On November 6, FDA will hold the second of its public meetings on what it calls Challenges and Solutions in Lot-Level Food Traceability. Anyone who wants to speak must register by October 26, six days from today, and no same-day registration is allowed. Written comments are due December 6 in Docket No. FDA-2014-N-0053. The meeting is billed as a chance to talk about implementation. Read the notice closely and it is something narrower: a conversation about how to make the rule’s core requirement, the lot code, easier to live without.

So here is where the Food Traceability Rule stands, how it got here, and why, more than fifteen years after Congress ordered it, FDA cannot spend a dollar enforcing a single line of it.

What the rule does

Section 204 of the Food Safety Modernization Act, signed in January 2011, told FDA to identify the foods most often tied to outbreaks and to require anyone who manufactures, processes, packs or holds them to keep more and better records. Those foods are on the Food Traceability List: leafy greens, tomatoes, cucumbers, peppers, melons, sprouts, fresh herbs, fresh-cut fruits and vegetables, shell eggs, nut butters, soft cheeses, finfish, crustaceans, some shellfish and ready-to-eat deli salads.

For those foods, the rule requires companies to record key data elements at critical tracking events, harvesting, cooling, initial packing, shipping, receiving and transformation, and to tie them together with a traceability lot code assigned when the food is first packed and carried down the chain. When FDA asks during an outbreak, a covered company must produce an electronic, sortable spreadsheet within 24 hours. FDA estimated the rule would reach more than 323,000 domestic businesses operating about 484,000 establishments, according to the Congressional Research Service.

Before the rule, most food handlers only had to know one step forward and one step back, and farms and restaurants did not have to keep even that. FDA’s own pilot projects, run by the Institute of Food Technologists in 2011 and 2012, found that one-up, one-down records often could not connect the restaurants and grocery stores where people got sick to the farm that grew the food. That is the gap the rule was written to close, and the lot code is what closes it.

The calendar

Congress gave FDA until January 2013 to propose the rule. FDA missed it. The agency did not deliver its report to Congress on the pilot projects until 2016. In 2018 the Center for Food Safety and the Center for Environmental Health sued, and a consent decree set the schedule FDA finally met: a proposed rule on September 23, 2020, and a final rule on November 21, 2022. It took effect January 20, 2023, with compliance set for January 20, 2026. That three-year runway was already a year longer than FDA had proposed, granted because industry asked for it, and at the time FDA said it would go no further. In September 2023 the agency added that it would not begin routine inspections until 2027.

In January 2024 the Government Accountability Office told FDA to finalize and document an implementation plan to hit the 2026 date. Two months later, Rep. Scott Franklin of Florida introduced H.R. 7563, which would have relieved retailers, restaurants and warehouses of the duty to keep and pass along lot codes and required more FDA pilot projects before any enforcement. It never got a vote. Its ideas went into House agriculture appropriations bills instead.

On March 20, 2025, FDA announced it intended to push compliance back 30 months, to July 20, 2028. I wrote at the time that FDA had again kicked the rule down the road. The agency’s stated reason was that even the companies ready for January 2026 depended on supply chain partners who were not, and it said plainly that the extension did not amend, and that it did not intend to amend, the requirements of the rule. The proposed rule to extend the date was published August 7, 2025. As far as I can find, it has never been finalized. It did not need to be.

Congress finished the job. On November 12, 2025, the bill that ended the government shutdown, H.R. 5371(Public Law 119-37), carried Section 780 of Division B, which says no funds appropriated by the Act may be used to administer or enforce the rule before July 20, 2028. The same section told FDA to meet quarterly with farms, restaurants, retailers and warehouses to identify and implement flexibilities for satisfying the lot-level tracking requirement, to give industry guidance on food waste recovery, reclamation, intracompany transfers and customer returns, and to test its own internal tracing system. As I explained in August, that distinction matters. A delay is an agency asking for patience. A funding bar is Congress taking away the agency’s ability to change its mind.

Why it has not been implemented

There are three honest answers, and they overlap.

The first is that part of industry asked for the delay, and its own filings say so. When I went through the federal lobbying disclosures in August, the International Foodservice Distributors Association described its fourth-quarter 2025 work on the rule as support for the 30-month extension and targeted flexibilities. The National Restaurant Association listed H.R. 5371 by number, for provisions regarding FDA traceability, in the quarter it became law. The National Retail Federation reported in 2024 that it sought support for H.R. 7563. Jessica Glenza at the Guardian named Publix among the companies lobbying on the rule.

The members of Congress who carried that ask should be named, because their names are on the record. Rep. Scott Franklin, a Florida Republican who lives in Lakeland, where Publix is headquartered, introduced H.R. 7563 and is vice chair of the House Appropriations Subcommittee on Agriculture, Rural Development, Food and Drug Administration. Rep. Sanford Bishop, a Georgia Democrat, was its lead cosponsor; he is the subcommittee’s ranking member and chaired it in the 117th Congress. Rep. Jimmy Panetta, a California Democrat whose district runs down the Monterey County coast, cosponsored it. And Rep. Andy Harris, a Maryland Republican, chairs the subcommittee whose fiscal 2025 bill and fiscal 2026 bill, H.R. 4121, carried H.R. 7563’s language. The fiscal 2026 version became Section 780. Harris also wrote the fiscal 2027 bill the House passed in June. Two Republicans and two Democrats, three of them sitting where FDA’s money gets written. The bill they introduced died. Its provisions did not; they became law through the one committee where they never needed a hearing, a floor debate or a vote of their own.

The Senate did not go along. In Senate Report 119-37, Senate appropriators directed that FDA may not use funds to delay enforcement beyond July 20, 2028, and that full compliance and enforcement begin by then. Section 780 is a House product.

The second answer is that the industry’s argument is not frivolous, and I should say so. Traceability only works when everyone in a chain is doing it. The 2024 Reagan-Udall Foundation roundtables found low awareness of the rule among small and foreign suppliers and nonchain restaurants, and distributors have real problems with mixed-lot pallets and pick slots. But Kroger, Albertsons and Walmart built programs to the rule before the delay was announced, and Western Growers says its members have collected this shipment data voluntarily for more than a decade. The associations that lobbied for the delay represented the ready and the unready alike. They spoke for the unready.

The third answer is that FDA is not ready either. Section 204 also directed FDA to build an internal system to receive and analyze traceability data. In January 2026 GAO reported that the system still does not exist. FDA told GAO in March 2025 it would be done by January 2026; after the delay, it moved that date to July 2028, too.

What happened while we waited

FDA has now closed its investigation into the largest Cyclospora outbreak in American history: 12,883 confirmed illnesses, 570 hospitalizations and two deaths in 21 states, linked to shredded iceberg lettuce from Taylor Farms de Mexico. Leafy greens are on the Food Traceability List. FDA held its first lot-level public meeting on June 15; CDC puts the first illness on June 14. A Michigan epidemiologist spent the first two weeks of July asking where the lettuce came from. In the meantime, lettuce sales fell, and California growers plowed under healthy crops that had nothing to do with the outbreak. That is what happens when the answer takes twelve days instead of one: the innocent pay for the guilty.

On August 12, eleven consumer, public health and labor organizations in the Safe Food Coalition asked congressional leaders to rescind Section 780. The International Fresh Produce Association says it opposes further delays, and Taylor Farms says it supports traceability legislation. Congress has not acted. The House passed its fiscal 2027 agriculture and FDA bill, H.R. 8646, on June 4 by a vote of 213 to 210, and bill summaries say it carries the same bar on enforcement before July 20, 2028. The Senate has not passed its version.

Where things stand today

On February 19, FDA announced a final exemption for certain cottage cheese, a draft question-and-answer guidance and a series of stakeholder sessions organized by the Partnership for Food Traceability, an industry membership nonprofit. The first session, on March 6, was open only to the partnership’s members.

On October 8, FDA issued draft guidance saying it would not enforce the rule against retail food establishments and restaurants whose average food sales over the previous three years are more than $250,000 but not more than $1 million, adjusted for inflation from 2020. Businesses at $250,000 and under were already exempt. FDA says the new carve-out covers only a small share of food sales while relieving many small operators. Maybe so. It is also the first time the agency has moved the rule’s boundary since Congress told it to look for flexibilities.

And then there is November 6. According to the Federal Register notice and Food Safety Magazine’s summary, FDA wants to hear about a “reasonable range” of traceability lot codes for distributors shipping to retailers; “inferred” lot codes, where a distribution center calculates rather than scans what it shipped; key data elements for individual items no longer in a case; returns and reclamations; food waste recovery; intracompany shipments; retailers that transform food and send it to other stores; and whether industry should settle on a data standard. FDA’s earlier discussion paper and the comments on it are in the docket.

Look at where every one of those topics sits. Distribution centers, grocery stores, restaurants, the back half of the chain, the same link H.R. 7563 tried to cut. The lot code is the rule. A “reasonable range” of lot codes is a range of possible farms, and a range of possible farms is what investigators have today.

Industry, it is your turn

The consumer, public health and labor groups have done their part. Eleven of them signed the August letter. They should not be the only ones asking.

The companies that are already ready have more to gain from this rule than anyone, and they have been far too quiet. So, by name. Kroger, Albertsons and Walmart built traceability programs before the delay was announced; you are the proof that the rule works, and Congress needs to hear it from you. The International Fresh Produce Association says it opposes further delays. Taylor Farms says it supports traceability legislation. Western Growers has already written its members a guidance document on how to comply. The California and Arizona Leafy Greens Marketing Agreements set the food safety rules for the crop at the center of this summer’s outbreak. Put your names on a letter to House and Senate leadership, and to Senators John Hoeven and Jeanne Shaheen, who lead the Senate appropriations subcommittee, asking that Section 780 be rescinded and kept out of the fiscal 2027 bill in conference.

To FMI and the National Grocers Association, which applauded the delay, and to the International Foodservice Distributors Association, the National Restaurant Association and the National Retail Federation, whose filings asked for it: your members include the companies that are ready. Let them speak for themselves on November 6, and do not ask FDA to carve the lot code out of the back half of the chain on their behalf.

And to everyone who grows, packs, ships or sells a food on the list: comply now. Nothing in Section 780 stops a company from keeping lot codes and producing that 24-hour spreadsheet tomorrow. Do it before the 2027 growing season and say so publicly. If the companies already doing this asked Congress to unwind the delay, the argument that industry cannot manage the rule would collapse, because it would be industry saying otherwise. California’s growers learned this summer what it costs when nobody can draw a line around one facility. This rule draws that line.

What I would say on November 6

If you care about this, register to speak before October 26 and file comments by December 6. Here is what I would ask for. No flexibility that lets the lot code stop at the distribution center door. Any flexibility FDA does adopt should first be tested in a real or simulated traceback, the same pilot-project standard the industry demanded of the rule itself. And FDA should repeat in public what it told Congress and the courts: that the full public health benefit of the rule depends on every covered entity complying.

And to Congress: rescind Section 780, keep it out of the fiscal 2027 bill, and let the Senate’s own language, full compliance and enforcement by July 20, 2028, and not a day later, survive conference this time.

Congress ordered this rule in January 2011 and set a deadline of January 2013. If nothing else slips, enforcement begins in July 2028. That is fifteen and a half years from deadline to enforcement. The records the rule requires already exist, in some form, in warehouses and distribution centers across the country. The only question left on November 6 is whether they will follow the food all the way to the people who eat it.

Where the recalled oysters went and where the 27 sick people live. Sources: VDH recall alert (Oct. 9), FDA safety alert and CDC notice (Oct. 19, 2026). Virginia’s 53 in 18 states is not broken out by state and cannot be mapped.

As of October 20, 2026, the Virginia Department of Health counts 53 people in 18 states sick with Salmonellaafter eating raw oysters, with illnesses beginning in late May and running to September 21. CDC’s October 19 notice and FDA’s safety alert, posted the same day, count 27 in 12 states, with five of 21 hospitalized and no deaths, tied to oysters Little Wicomico Oyster Company harvested from the Little Wicomico River. Virginia says the 27 are a genomic subgroup of the 53. Georgia’s Department of Agriculture says Publix sold the recalled oysters as “James River Oysters” through October 14, five days after the recall, and two of CDC’s 27 live in Colorado, where Marler Clark represents a resident who became ill after eating raw oysters.

Twenty-six sick more people are in Virginia’s count and missing from CDC’s and FDA’s.

Virginia’s outbreak page, updated today, says that as of October 14 there were 53 cases in 18 states with onsets from late May through September 21, and that CDC’s 27 cases in 12 states are “a genomic subgroup” within that total. The arithmetic is 53 minus 27: twenty-six people with Salmonella Telelkebir in the same window whose isolates fall outside the subgroup, in six states no federal page names. CDC’s notice and FDA’s safety alert, both posted October 19, five days after Virginia’s count date, carry only the 27 and do not mention the 53. Within the 27, ten reported eating raw oysters; tag and traceback records exist for seven; two of the seven ate oysters from a single Virginia source and five from several; and the company’s Little Wicomico River oysters were “the only oysters in common to the seven ill people included in traceback.” Nothing on any page says whether the other 26 ate oysters or what the sequencing showed. A genomic subgroup is a decision about how tightly to draw the circle. The people outside it are just as sick, and if they ate oysters, the water that made them sick is not in the recall.

The recalled oysters went to sixteen jurisdictions, and the cases from that harvest have not been counted yet.

The gap between the map’s blue states and its red ones is a calendar, not a contradiction. The recall covers oysters harvested September 18 through October 8. CDC’s last illness onset is September 14, four days before the first recalled harvest day. Every one of the 27 counted so far ate oysters harvested before the recall window, which is why I said yesterday that the recall could not help any of them. Anyone sickened by the recalled harvest would have fallen ill from late September on, and CDC says it takes three to four weeks for a stool culture to become a sequenced isolate that is matched to the outbreak and counted. Those people are in the pipeline now. Seven jurisdictions on the receiving list — California, the District of Columbia, Georgia, Missouri, North Carolina, Nevada and Texas — have no case yet because the oysters that went there are the ones whose illnesses are still being confirmed, and because the summer customers who ate the earlier harvest may not be the fall customers who got the recalled one. Expect the blue states to turn red.

The same lag explains part of the 53-versus-27 gap, and the serotype explains the rest. Virginia’s 53 is every Salmonella Telelkebir illness in the window reported to it as the harvest state through the shellfish-sanitation channel, as of October 14. CDC’s 27 is the set whose isolates sequence closely enough together to be called one strain, pulled at an earlier data cutoff and posted October 19. Telelkebir was the serotype in last year’s oyster outbreak too, 80 sick in 23 states with no harvest area ever found, so some of the 26 outside the subgroup may carry a related strain from other water, and five of the seven traced patients ate oysters from more than one source. None of that is in dispute. What is missing is the sentence that says so: that 53 are sick, 27 sequence together, the rest are under review, and more are coming. Virginia wrote half of it. The federal pages wrote none.

The Publix label said James River, not Little Wicomico.

Yesterday I wrote that the consignee list has not been published; every federal page still says only, “restaurants and retailers.” Georgia’s Department of Agriculture recall page, posted October 15, adds a block headed “Information regarding oysters sold at Publix affected by this recall”: product description “James River Oysters,” GTIN 29953900000, purchase window “9/27/2026 thru 10/14/2026,” discard and do not eat, with a link to Publix’s corporate recall notice. Georgia does not say which stores.

The purchase window runs five days past the recall date.

Virginia’s October 9 recall alert asked receiving states to verify that companies “have stopped selling, serving, and distributing the affected oysters.” Publix’s purchase window closes October 14. Neither Georgia’s page nor the Publix notice says when the dealer’s notice reached Publix or whether oysters stayed on sale after it arrived. A shopper who read Virginia’s alert, FDA’s or CDC’s would have found “Little Wicomico” in all three and “James River” in none; the harvest tag stays with the container, and the label goes home.

Colorado has two of the 27, and the oysters were shipped there.

CDC’s map puts two of the 27 in Colorado, and Colorado is on both receiving lists, so no one must reach for travel to explain them. Marler Clark represents a Colorado resident who became ill after eating raw oysters, and our epidemiologists have requested the state and local health department records. Indiana, Iowa and Kentucky, with one or two cases each, are on neither list; travel or the further distribution FDA mentions could explain that, and the agencies have not said which.

StateSickStateSick
Pennsylvania8Connecticut1
New York4Florida1
Virginia3Illinois1
Colorado2Iowa1
Indiana2Kentucky1
Maryland2New Jersey1

Source: CDC, Where People Got Sick, Salmonella outbreak linked to raw oysters, posted Oct. 19, 2026. Check: 8 + 4 + 3 + 2 + 2 + 2 = 21, plus six states with one case each = 27, matching CDC’s subgroup total; Virginia’s 53 in 18 states as of Oct. 14 is not broken out by state.

Virginia pulled 400 oysters from the river on October 13, and the results were due today.

Virginia collected 400 of the company’s oysters from cages in the river on October 13 for Salmonella testing, with results expected October 20; as I write, the page does not report them. Reopening requires three negative rounds, the second about October 21 and the third the week after, and the closure is expected to run 30 days. Pennsylvania’s positive distributor sample is still awaiting sequencing. Publish the 400-oyster result, the sequencing result and the sanitary survey of the river before anyone harvests there again.

The first retailer’s name came from a state agriculture page, ten days before the federal notices.

The first public word was Virginia’s October 9 alert, written for regulators. The first retailer’s name came six days later from a receiving state. The federal notices came four days after that, carrying half of Virginia’s count. Both of my pending citizen petitions ask FDA to publish the consignee list in every recall, so the next shopper does not have to depend on which state she lives in.

If you bought James River Oysters at Publix between September 27 and October 14, do not eat them.

Throw them out or return them. If you ate raw oysters since mid-September and have diarrhea, fever or stomach cramps, see a doctor, say that you ate oysters and ask for a stool culture; the culture is what lets a state laboratory sequence the isolate and count you. Restaurants and retailers should check shellstock tags for VA-1421-SP and harvest dates from September 18 through October 8.

How many people are sick from the Little Wicomico River oysters?

Virginia counts 53 in 18 states as of October 14; CDC’s October 19 notice counts 27 in 12 states, the genomic subgroup tied by traceback to the Little Wicomico River, with five of 21 hospitalized and no deaths. Both counts are expected to grow, since it takes three to four weeks to link a new case.

Were the recalled oysters sold at Publix?

Yes. Georgia’s Department of Agriculture posted on October 15 that Publix sold them as “James River Oysters,” GTIN 29953900000, purchased September 27 through October 14, 2026, and Publix has its own notice under that title.

Which oysters are recalled?

Whole shellstock oysters harvested by Little Wicomico Oyster Company (VA-1421-SP) from the Little Wicomico River and Little Wicomico River Lower between September 18 and October 8, 2026. Other Virginia growing areas are not included.

CDC announced today that a multistate outbreak of Salmonella Telelkebir infections has been traced to raw oysters harvested by the Little Wicomico Oyster Company of Heathsville, Virginia, from the Little Wicomico River, a small tributary of the Potomac on Virginia’s Northern Neck. As of October 19, 27 people in 12 statesare confirmed sick, five of the 21 with information available (24%) have been hospitalized, and no one has died. Illnesses began between May 29 and September 14, 2026. The company has recalled oysters harvested from September 18 through October 8, Virginia closed the river to shellfish harvesting on October 9, and the advice from CDC, FDA and Virginia is the same: do not eat, sell or serve recalled oysters from Little Wicomico Oyster Company (dealer certification VA-1421-SP).

What the agencies know, and how they know it

Of the 27 sick people, 13 have been interviewed, and 10 of them (77%) reported eating raw oysters, against a 1.6% background rate in CDC’s FoodNet Population Survey. Traceback was possible for seven of the ten; their oysters came from several harvest areas, and Little Wicomico River oysters from Little Wicomico Oyster Company were the only ones all seven had in common. Pennsylvania officials also sampled oysters at a distributor that supplied the restaurants where sick people ate; one sample harvested by the company from the Little Wicomico River was positive for Salmonella. Whole genome sequencing to determine whether it matches the outbreak strain is pending.

Where the sick people live: Pennsylvania 8, New York 4, Virginia 3, Colorado 2, Indiana 2, Maryland 2, and one each in Connecticut, Florida, Illinois, Iowa, Kentucky and New Jersey. When they got sick, by month of onset: one in May, one in June, 13 in July, eight in August and four in September, the last on September 14. Patients range in age from 1 to 86, with a median age of 41.

The recall and the closure

The Virginia Department of Health’s Division of Shellfish Safety and Waterborne Hazards issued a recall alerton October 9 for shellstock oysters harvested from the Little Wicomico River and Little Wicomico River Lower from September 18 through October 8. The recall was triggered by an outbreak declared by the Pennsylvania Department of Agriculture after patients were epidemiologically linked to the company’s oysters. Virginia lists distribution to Colorado, Connecticut, the District of Columbia, Florida, Georgia, Illinois, Maryland, Missouri, North Carolina, New Jersey, Nevada, New York, Pennsylvania, Texas and Virginia; FDA adds California and notes the oysters may have been further distributed. The same day, Virginia issued Emergency Condemnation 010A, closing the downstream portion of the river to all shellfish harvesting through November 7, which Food Safety News noted was eight days into Virginia’s fall oyster season. Oysters from other Virginia growing areas are not part of the recall.

What the dates tell us

First, every one of the 27 reported illnesses began before the recall window opened. The last reported onset is September 14; the first recalled harvest date is September 18. Salmonella symptoms begin six hours to six days after exposure, so none of the 27 could have eaten a recalled oyster. The oysters that made them sick were harvested in summer, sold and eaten months before anyone connected the cases to a river. The recall protects whoever would have eaten the next bag; it cannot do anything for the 27. Later illnesses may still surface, since linking a new case to an outbreak takes three to four weeks, and if any fall inside the window the recall will have earned its keep. None of this is a criticism of Virginia or of the company, which acted within a day of the Pennsylvania finding. It is a description of the system we have: the outbreak is over, or nearly so, before the recall exists.

Second, three of the twelve states with sick people, Indiana, Iowa and Kentucky, are not on the distribution list for the recalled oysters. Those patients may have traveled, the company’s summer customers may differ from its fall customers, or the oysters were further distributed, which FDA says may have happened. CDC and FDA have not said which. The consignee list, the distributors, restaurants and retailers that received these oysters, is what I have been pressing FDA to require in every recall. It has not been published here.

Third, the first public word of this outbreak was not a consumer notice. It was Virginia’s October 9 recall alert, written for receiving states and shellfish regulators. Food Safety News reported it on October 12. CDC’s and FDA’s consumer notices arrived today, October 19, almost five months after the first person got sick on May 29.

The second Telelkebir oyster outbreak in a year

Last December I wrote about a Salmonella Telelkebir outbreak linked to raw oysters that CDC closed on February 24, 2026: 80 sick in 23 states, 34 of 68 hospitalized (50%), with illnesses from June 21 to December 22, 2025. Traceback was done for 19 patients, no single common source was found and nothing was recalled. The map of that outbreak also had Pennsylvania (10), New York (7), Virginia (6) and New Jersey (6) at the top. CDC’s notice today does not say whether this year’s strain is related to last year’s. If the same uncommon serotype is back for a second summer in the same part of the country, the first question is where it lives between seasons.

Salmonella in an oyster is not Vibrio in an oyster

Vibrio is a natural resident of warm coastal water. Salmonella is not. It lives in the intestines of people, livestock, birds and wildlife, and it reaches an oyster bed in sewage, septic drainage, farm runoff and droppings. Under the National Shellfish Sanitation Program, growing waters are classified by sanitary survey and fecal coliform standards. A 2005 study in Applied and Environmental Microbiology tested 1,296 market oysters from 36 bays on all three U.S. coasts, all harvested from approved waters, and found Salmonella in 7.4% of them, concentrated in particular bays and far more common in summer than in winter. It found no relationship between fecal coliform levels and the presence of Salmonella, and it concluded that oyster meat should be tested for Salmonella regularly, in every bay open for harvest, rather than relying on coliforms. Twenty-one years later, growing areas are still classified by the indicator, not by testing for the pathogen. Virginia said when it closed the river that it was testing shellfish from the implicated area. Those results, the sanitary survey of the Little Wicomico and the pollution source it points to should be published before the river reopens on November 7, along with FDA’s sequencing result on the Pennsylvania sample.

If you ate raw oysters

If you have eaten raw oysters since mid-September and have diarrhea, fever and stomach cramps, see a doctor and ask for a stool culture; a culture is what connects a patient to an outbreak. Restaurants and retailers should check with their distributors for Little Wicomico Oyster Company (VA-1421-SP) oysters harvested September 18 through October 8 and dispose of them.

Salmonella

Salmonella sickens more than a million Americans every year and is the leading cause of hospitalizations and deaths among foodborne bacteria. Its reach is remarkable—it turns up in produce, poultry, pantry staples, and pet food alike. Marler Clark, The Food Safety Law Firm, is the nation’s leading firm representing victims of Salmonella outbreaks. Our Salmonella lawyers have represented thousands of victims in the only U.S. law practice devoted exclusively to foodborne illness, which has recovered more than $850 million for victims and families nationwide. We have tied Salmonella outbreaks to cantaloupe, tomatoes, ground turkey, salami, sprouts, cereal, peanut butter, and restaurant meals, and we have brought cases against Cargill, ConAgra, Peanut Corporation of America, Sheetz, Taco Bell, Subway, and Wal-Mart. The Peanut Corporation of America outbreak ended in federal criminal convictions, with the company’s owner sentenced to twenty-eight years in prison. We are also pressing the federal government to treat Salmonella as an adulterant on raw chicken, so that contaminated poultry can be kept off store shelves the way dangerous E. coli already is.

If you or a family member became ill with a Salmonella infection—including lasting complications such as reactive arthritis or irritable bowel syndrome (IBS)—after eating contaminated food, contact the Marler Clark Salmonella attorneys for a free case evaluation.

FAQ

What is Salmonella?

Salmonella is a group of bacteria behind one of the most common—and most deadly—foodborne infections in the United States, sickening more than a million people a year. Symptoms usually begin between 6 hours and 6 days after eating contaminated food and include diarrhea, fever, and stomach cramps that last four to seven days.

What foods cause Salmonella outbreaks?

Poultry, eggs, ground beef, produce, peanut butter, flour, and even pet food have all caused outbreaks. Salmonella can survive for months in dry, low-moisture foods.

Can a Salmonella infection cause long-term problems?

Yes. Some victims develop reactive arthritis—joint pain and swelling that can persist for months—or irritable bowel syndrome after the initial illness resolves.

Who is most at risk?

Young children, adults over 65, and people with weakened immune systems are the most likely to be hospitalized or to suffer severe, invasive infection.

Can I sue for a Salmonella infection?

If your illness is linked to a contaminated product or a recognized outbreak, you may have a claim against those in the chain of distribution. A stool culture matching the outbreak strain is powerful evidence.

Additional Resources:

First, a shout-out to Jerry Hagstrom, who has been covering farm and food policy in Washington longer than just about anyone, for his piece in The Fence Post on the International Fresh Produce Association’s push to get members to sign up for its Advocacy Action Network. Jerry reads the fine print so the rest of us don’t have to.

According to Jerry, IFPA’s new chair, Elyse Lipman of Lipman Family Farms, told members in Orlando that surveys show 88% of consumers lost confidence in fruits and vegetables after this year’s Cyclospora outbreak. Her pitch: join the network so the industry can be mobilized quickly to contact Washington on everything from the farm bill to tariffs to programs that increase consumption. “We need to advocate for fresh with one voice,” she said.

I agree. I have spent more than three decades cleaning up after foodborne illness outbreaks, and I want Americans eating more fruits and vegetables, not fewer. My firm represents many of the people sickened in this year’s Cyclospora outbreak, and I can tell you none of them want to be afraid of a salad.

But one line in the story stopped me. The framing was that the industry’s Cyclospora problem was more with consumers and social media than with government. With respect, the problem was with a parasite that made people sick. Consumers did not lose confidence because of TikTok. They lost it because they, or someone they knew, got sick, and because they were too often left guessing which product, which company and which store. IFPA hired a global crisis communications firm and stood up a Cyclospora Rapid Response Task Force. Fine. But you cannot message your way out of a food safety problem.

So, if IFPA is building one voice, here is what I would like to hear it say:

Bring back the Microbiological Data Program. This one should be first on the list, and IFPA is uniquely positioned to make it happen, because IFPA’s predecessors helped kill it.

A little history. In 2001, USDA’s Agricultural Marketing Service launched the Microbiological Data Program (MDP) to collect data on harmful bacteria in fresh produce. State agriculture labs tested fruits and vegetables pulled from more than 600 distribution sites for Salmonella, Listeria, Shigella and pathogenic E. coli. Sampling ran through 11 states representing roughly half the U.S. population, and MDP supplied close to 90 percent of all available bacterial pathogen data on produce and fed CDC’s PulseNet. It cost about $5 million a year.

It worked. MDP testing prompted at least 19 produce recalls in two years, and it helped pinpoint outbreak sources, including Salmonella Javiana in green onions in 2009. That was the problem, at least for some in the industry. United Fresh and other trade groups lobbied for years to eliminate the program, arguing it cost growers millions in recalls and unfairly targeted farmers. The FY2013 budget zeroed it out, and MDP shut down in December 2012, which, by Food Safety News’ analysis, eliminated more than 80 percent of public pathogen testing of produce. FDA never filled the gap. I called for MDP’s return in 2024, and I am calling for it again.

United Fresh and PMA are now IFPA. If the industry wants to tell consumers it is serious about safety, there is no more credible message than asking Congress to restore and fully fund the surveillance program it once lobbied to end, and to expand it to cover parasites like Cyclospora, which MDP never tested for. Independent, random testing is how you find problems before people get sick and how you show the public the produce aisle has nothing to hide.

Traceability, on time. IFPA says it is pushing for implementation of traceability requirements. Good. Say it loudly, and mean the FSMA 204 Food Traceability Rule, with no more delays. The faster an outbreak is traced, the shorter the hit to the entire category.

Transparency, not secrecy. Nothing kills consumer confidence like a public health notice about an “unnamed” product. I am finalizing citizen petitions asking FDA to stop treating the identity of outbreak-implicated firms as confidential and to require retail consignee lists in Class I recalls. If consumers know exactly what to avoid, they don’t have to avoid everything. I would welcome IFPA’s comments and support.

Fund the people who solve outbreaks. CDC, FDA and state and local public health labs are the industry’s best friends when something goes wrong. Cutting them is a false economy that the produce aisle pays for.

Prevention on the farm. Clean water, clean hands, and audits that actually mean something. The cheapest outbreak is the one that never happens.

The fastest way to win back that 88% is to make sure there is not a next time, and when there is, to tell people quickly and specifically what to throw out. If the Advocacy Action Network wants to carry that message to Washington, sign me up. My number has not changed.

Most of the email that lands in my inbox about E. coli O157:H7 is not funny. It is a mother describing her child on dialysis. It is a health department records request coming back half redacted. It is a defense lawyer explaining that the hamburger was cooked just fine.

This morning was different.

The email came from Ashleigh Diehl, Director of Marketing at GIANTmicrobes in Stamford, Connecticut. It opened with “Hi William” – a name used mostly by my mother and opposing counsel – and said she had just read my recent post, FDA Has Seventeen Open Outbreak Investigations, 14,197 Sick and Two Dead. Ten Name No Food. To her credit, she got the point. Seventeen open FDA files, ten with no food named, an egg investigation FDA calls complete on its own advisory while its table still lists it as active, and USDA’s first active investigation of 2026 being the joint E. coli file. She called it “a snapshot of how murky the attribution still is.” That is about right.

Then came the pitch. GIANTmicrobes makes an E. coli plush, and she thought it might be a fun, visual add to a future story – “or just make you laugh.”

It did.

So, who are these folks?

GIANTmicrobes makes stuffed toys of microbes – and these days cells, organs and other health-related things. The company was founded in 2002 by Drew Oliver, who, according to a 2020 profile by the National Medical Library of the Czech Republic’s museum, began designing plush germs in 2001 while he was a student at the University of Chicago Law School. So a law student who went into microbes. I went the other direction.

Oliver told Nature Microbiology in 2016 that part of the idea came from Richard Feynman’s description of watching a microscopic organism struggle to survive in an evaporating drop of water, and part from a P.J. O’Rourke piece lamenting that he had no way to explain to his daughter what was going on during a stomach ache. Scholastic tells it more simply: he made a stuffed common cold to teach his kids why to wash their hands.

Andrew Klein has been the company’s president since 2015. He has talked to Infection Control Today about infection preventionists using the plush to educate staff and patients, and in her email Ms. Diehl offered him up to talk about why humor helps health education and how plushies can help with vaccination rates.

Per the company, it has sold more than 15 million of these in over 30 countries, in museums, hospitals, bookshops, toy stores, Amazon and Target, and has shown up on NPR’s Wait Wait…Don’t Tell Me, CNN and Fox News, and in The Atlantic and Wired. They also launched a plush SARS-CoV-2 on March 19, 2020, which not everyone found charming.

And E. coli is hardly lonely. The company says it offers more than 150 different microbes and cells, in sizes from three-inch keychains to two-foot “Gigantics.” For those of us in the foodborne illness business, the lineup reads like my case list: there is a Salmonella (“Which came first: the chicken or the egg or the Salmonella?”), a Listeria, a Norovirus pitched as a cruise-ship stowaway, and a C. diff. They even sell a Germs Deluxe 10-packthat bundles E. coli, Salmonella, Listeria and Norovirus with Ebola, measles, flu, the common cold and a sore throat. Beyond the germs there are cells, organs, MRSA and, per their own blog, a leprosy plush that once won something called March Microbe Madness.

Why I don’t mind a cuddly E. coli

I have spent more than thirty years trying to get people to understand something they cannot see. The kids I have represented with hemolytic uremic syndrome did not get sick from a cartoon with big eyes. E. coliO157:H7 does not smile; it produces Shiga toxin and goes after the kidneys.

But if a plush bacterium gets a seven-year-old to wash her hands, or gets a parent to ask why ground beef needs to hit 160°F, or why raw milk and sprouts are a bad bet, I am for it. Fear does not teach people much. Understanding does. And you cannot understand something you can’t picture.

My only suggestion for the product line: a plush for the food that FDA has not named. Ten of the seventeen open investigations would need one. And since Cyclospora accounts for 13,195 of the 14,197 sick on FDA’s table, if they ever make a Cyclospora, I will take that too.

Ashleigh, send the E. coli. It can sit on my desk next to the outbreak tables. It will be the only thing on that desk that makes me laugh.

A tip of the pen to Dr. Kalmia Kniel of the University of Delaware, whose editorial in ASM Food Microbiology, “The 2026 Cyclospora outbreaks: a wake-up call for food safety, surveillance, and prevention,” was published October 12. It is a scientist’s account of this summer: the strange biology of the parasite, why fresh produce stays vulnerable, and why finding the source takes so long. It is open access, and it should be required reading for every regulator, grower, processor and buyer of leafy greens before the 2027 season.

Dr. Kniel ends with five questions. How did a human parasite reach a lettuce field? What safeguards failed? How do we get better traceback? How do we strengthen surveillance? How can industry stop contamination before it happens? I have been trying to answer those questions since July. Here is where I have landed, question by question. But first, the numbers, because they are what make the questions urgent.

The numbers

Dr. Kniel wrote with the late-August figures. The count kept climbing. When CDC stopped updating its case data page weekly on September 15, it had recorded 19,883 laboratory-confirmed, domestically acquired cases with onset between May 1 and August 31, 1,064 hospitalizations and two deaths, in 49 states and the District of Columbia. In the same stretch of 2025, there were 1,180. That is nearly seventeen times as many.

That is only the confirmed domestic count. CDC listed another 9,765 laboratory-confirmed cases still waiting to be sorted into domestic or travel-related, and at least 6,594 more that were not lab-confirmed or were pending interview. Put those together and CDC knew of more than 36,000 cases in some form, on top of 2,078 people who picked up the parasite abroad.

The multistate outbreak tied to iceberg lettuce from Taylor Farms de Mexico closed at 12,883 illnesses, 570 hospitalizations and two deaths across twenty-one states. Subtract that from the season and roughly 7,000 confirmed domestic cases have no food named beside them. Dr. Kniel is right that this was almost certainly several outbreaks at once, and the clusters that remain on FDA’s table deserve answers too.

And all that undercounts. CDC has a number for how much. In 2011, Elaine Scallan and colleagues published “Foodborne Illness Acquired in the United States—Major Pathogens” in CDC’s own journal, Emerging Infectious Diseases (vol. 17, no. 1, pp. 7–15). It is the study behind the federal estimates of how many Americans get foodborne illness each year. For each pathogen, the authors calculated how many illnesses never get counted because people do not see a doctor, are not asked for a stool sample, or are not tested for the right bug. For Cyclospora cayetanensis, the underdiagnosis multiplier is 83.1, with an underreporting multiplier of 1.0. In plain terms: for every laboratory-confirmed case, about 83 people are sick.

Here is the math. All 19,883 confirmed domestic cases this season, times 83.1, is about 1.65 million infections. The 12,883 tied to the Taylor Farms de Mexico lettuce, times 83.1, is about 1.07 million. The roughly 7,000 confirmed cases with no food named, times 83.1, is about 582,000.

For perspective, Scallan’s team started from 239 laboratory-confirmed cases a year and estimated about 19,808 total Cyclospora illnesses annually in the United States. This summer the confirmed count alone, 19,883, passed that estimate of everyone who gets sick in a typical year.

Treat the multiplied figures as orders of magnitude, not counts. The multiplier was built before multiplex stool panels made Cyclospora testing more common, and testing this summer was unusually aggressive, so 83.1 probably runs high today. Cut it in half and the season is still about 826,000 infections. Cut it to a quarter and it is about 413,000. However, you slice it, hundreds of thousands of Americans were sickened, and almost none of them will appear in any government table.

1. How did a human parasite reach a lettuce field?

Dr. Kniel explains the biology better than I can. Cyclospora lives only in people. Infected people shed oocysts that are not yet infectious; they have to mature in the environment first. That is why person-to-person spread is so unlikely, and why investigators look at agricultural water, worker sanitation, septic leaks, sewage and irrigation systems.

When FDA closed its investigation on October 2, it said where it found the parasite: in a tank of wastewater leaving the Taylor Farms de Mexico plant that shredded the lettuce, and in a drainage ditch at a grower that supplied it. That is the same plant at Doctor Mora, Guanajuato, tied to the 2013 salad mix outbreak, when Taylor Farms said it was testing the water. (A disclosure: my firm represents people sickened this summer and has filed cases naming Taylor Farms entities.)

My answer: treat human waste as the hazard it is. Verify toilets, handwashing and worker health on every leafy greens farm supplying U.S. buyers, using auditors who are not paid to pass. Keep processor wastewater away from any water that touches a crop, test it for parasites, and publish the results. And bring back enforceable numeric standards for pre-harvest water on leafy greens, with parasite testing in the growing regions where outbreaks keep happening.

2. What safeguards failed?

The rules are on the books. The Produce Safety Rule covers agricultural water, worker hygiene and sanitation. The preventive controls rule requires supplier programs. The Foreign Supplier Verification Program is supposed to make imported produce meet U.S. standards. In September FDA sent a letter to industryreminding everyone of all three.

What failed is enforcement and the absence of a backstop. Section 780 of the November 2025 appropriations law restricts FDA’s money to enforce the Produce Safety Rule and its pre-harvest water requirements, and bars spending to enforce the traceability rule before July 2028. And there is no validated kill step for Cyclospora on ready-to-eat produce. Dr. Kniel says the same thing in plainer terms: rinsing cannot reliably remove the oocysts, and cooking is what works. Nobody cooks the lettuce on a taco.

My answer: lift the spending restrictions. Hold imported greens to exactly the same standard as greens grown in California and Arizona, backed by on-site FDA inspections. And because there is no kill step, limit how much product from different fields is commingled on one line, keep lots small, and hold and test finished product by lot using FDA’s validated method before it ships.

3. How do we get better traceback?

Dr. Kniel predicts greater investment in digital traceability. I hope she is right, because Congress has gone the other way. FDA’s Food Traceability Rule covers leafy greens and was supposed to take effect in January 2026; compliance was pushed to July 2028. FSMA ordered FDA in 2011 to build a system to receive traceability data, and a Congressional Research Service report says it still does not exist.

My answer: rescind the delay, as the Safe Food Coalition has asked, and in the meantime every company handling leafy greens should comply now. Agree on a standing protocol among growers, processors, chains and grocers to hand sourcing and shipping data to FDA and CDC within hours. And tell the public what investigators know: my FDA citizen petitions ask the agency to stop treating the names of outbreak-implicated firms as confidential and to publish retail consignee lists in Class I recalls, as USDA has done for meat and poultry since 2008.

4. How do we strengthen surveillance?

Dr. Kniel credits the surveillance system with spotting the rise, and it deserves credit. But we entered the worst Cyclospora year on record with less of it. On July 1, 2025, CDC made FoodNet reporting optional for Cyclospora and five other pathogens. It still takes about six weeks to know whether a sick person belongs to an outbreak, which is why the recall came after most people had already eaten the lettuce. And USDA’s Microbiological Data Program, which sampled produce in commerce, was killed in 2012 to save about $5 million.

My answer: put Cyclospora back on the mandatory FoodNet list. Fund the state and local labs and the epidemiologists who interview the sick. Make Cyclospora genotyping routine and push it to state labs, since it is the one tool that does not depend on someone remembering a June salad. Set a public target for shrinking the six-week lag. And rebuild MDP, this time with a parasite on the panel.

5. How can industry prevent contamination before it happens?

Dr. Kniel points out that there is still no cell culture or animal model for Cyclospora, and she calls for academic, government and industry scientists to work together on the gaps. Yet Congress did not fund two federal Cyclospora research projects in the FY2026 agriculture bill, and a third is being moved away from the scientists running it, as reported in August.

My answer: restore that research money. Fund and validate a detection method growers can actually run on irrigation water, sediment and soil at commercial speed, because no one can manage a hazard they cannot measure. And fund the search for a treatment step that kills parasites on lettuce without destroying it.

What FDA has promised

On October 2, FDA announced ten activities it plans before the 2027 season, including an Outbreak Investigation Report, more surveillance of commodities historically linked to Cyclospora, grower outreach and an update to its Cyclospora Prevention, Response and Research Action Plan. That is welcome. It needs dates, dollars and a commitment to publish what investigators did not find as well as what they did. The Action Plan existed before this summer, and so did the outbreak.

I laid out a longer list in September, ten things FDA should do before next May, and last week a list focused on leafy greens. None of it requires new science to start. It requires money and will.

Congress wrote. Nobody answered.

Dr. Kniel’s editorial stands out partly because so many have moved on. Members of the House and Senate sent at least fifteen letters about this outbreak to HHS, FDA, CDC, USDA and Taylor Farms, asking about the source, the staffing, the surveillance cuts and the research money. The deadlines came and went. What answers there were showed up in press statements and on a company FAQ page, not in any published reply. There has been no hearing. Fiscal 2027 began October 1, and Congress can restore the Cyclospora research and lift the traceability delay in the final funding bill without writing another letter. The people who got sick deserve a hearing on the record before the next crop comes up.

The next one

Dr. Kniel writes that “prevention, rather than reaction, remains the cornerstone of food safety,” and that the goal is not to solve the last outbreak but to prevent the next one. I could not agree more. The next growing season in central Mexico is already under way. Nearly 20,000 confirmed illnesses, more than a thousand hospitalizations and two deaths are the price of reacting. Her five questions are the right ones. The people who can answer them have about six months.

As of October 18, 2026, the FDA’s table of active outbreak investigations still lists 50 people sick with E. coli O157:H7 in the outbreak tied to ready-to-eat, multi-ingredient chicken wraps, FDA reference #1421 and FSIS 2026-04. The row has not changed since October 15. The product column still reads “Multi-ingredient wrap with chicken (suspect).” There is still no recall, no brand, no manufacturer, no list of states and no hospitalization or HUS count. And FSIS, the agency with jurisdiction over a product made with chicken, still has no outbreak page of its own, only a link back to the FDA’s table.

The supply chain, however, is talking. It just isn’t talking to the public. My client in Pembroke, North Carolina, one of the two women in the first lawsuit filed in this outbreak, was contacted by the quality systems coordinator of a New Jersey fresh-food manufacturer whose product line includes prepared wraps and sandwiches. She bought Tarragon Chicken Salad Wraps at a Trader Joe’s in Myrtle Beach, was hospitalized with E. coli O157:H7, and was classified by her state health department as a confirmed case in this multistate outbreak. A manufacturer’s quality department does not go looking for someone like her without a reason.

The government knows where these wraps were made

The FDA’s table shows that traceback, on-site inspection and sampling have all been initiated. Chicken wraps made under federal inspection carry an establishment number inside the USDA mark of inspection. A traceback from a sick person’s wrap ends at that number, and an on-site inspection happens at a specific plant. The agencies are not withholding the manufacturer’s name because they don’t know it. They are withholding it because, in their view, the wraps were pulled from the market before the outbreak link was confirmed and are past their shelf life, so there is nothing for consumers to do.

The name is getting out anyway, just not from the agencies

The first brand on the record, Trader Joe’s, came from a federal complaint and from the plaintiffs’ own health department records, not from the FDA or FSIS. Since we filed on October 15, the name has spread through Patch, WPDE, Hoodline and Yahoo. Patch noted that the FDA’s public listing does not identify Trader Joe’s or any brand, and Hoodline reported that Trader Joe’s did not immediately respond to the suit. Now a manufacturer’s quality staff is reaching out to the people who got sick.

That is backwards. Lawyers, reporters and a company’s own quality department are filling a gap the public health agencies chose to leave open. The information is arriving piecemeal and unverified, and without the facts only the agencies can supply: production dates, lots, distribution, and whether anything else was made with the same ingredients. As I wrote in Saturday’s Weekend Wrap-Up, the first brand name came from a courthouse. The second may come from a manufacturer reaching out to a patient.

Why the manufacturer matters more than the retailer

Retailers like Trader Joe’s rarely make their own prepared foods. Co-packers do, and a co-packer that makes wraps for one retailer often makes salads, sandwiches and wraps for several others, sometimes on the same lines and with the same ingredients. The agencies say they are looking at possible ingredient-level contamination, and Food Safety News reports that USDA is trying to determine whether the chicken went into other products. If the problem was the chicken, the tarragon chicken salad, or something else handled at that plant, the question is not just which wrap made people sick. It is what else came off that line, under that establishment number, and where it went.

When FSIS announces a recall, it names the establishment, prints the establishment number, and posts a retail distribution list. A product that is “proactively removed” before a link is confirmed gets none of that. The establishment number stays in an agency file, and the public is left to piece it together from lawsuits and news stories.

This is what my petitions are about

My two draft citizen petitions, one on disclosing the firms implicated in outbreak investigations and one on publishing consignee lists that show where product went, are still being edited and remain open for comment. I laid out how this outbreak bears on both in an earlier post. This week’s development makes the point more sharply. The identity of the firm is not a secret inside the industry. Trader Joe’s knows who made its wraps. The manufacturer knows who it sold them to. The agencies know both. The only people who don’t know are the 50 confirmed patients, the many more who were never tested, and the doctors treating them.

What should happen now

FSIS should name the establishment and its number, say whether any other products were made with the same chicken or the same chicken salad, and publish the states, the hospitalizations and the HUS count. Trader Joe’s should tell its customers who made the Tarragon Chicken Salad Wraps and which stores sold them. And if the New Jersey company that reached out to Ms. Gottlieb made these wraps, it should say so publicly instead of contacting the sick one person at a time.

Marler Clark represents Ms. Gottlieb and Meaghan Egan Bright of Somerville, Massachusetts, and is investigating other illnesses in this outbreak. If you ate a prepared chicken wrap in August or September and developed diarrhea, especially bloody diarrhea, see a doctor, ask for a stool test for Shiga toxin-producing E. coli, and report the illness to your local health department.

A tip of the pen to Gina Shaw, whose piece in the October issue of Infectious Disease Special Edition, “Invaders From Planet Lettuce,” asks the right question: how did produce become the vehicle that makes so many of us sick? It is the clearest explanation I have read of why this summer’s Cyclospora season was the largest ever recorded in the United States, and it is written for the doctors who must diagnose it. It deserves a wider audience, especially among the people who can do something about it.

Shaw lays out the scale. By CDC’s early September update, more than 18,000 laboratory-confirmed cases had been acquired in the U.S. since May 1, with thousands more awaiting confirmation, against 1,180 in all of 2025. Most trace to the outbreak linked to iceberg lettuce from Taylor Farms de Mexico, which FDA closed at 12,883 illnesses, 570 hospitalizations and two deaths. Michigan alone reported more than 14,000 cases, in a state that saw 50 the year before.

Her sources, Dr. David Freedman of the University of Alabama at Birmingham and Dr. Craig Hedberg of the University of Minnesota, explain why this one got so big. Cyclospora used to ride in on raspberries, basil and cilantro, foods most people eat rarely or in small amounts. This time it rode in on lettuce, which about 80 percent of Americans eat in any given week, served through a fast-food chain. As Dr. Freedman put it, “Now the lettuce is bringing it to us.” Industrial processing then does the rest: millions of heads mixed in one plant, so a few bad ones contaminate thousands of bags. And the parasite survives for weeks in soil and water and shrugs off chlorine washes.

Then Shaw describes a system under strain. The outbreak hit a year after CDC made FoodNet surveillance optional for Cyclospora and five other pathogens. Case reports piled up faster than short-staffed health departments could interview patients, which slowed the search for sources. Michigan pulled people from other programs to keep up. And unlike Salmonella or E. coli, there is no real-time genetic fingerprinting for Cyclospora, so investigators could not always tell whether a patient belonged to the big outbreak or one of the smaller clusters. Her table of open FDA investigations as of late September is sobering on its own: sprouts, raw milk cheese, jalapeños, eggs, infant formula and a long run of outbreaks with the food listed as unknown.

Shaw ends with Dr. Freedman’s warning that preventing a repeat in 2027 will take more inspection and mitigation of at-risk produce, and more people to do it, than we have now. I agree. So here is my list, focused on the food that made so many of us sick this summer: leafy greens.

What needs to be done about leafy greens

  1. Unwind the traceability delay. FDA’s Food Traceability Rule covers lettuce, spinach and other leafy greens, and it was supposed to take effect in January 2026. Compliance was pushed to July 2028, and Congress wrote the delay into law. The Safe Food Coalition has asked Congress to rescind it. It should. In the meantime, every grower, processor, distributor and restaurant chain handling leafy greens should comply now. This summer, perfectly good lettuce was plowed under because nobody could say quickly where the bad lettuce went.
  2. Fix the water in the fields and the plants. In 2024 FDA replaced numeric water testing with “systems-based” assessments. For a crop grown close to the ground and eaten raw, that is not enough. We need enforceable numeric standards and routine testing for pre-harvest water on leafy greens; parasite testing in irrigation and wash water in the growing regions where outbreaks have occurred; processor wastewater controls (FDA found Cyclospora in a Taylor Farms de Mexico wastewater tank, and Taylor Farms promised water testing there in 2013); and research money for a wash or treatment step that actually kills parasites on lettuce without destroying it.
  3. Break up the batches and test before it ships. Shaw’s sources explain how a single plant mixing millions of heads lets a few bad ones contaminate thousands of bags. Processors should limit how much product from different fields and growers is commingled on one line or shift, keep lots small enough to trace and recall precisely, and test finished product by lot, holding each lot until results come back negative, using FDA’s validated methods for the pathogens of concern, including Cyclospora. Testing will not catch every contaminated bag, but small, held lots shrink an outbreak when one gets through and give investigators a lot number instead of a guess.
  4. Keep cattle away from lettuce. Setbacks between animal feeding operations and leafy greens fields, and the canals that irrigate them, should be required by rule, not left to buyer guidance. The 2018 Yuma romaine outbreak traced back to irrigation canal water near a large feedlot, and the Salinas Valley outbreaks that followed pointed to cattle on adjacent land. We keep finding the same source and keep acting surprised.
  5. Hold imported greens to the same standard. Lettuce grown in Mexico for U.S. buyers should meet the same water, setback and sanitation requirements as lettuce grown in California and Arizona, enforced through the Foreign Supplier Verification Program and backed by on-site FDA inspections.
  6. Verify sanitation on the farm. Toilets, handwashing, worker health and water at leafy greens farms supplying U.S. buyers should be checked by auditors who are not paid to pass. Training is not verification.
  7. Rebuild the trip wire. USDA’s Microbiological Data Program sampled lettuce, spinach and other produce in commerce for about $5 million a year and triggered dozens of recalls before it was killed in 2012. Bring it back, with leafy greens at the top of the list. It is the cheapest insurance the produce industry will ever be offered.
  8. Count the sick again. In July 2025, CDC made FoodNet reporting optional for Cyclospora and several other pathogens. The largest Cyclospora outbreak in history arrived a year later, on a food 80 percent of us eat every week. When the vehicle is that common, every day of delay means more people sick. Restore full FoodNet surveillance, fund the state and local labs and epidemiologists who interview the sick and find the source, and pay for real-time Cyclospora genotyping so investigators know which cases belong to which outbreak.
  9. Agree on data sharing before the next outbreak. A standing protocol, worked out now among leafy greens growers, processors, fast-food chains and grocers, for handing shipping, sourcing and test data to CDC and FDA within hours, not days.
  10. Publish root cause analyses and solve the unsolved. FDA has promised a report on the lettuce outbreak. Every leafy greens outbreak should get one, published promptly and in full, and the four other Cyclospora outbreaks this year that still list the source as “not yet identified” deserve an explanation of why.
  11. Stop rolling back FSMA. The Produce Safety Rule is the floor for every leafy greens farm in the country. Every “burden” removed in the name of deregulation was a tool someone needed this summer.
  12. Name the firms, publish the consignee lists and let FDA sample where the evidence points. I have drafted three FDA citizen petitions that go straight at how leafy greens outbreaks are handled: one asking FDA to stop treating the identity of outbreak-implicated firms as confidential, so people know which lettuce to throw out; one asking it to publish retail consignee lists in Class I recalls, as USDA has done for meat and poultry since 2008, so people know which stores sold it; and one asking it to use its communicable disease authority to require sampling at animal feeding operations implicated in multistate outbreaks, so investigators can finally test the cattle next to the field.

The cheapest outbreak

Gina Shaw’s headline calls it the rise of foodborne illness. It does not have to keep rising. None of the items on my list requires new science. Most require money, will and the willingness to make the people who grow, process and sell our leafy greens do things differently. Dr. Freedman is right that whether we can get it done before next summer is the big question. The 2027 growing season is about six months away. The cheapest outbreak is the one that never happens.

Saturday, October 17, 2026

Last Saturday’s edition closed on a seed lot with no named supplier, a chicken wrap outbreak with no brand, and 619 sick on FDA’s files with no food named. This week the wrap file reached 50 and the first brand on the record came from a federal complaint, not an agency page; Oregon told its doctors about a multistate Salmonella Enteritidis outbreak and no one else; FDA explained three closed outbreaks six weeks to nine months late; the ledger rose to 626; and a 99-person Salmonella Oranienburg outbreak ended with no food named. The through-line of this fifteenth column: the information exists. It reached the public through the wrong door, or not at all.

The chicken wrap file reached 50 sick on Thursday, and the first brand name on the public record came from a courthouse. FDA’s outbreak table, re-stamped October 15, moved reference 1421 — the E. coli O157:H7 file shared with USDA’s Food Safety and Inspection Service as 2026-04 — from 38 to 50; it was 25 on September 23. Product: multi-ingredient wrap with chicken, suspect; traceback, inspection and sampling checked; recall blank; no FDA advisory, no CDC notice; USDA checking whether the chicken went into other products. The same Thursday, Marler Clark filed Gottlieb and Bright v. Trader Joe’s Company, No. 2:26-cv-12017, Central District of California, the first lawsuit in the outbreak, for two women — Gayle Gottlieb of Pembroke, North Carolina, and Meaghan Egan Bright of Somerville, Massachusetts — hospitalized after eating Trader Joe’s Tarragon Chicken Salad Wraps bought August 23 and 24; the complaint alleges Trader Joe’s pulled the wraps before the link was confirmed, and as of Friday the company had not answered the reporters who asked. Why the name still matters, and what the two pending petitions would have done about it, is Friday’s post.

Oregon told its doctors about a multistate outbreak and told no one else. Deschutes County, with the Oregon Health Authority, sent a Health Alert Network advisory to clinicians in three Central Oregon counties about a multistate Salmonella Enteritidis outbreak — exposure histories wanted on eggs, raw milk, animal contact and restaurants; reflex cultures after every positive PCR — with no case count, no food, no other states, no onset dates, and no public notice from the county, the state, CDC or FDA as of Friday. The public learned of it from Food Poison Journal Sunday and Food Safety News Monday. CDC’s current outbreaks page, updated Wednesday, counts fifteen active multistate Salmonella investigations, seven E. coli, six Listeria and two Campylobacter — thirty in all. Monday’s post is here.

FDA explained three closed outbreaks this week, and every explanation arrived long after the file closed. Thursday’s table moved the egg, moringa and infant formula files to Closed 2026 with an Executive Incident Summary Abstract each. The egg abstract — 134 sick in 18 states, 34 hospitalized, 1,589,577 dozen eggs recalled — closes on a Form 483 with three Egg Rule observations at Midwest Poultry Services, the plainest that a henhouse which tested positive during the prior flock was not cleaned and disinfected before new hens went in; Food Poison Journal has the detail. The moringa abstract: a file opened in January, 131 sick in 38 states, 36 hospitalized, four Salmonella strains, three recalling brands, a Form 483 with thirteen observations, a contract manufacturer and foreign supplier redacted even now. The infant formula abstract: the three infants sickened since April after eating Nara Organics powdered formula got lots made with fluid milk from Organic West Milk, spray-dried by Dairy Farmers of America; the ingredient suppliers overlapped with the November 2025 ByHeart outbreak; the Washington infant’s isolate sits five SNPs from six ByHeart isolates; not enough evidence to confirm sources and routes of contamination. That is the record behind the draft petition for Clostridium botulinum testing of powdered formula and its dairy ingredients, open through October 30, and Friday’s two questions on every ingredient and every tanker of raw milk.

The no-food ledger rose to 626, and a 99-person outbreak ended with no food ever named. Nine rows on FDA’s active table carry a case count and no product: Listeria 1415 at 11 and 1417 at 46, up two; Salmonella Newport 1403 at 116; Salmonella I 4,[5],12:i:- 1402 at 43; Cyclospora 1398 at 22, 1392 at 240, adjusted up five, and 1385 at 38; Listeria 1379 at 11; and Salmonella Oranienburg 1387 at 99, opened July 8, traceback, inspection and sampling run, now simply ended. Five of the nine read Ended with no food named, 442 sick among them. Listeria 1415 has started sample collection. Last Saturday the nine rows summed to 619.

The sourced outbreaks did not move, except for one line about Korea. Evergreen Fresh Sprouts broccoli sprouts stand where CDC left them September 24: 32 sick, three hospitalized, six states, 24 in Washington. FDA’s advisory, current to October 9, lists all three retail recalls of seed lot W4105 — EAK Distribution, Eden Brothers and Gardeners Basics — and not the supplier. Alfalfa sprouts hold at 76 sick in 16 states, six hospitalized, 27 in Wisconsin, unchanged since October 1. Graziers raw milk cheese holds at 13 sick, eight hospitalized, three with HUS, nine states, more than half five or younger; the October 9 re-stamp of FDA’s advisory added one undated section — the recalled cheeses were also shipped to the Republic of Korea — and FDA is now collecting samples. Pet turtles: 20 sick in 12 states since September 18. Comments on the draft raw sprout and raw milk cheese warning petitions closed Thursday; the last call is here.

Three state clusters are running on local reporting alone, and a fourth got its final report.Virginia closed the Little Wicomico River in Northumberland County to shellfish harvest from October 9 through November 7 after Pennsylvania traced at least five Salmonella cases to its oysters and clams; the closure order cites FDA’s recommendation and names no harvester, Pennsylvania has issued nothing, and Food Poison Journal has what is public. In Lebanon, Tennessee, the state calls the illness that sent 19 Walmart fulfillment-center workers to hospitals October 11 a possible foodborne outbreak associated with a meal; 911 recordings point at homemade rice, no pathogen is named, and Food Poison Journal followed it. Monongalia County, West Virginia, which reported 16 lab-confirmed Salmonella cases in two weeks at the start of the month, has posted nothing since. Maricopa County’s final report on the 2025 Arizona State Fair outbreak — 18 sick, 11 hospitalized, six with hemolytic uremic syndrome, a petting zoo odds ratio of 190, and 73 percent of the animal-area visitors using hand sanitizer — is in Monday’s post.

Six food warning letters posted Tuesday, and the one from Bakersfield describes five years of positive swabs. FDA’s September 29 letter to Dreyer’s Grand Ice Cream, the Froneri plant in Bakersfield: the company’s own environmental monitoring from 2021 through February 5, 2026 showed ongoing findings of Listeria monocytogenes — 19 Listeria positives and eight confirmed L. monocytogenes in the first five weeks of 2026 alone — the Round Top Drumsticks hazard analysis did not treat environmental Listeria as a hazard needing a preventive control, and the ready-to-eat cones, sandwiches and fruit pops are adulterated; no recall, no illness, no finished-product result. The August 5 letter to Lewis Sausage Company of Burgaw, North Carolina — Phyllis Entis identifies it as the contract manufacturer FDA redacted from Viva Raw’s letter — rests on the August 2025 state samples that found Salmonella Enteritidis and Listeria monocytogenes in raw pet food.

The recalls ledger, with the standing credit to Phyllis Entis at eFoodAlert. Family Farmstead Dairy of Worcester, New York, recalled its 8-ounce Raw Mild Cheddar, lot and best-by 7/22/27, on Thursday; the first version named E. coli O157:H7 and grocers in New York and New Jersey, and the corrected notice posted the same day names O26:H11 and O45:H2 — two strains, from two samples at a Maryland laboratory in FDA’s state sampling program — and seven states; no illnesses; Thursday’s note and Food Safety News carry the detail. Krasnyi Oktyabr of Brooklyn, a week after its Starfish Seafood trio, recalled Starfish cold-smoked Norwegian salmon Friday — lot D080-148, best before November 27, 2027, six states — after New York agriculture sampling found Listeria monocytogenes; no illnesses. Entis’s October 10, 11–12, 13, 14 and 15–16ledgers carried the foreign items — Canadian and Québec cheeses for Listeria and Shiga toxin E. coli, and the United Kingdom’s second egg strain, 36 cases traced by sampling to a Ukrainian packer while the main imported-egg outbreak passes 650 sick. No FSIS recall or alert surfaced in any named outlet; FSIS’s pages would not load from here.

The lettuce outbreak’s after-action moved to the people writing about it, and the argument is published separately. Consumer Reports’ Paris Martineau wrote the clearest outside account of the outbreak and its four gaps, and Tuesday’s post adds to each; IFPA’s Cathy Burns gave the industry’s State of the Industry version in Orlando; FAO’s fact sheet says washing and chlorine do not reliably remove the parasite, which Wednesday’s post takes upstream to Guanajuato; the counts as of October 15 are unchanged at 12,883 in the outbreak and 19,883 in CDC’s season; and what the beef industry did after 2002 and what Salinas said in 2007 are linked, not repeated.

The policy week was short, and the one durable item came from FSIS. Food Safety News reports FSIS will take its Listeria environmental swabbing program nationwide between January and March 2027 — four swabs, two on food-contact surfaces and two off, every time an inspector samples ready-to-eat product at a federally inspected plant, whole genome sequencing on every Listeria monocytogenes positive — after a pilot in the Alameda, Chicago and Raleigh districts whose results have not been published. Judge Vince Chhabria on Tuesday stayed USDA’s reorganization indefinitely; FSIS is not among the components listed. The Senate is out until November 9, so Heidi Overton’s FDA nomination waits; FDA’s virtual lot-level traceability meeting is November 6, speaker registration closing October 26. Friday’s Publisher’s Platform puts all seven petitions — formula botulism testing, raw milk cheese and sprout warnings, Salmonella as an adulterant in poultry, sampling at animal feeding operations, naming implicated firms, and consignee lists — in one place, formula-petition comments open through October 30.

That’s the week — 50 sick from a chicken wrap whose name is in a federal complaint and not on a federal page; an Oregon advisory the public was not meant to see; three outbreaks explained after they closed; 626 sick on nine files with no food; a 99-person outbreak that ended with no answer; and a raw cheddar whose serotype changed between the morning and the afternoon. 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.

Last week I posted a draft citizen petition asking FDA to add botulinum neurotoxin-producing clostridia to the short list of organisms that powdered infant formula must be tested for before it ships, and to require that the dairy ingredients formula is made from be tested before they are used. Comments are due by Friday, October 30, 2026. Thank you to everyone who has already written. Two of the comments raise questions I do not think I should answer alone, so I am putting them to all of you.

https://www.marlerblog.com/files/2026/10/C.-botulinum-Petition-10.15.26-WDM-1.pdf

The petition will not change on these points for now. It asks FDA for what the record from the ByHeart and Nara outbreaks supports, and I want to keep it that way. But that record was built after fifty-two babies were hospitalized, and the next outbreak will not wait for a better one. If there is a case for reaching further upstream, I want to hear it before I file, not after.

Question one: should testing reach every ingredient, not only dairy?

The draft requires lot-by-lot testing of dairy-based ingredients that have not been through a validated spore-reduction step: whole milk powder, nonfat dry milk, milk protein concentrate, whey and its concentrates and hydrolysates. That is where the evidence points. FDA’s inspections found the organism in a lot of organic whole milk powder at the dryer, and when the agency went looking in powdered milk this February, it found a spore-former carrying a botulinum toxin B1 gene in two of 48 samples on the first pass.

But spores live in soil, not only in milk. Plant proteins, starches and carbohydrate powders can carry them too, and a rule that says “dairy-based” invites arguments about what counts. On the other side, ByHeart’s own testing of its non-dairy ingredients, reported to the July roundtable, came back negative, and requiring a 200-gram test of every lot of a vitamin premix that makes up a fraction of a percent of the can is a lot of cost for little protection. One middle path would presume that every ingredient needs testing unless the manufacturer documents in writing why a particular one does not and keep those records available to FDA.

So: which non-dairy ingredients, if any, belong in a testing requirement? Is a written-justification approach workable, or does it hand the decision right back to the companies that made it before?

Question two: should the milk be tested before it is ever dried?

One commenter argued that the draft does not go far enough upstream. In that reader’s view, any company drying milk into whole milk powder or other dairy powders for infant formula should test every incoming tanker of raw milk for botulinum neurotoxin-producing clostridia, on the theory that the smaller and more targeted the haystack, the better the odds of finding the needle.

The idea has real force. A tanker comes from a handful of farms on one route. Once it is pumped into a silo and dried with dozens of other loads, a contaminated delivery is diluted across a large powder lot, and the link back to the farm is gone. Reporters found that FDA traced the milk in these outbreaks to 22 dairies that have never been publicly named. Testing at the tanker would tie any positive to a small set of farms, keep that milk out of infant formula, and generate the farm-level data that nobody has today.

It also raises hard questions, and I would rather hear them answered by people who run dryers and labs than guess at them myself:

Timing. The confirmed methods for this organism take days. Raw milk cannot sit in a tanker for days waiting on a result. Does tanker testing work as a gate, or only to hold and trace the powder made from that milk until results come back?

Sensitivity. Drying concentrates whatever is in the milk, so a gram of powder represents several grams of milk. Is a sample from a tanker more likely, or less likely, to find low-level, unevenly spread spores than a sample from the powder?

Who is responsible. FDA’s infant formula rules govern formula manufacturers, not the dryers that supply them. Should the requirement sit with the dryer directly, or with the formula maker, as a condition its suppliers must meet before their powder is used?

Cost and volume. How many tankers a day does a dryer serving infant formula receive, and what would testing each one cost against what one outbreak costs?

How to weigh in

Send comments to me at bmarler@marlerclark.com, by Friday, October 30, 2026. If you work at a dryer, a formula plant or a testing lab, tell me so; that experience is exactly what these questions need. With your permission, I may quote you, with or without your name, in the petition docket or in a follow-up post. FDA has said, in its July 13 letter, that it wants to work with industry on best practices. Here is a place to start.