 1 through August 31. Subtract the 12,883 in the lettuce outbreak and 7,000 confirmed cases remain with no food named. Those investigations never got far enough to hit a supply chain wall. On September 15 CDC stopped updating that page weekly. Over the coming months, epidemiologists and I will be working to learn where those people were infected.

Questions readers are asking

Did investigators find the source of the 2026 Cyclospora outbreak?

They found the supplier, Taylor Farms de Mexico, within a day of announcing the investigation. As of the September 11 close, the FDA had not named a farm or a root cause, and sample results from the Mexican growers were pending.

Would the Food Traceability Rule have helped?

The rule requires harvest location and lot codes to travel with leafy greens and requires records within 24 hours. Enforcement is barred until July 20, 2028.

Test the system

Negrón’s advice to companies is to test the traceability system rather than keep a plan on paper. Congress should take the same advice. The country had a system, with a written rule and a compliance date, and Congress took the money away before anyone could test it. Eleven consumer, public health and labor organizations have asked Congress not to push the rule back further. The fastest way to shrink the next supply chain mystery is to let the FDA enforce the rule it already wrote.