Eggs dated into August, frozen berries marked for 2028 and unresolved outbreaks involving hundreds of patients expose a serious weakness in America’s food-warning system

A bag of frozen blueberries implicated in an outbreak of potentially dangerous E. coli carried a best-before date of 9 February 2028.

More than 1.5 million dozen eggs recalled during a Salmonella outbreak carried sell-by or best-by dates extending to 17 August 2026.

Iceberg lettuce connected to one of the largest publicly identified American foodborne illness clusters of the year was supplied not only to supermarkets, but also to restaurants and food-service businesses across at least 27 states.

These were not three isolated recall notices.

They formed part of a crowded federal outbreak table that, as of 4 August 2026, included several simultaneous investigations in which officials had not publicly identified the food at all. One involved 212 reported Salmonella Javiana cases. Another involved 93 Cyclospora cases. A separate Listeria monocytogenes investigation involved nine cases and had triggered traceback, inspection and sampling activity without a named product appearing on the FDA’s public table. (fda.gov)

For consumers, this creates two different risks.

The first is physical: recalled food may remain in refrigerators, freezers, cupboards or commercial kitchens long after the original news alert.

The second is informational: the United States publishes a large volume of food-safety data, but divides it across outbreak tables, individual advisories, company recall announcements, enforcement reports, retailer pages, state notices and separate FDA, CDC and USDA systems.

The result is an American warning system rich in information but poor at answering the question that matters most inside a household:

Is there something in my kitchen that I should not eat?

What the current outbreak data shows

The FDA’s Coordinated Outbreak Response and Evaluation Network, known as CORE, maintains a table of investigations involving FDA-regulated food.

Its public table is not a conventional list of supermarket recalls. It also contains emerging disease clusters, investigations whose source is unknown, outbreaks that have ended but remain administratively active and cases in which traceback or sampling is taking place without a consumer-facing advisory.

On 4 August, the table showed the following recent investigations.

FDA reference	Pathogen	Publicly identified food	Reported cases	Public status
1395	Salmonella Javiana	Not identified	212	Ongoing
1390	Cyclospora	Iceberg lettuce	At least 1,947	Ongoing
1392	Cyclospora	Not identified	93	Ongoing
1378	Salmonella Enteritidis	Shell eggs	98	Ongoing
1387	Salmonella Oranienburg	Not identified	90	Ongoing
1389	E. coli O145:H28	Frozen blueberries and mixed berries	12	Ongoing
1385	Cyclospora	Not identified	10	Ongoing
1379	Listeria monocytogenes	Not identified	9	Ongoing
1383	Clostridium botulinum	Powdered infant formula	See advisory	Ongoing
1380	Listeria monocytogenes	Requesón or soft ricotta-style cheese	See advisory	Ongoing

The table also contained several investigations marked as ended but not yet moved into the FDA’s closed-investigation section. (fda.gov)

This distinction matters.

An outbreak can be epidemiologically over while regulatory, sampling or prevention work continues. Conversely, an investigation can be active without there being enough product information for a consumer to search a refrigerator or scan a barcode.

Among the recent entries carrying numerical case totals on the table, at least six investigations involved foods that had not been publicly identified. Together, the listed unidentified-food clusters accounted for hundreds of reported illnesses, although the cases cannot simply be added into one national outbreak because they involve different pathogens and separate investigations. (fda.gov)

That is the central limitation of the present system: the federal government can know that an outbreak exists before it knows—or before it can publicly disclose—what consumers should avoid.

Three recalls, three different kinds of household risk

The egg, berry and lettuce investigations illustrate why recall relevance should not be measured only by the date an announcement appeared.

A better measure would consider:

how long the product remains usable;
whether consumers commonly freeze or store it;
whether packaging carries searchable codes;
whether distribution was retail, restaurant or food service;
how widely the product travelled;
whether all implicated sources have been identified;
and whether the recall later expanded.

A household-persistence assessment

Product	Storage profile	Latest relevant date	Geographic reach	Hospitalisation rate	Household persistence
Frozen GreenWise berries	Frozen; long-life	Best by 9 February 2028 on original lot; later recall expanded to all lots of four products	Eight states	33.3%	Very high
Midwest Poultry eggs	Refrigerated	Sell-by or best-by dates to 17 August 2026	Retail or food service in at least six states; illnesses in 17 states	26.5%	High until dates expire
Taylor Farms iceberg lettuce	Fresh; short-life	Marketside dates to 3 August 2026	At least 27 states, including restaurants and retailers	Approximately 5.0% using the published 98 hospitalisations and 1,947 cases	Falling in homes, but food-service exposure and delayed symptoms remain relevant

The “household persistence” classification above is an editorial assessment based on stated storage type, date marking, recall breadth and distribution. It is not an FDA risk classification.

Hospitalisation percentages were calculated from the most recent published case and hospitalisation totals. They should not be interpreted as the probability that any individual consumer will require hospital treatment. Surveillance is incomplete, milder illnesses are less likely to be diagnosed and the denominator may change as investigations develop. (U.S. Food and Drug Administration)

1. Frozen berries: the recall that could remain relevant into 2028

The frozen-berry investigation presents the clearest example of a recall outliving its news cycle.

Federal and state officials connected an outbreak of E. coli O145:H28 to GreenWise organic frozen blueberries supplied by Frutas y Hortalizas del Sur S.A. in Chile and sold through Publix.

The first recall, announced on 3 July, covered a 10-ounce GreenWise Organic IQF Blueberries product carrying lot code 60401, printed as “6 040 01”, and a best-by date of 9 February 2028. (U.S. Food and Drug Administration)

That date is crucial. It means the original product could reasonably remain inside a freezer for approximately 18 months after the outbreak announcement.

But the recall did not remain limited to that single lot.

On 29 July, Publix expanded the action to all lots of four GreenWise products:

Product	Size	UPC
GreenWise Organic Whole Blueberries	10 oz	41415-06453
GreenWise Organic Whole Blueberries	48 oz	41415-12053
GreenWise Organic Whole Mixed Berries	10 oz	41415-06753
GreenWise Organic Whole Mixed Berries	48 oz	41415-12153

The products were supplied to Publix stores in Alabama, Florida, Georgia, Kentucky, North Carolina, South Carolina, Tennessee and Virginia. Publix had conducted an internal stop-sale by the end of June, but the later public recall covered products that had already reached consumers. (U.S. Food and Drug Administration)

FDA said its remote assessment of the Chilean supplier produced concerns about at least one additional lot used in other GreenWise products. Products from the firm were subsequently added to an import alert, allowing relevant frozen blueberries offered for entry into the United States to be detained without physical examination. (U.S. Food and Drug Administration)

The published outbreak totals were small compared with the lettuce cluster: 12 illnesses, four hospitalisations and no deaths, with cases reported in Florida and Georgia.

Yet the hospitalisation rate among reported cases was approximately 33%.

Seven of the nine patients interviewed reported eating frozen blueberries, and five specifically reported the GreenWise product. Officials said they were still working to determine the source of contamination. (U.S. Food and Drug Administration)

This illustrates a recurring problem in foodborne epidemiology.

The number of confirmed patients does not necessarily measure the full reach of the implicated product. A comparatively small cluster can involve food distributed across several states and stored for years.

For a household checker, frozen berries should therefore rank near the top—not because they produced the largest outbreak, but because they have the greatest potential to remain unnoticed.

A consumer who discarded the outer packaging, poured the fruit into a reusable container or combined it with another bag may no longer possess the lot code or UPC needed to make a confident comparison. FDA’s own guidance says that consumers who froze a product without its original packaging and cannot determine whether it is recalled should throw it away. (U.S. Food and Drug Administration)

2. Eggs: 19 million eggs and a recall date still in the future

The scale of the shell-egg recall was much larger.

Midwest Poultry Services recalled 1,589,577 dozen eggs on 22 July because of possible Salmonella Enteritidis contamination. That is equivalent to approximately 19.1 million individual eggs. The products included white eggs and brown cage-free eggs sold in cartons ranging from six to 60 eggs, as well as bulk supplies intended for commercial use. (CDC)

They were produced at Texas farms between 6 June and 3 July and carried sell-by or best-by dates between 20 July and 17 August 2026.

For consumers checking a carton, the critical identifiers are:

plant code P-1950 or 0840962;
a Julian date from 157 through 184;
and one of the listed product or UPC combinations.

Recalled brands included Kroger, Simple Truth, Brookshire’s, Country Morning and Cal-Maine Sunups. The full FDA table contained more than 20 retail product configurations, plus bulk food-service egg formats for which no consumer UPC was available. (U.S. Food and Drug Administration)

CDC reported that recalled eggs had gone to food-service and retail customers in Texas, Oklahoma and Louisiana. Consumer availability included Kroger stores in Texas and Louisiana; Brookshire Grocery stores in Texas, Oklahoma, Arkansas and Louisiana; and smaller outlets in those states as well as Mississippi and New Mexico. (CDC)

FDA’s confirmed product-distribution list named Arkansas, Louisiana, Mississippi, New Mexico, Oklahoma and Texas, while warning that distribution could have extended further. Illnesses were reported in 17 states, extending far beyond the confirmed retail footprint. (U.S. Food and Drug Administration)

That difference does not prove unreported retail distribution. People travel, eat in restaurants and consume food away from the state where it was purchased. It does, however, show why a simple state-only recall search can be misleading.

The outbreak involved:

98 reported illnesses;
26 hospitalisations;
no reported deaths;
and a last recorded illness onset of 30 June.

That gives a reported hospitalisation rate of approximately 26.5%. (U.S. Food and Drug Administration)

Laboratory, epidemiological and traceback evidence implicated Midwest Poultry Services eggs as a likely source. Samples collected at Texas farms tested positive for Salmonella, and whole-genome sequencing found that some matched the outbreak strain.

FDA nevertheless said the producer did not account for all reported illnesses and that investigators were examining whether other sources existed. (U.S. Food and Drug Administration)

That qualification should be prominent in any consumer tool.

The public message is not simply “one producer caused 98 cases”. The more accurate formulation is:

The recalled eggs were identified as a likely outbreak source, but federal investigators said they did not explain every reported illness.

This is precisely the type of nuance that disappears when recall announcements are condensed into headline alerts.

3. Iceberg lettuce: the largest outbreak, but the shortest household shelf life

The iceberg-lettuce outbreak produced by far the greatest number of reported illnesses.

By 24 July, FDA listed:

1,947 illnesses;
98 hospitalisations;
cases in nine states;
and confirmed product distribution across 28 states.

CDC described distribution to consumers, restaurants and retailers in at least 27 states. The agencies’ totals are not necessarily contradictory: they may reflect different update times, counting methods or wording, but a live checker should flag the difference and timestamp each source. (U.S. Food and Drug Administration)

Taylor Farms de Mexico recalled all iceberg lettuce sourced from central Mexico on 17 July.

Consumer-facing products included Marketside iceberg salad and shredded lettuce sold at Walmart, with “best if used by” dates running from 18 July through 3 August 2026. The wider action also covered multiple food-service lots and package formats. (U.S. Food and Drug Administration)

By 4 August, those labelled dates had passed.

That greatly reduces the probability that an unopened retail bag remains in an ordinary refrigerator. It does not make the investigation irrelevant.

There are at least four reasons.

First, Cyclospora symptoms commonly begin around one week after infection, but CDC gives a possible range from two days to two weeks or longer. Untreated illness can persist for weeks. (CDC)

Second, the product was sold through restaurants and food-service channels. Consumers may have been exposed without ever seeing a package, brand, lot code or recall notice.

Third, the investigation grew beyond its initial five-state Taco Bell cluster. FDA’s first detailed update said 1,644 people reporting Taco Bell exposure had been identified in Indiana, Kentucky, Michigan, Ohio and West Virginia. By 24 July, the wider outbreak total had reached 1,947 cases across nine states. (U.S. Food and Drug Administration)

Fourth, the scientific evidence was more complex than a positive product test.

FDA initially reported a preliminary positive finding from a lettuce sample collected at the border. Laboratory experts then reviewed the result and concluded that it did not represent true amplification and should be treated as a false positive.

As of 19 July, FDA said there were no confirmed positive product samples for Cyclospora.

The agency did not withdraw its outbreak conclusion. It said traceback and epidemiological evidence continued to converge on shredded iceberg lettuce from Taylor Farms and growers in central Mexico. In one ingredient-level analysis, 90% of interviewed Taco Bell-associated patients reported eating iceberg lettuce. (U.S. Food and Drug Administration)

This is a valuable example of why a negative or invalidated food test does not automatically disprove an epidemiological link.

Food contamination may be intermittent. The particular unit sampled may not contain the pathogen. The contaminated portion may already have been eaten. Some organisms are technically difficult to detect. Investigators therefore combine interviews, purchase histories, restaurant records, traceback information and laboratory evidence rather than relying on a single sample.

The unresolved death-count discrepancy

The Cyclospora figures require a publication warning.

CDC’s page, updated on 3 August, states that the agency is aware of two cyclosporiasis-related deaths in Michigan among people with underlying health conditions and says its website will be updated to include laboratory-confirmed cases, hospitalisations and deaths.

However, the “Fast Facts” section on the same page still lists zero deaths.

FDA’s latest accessible count also lists zero deaths. (CDC)

A responsible publication should not report either number as fully settled.

The clearest wording is:

Federal outbreak totals continued to list zero deaths, but CDC said on 3 August that it was aware of two related deaths in Michigan and had not yet incorporated them into its confirmed headline totals.

This article should be updated when CDC reconciles the figures.

Unknown food, known outbreak

A recall checker can only search a package when investigators have identified a food.

Several current investigations offered no such option.

As of 4 August, FDA’s table included:

212 Salmonella Javiana cases linked to an unidentified food;
93 Cyclospora cases linked to an unidentified food;
90 Salmonella Oranienburg cases linked to an unidentified food;
ten Cyclospora cases linked to an unidentified food;
and nine Listeria monocytogenes cases linked to an unidentified food.

Traceback had been initiated in each. Sampling had begun in the Salmonella Javiana and Listeria investigations, while FDA also recorded an inspection in the Listeria cluster. (fda.gov)

The agency normally withholds product details at this stage for a defensible reason: the evidence may not yet justify naming a food or company.

Premature identification could mislead consumers, damage an uninvolved business and distort interviews with patients.

But the current interface leaves the public with little context beyond pathogen, case count and investigatory tick marks.

A better system would explain, in plain language:

whether patients are concentrated in particular states;
whether interviews point towards a common food category;
when illnesses began;
whether cases are still occurring;
what phase of investigation has been reached;
why no product has been named;
and what evidence would be required before a public warning is issued.

The absence of a named food is not proof that consumers face no continuing exposure. It means officials have not publicly established a sufficiently specific product link.

The FDA has data. It does not have one consumer journey

The United States does not lack recall information.

FDA’s openFDA food-enforcement API contains publicly releasable recall records from the Recall Enterprise System from 2004 to the present and is updated weekly. It provides machine-readable JSON data suitable for external databases and applications. (Open FDA)

The weekly Enforcement Report goes beyond prominent press announcements. FDA states that not every recall appears on its media-release page, but all FDA-monitored recalls are placed in the Enforcement Report after classification. (Open FDA)

The problem is fragmentation.

A consumer investigating one item may need to consult:

FDA’s CORE outbreak table;
an FDA outbreak advisory;
CDC’s outbreak page;
a company recall announcement;
the weekly Enforcement Report;
a supermarket’s notice;
a state health department;
USDA’s FSIS database if the product contains meat, poultry or certain egg products;
and possibly a restaurant or food-service notice.

These sources serve different regulatory and epidemiological purposes. They were not designed as one unified household search tool.

A single checker should not replace them. It should index them.

How America compares with other developed food-warning systems

No peer country reviewed for this article offers a perfect household recall tool. Several, however, provide useful features that the United States could combine.

United Kingdom: a documented public alerts API

The UK Food Standards Agency provides a public Food Alerts API covering current and recent:

allergy alerts;
product recall information notices;
and food alerts for action.

The data can be queried and reused, giving publishers and developers a supported route to build localised services on top of the official feed. (Food Data Cymru)

This is a meaningful advantage for third-party tool development.

The United States also supplies machine-readable recall data through openFDA, but FDA outbreak information and CDC epidemiological updates are not presented as one equivalent consumer API.

Canada: one cross-government recall search and multiple alert channels

Canada operates a combined Recalls and Safety Alerts service incorporating products overseen by the Canadian Food Inspection Agency, Health Canada and Transport Canada.

The public can search notices, subscribe by email and use RSS feeds. Current alerts display product imagery, publication dates and categories within a unified government interface. (Recalls Canada)

CFIA says it oversees approximately 155 food-recall incidents annually, posts food recalls on the central service and issues media warnings for high-risk recalls. (Canadian Food Inspection Agency)

Canada also distinguishes among public communications more explicitly. Its process describes:

a food recall warning;
a food safety warning when a recall cannot be conducted;
a consumer advisory;
and “in distribution” actions affecting hotels, restaurants, institutions, retailers or warehouses.

That vocabulary provides a clearer route for explaining whether a product reached consumers directly or remained within commercial distribution. (Canadian Food Inspection Agency)

Australia: clearer state and date presentation

Food Standards Australia New Zealand publishes searchable consumer recall alerts displaying the relevant states or territories, product date markings, the reason for recall and consumer instructions.

For example, a July 2026 abalone recall displayed the affected state, a best-before date extending to May 2027, the specific biotoxin hazard and advice to return the product. (Food Standards Australia New Zealand)

FSANZ also publishes structured recall statistics. It recorded 92 recalls in 2025, compared with a ten-year average of 87. Undeclared allergens accounted for 38% of the year’s recalls. Across 2021–2025, Australia recorded 91 microbial-contamination recalls; Listeria accounted for 36, Salmonella for 18 and E. coli for 13. (Food Standards Australia New Zealand)

The statistical reporting does not solve product-level lookup by itself, but it enables scrutiny of longer-term patterns that is difficult for a typical consumer to derive from individual notices.

European Union: cross-border notification history

The EU’s Rapid Alert System for Food and Feed, or RASFF, connects national authorities and publishes a consumer portal for recalls and public health warnings linked to RASFF notifications.

Its public search system provides recent notifications and an archive currently covering records from 2020 onwards. (Food Safety)

RASFF is especially useful where food crosses national borders. A notification can show the reporting country, countries to which products were distributed, actions taken and related warnings.

Its weakness for household use is similar to America’s: regulatory notification data can be technical, and the precise consumer-facing recall may still reside on a national website in another language or format.

Comparative assessment

Feature	United States	United Kingdom	Canada	Australia	European Union
Searchable public recall notices	Yes	Yes	Yes	Yes	Yes
Official machine-readable recall feed	Yes, openFDA enforcement API	Yes, FSA alerts API	Government alert feeds available, plus RSS	Public web listings and statistics	Searchable RASFF system
Outbreak case totals alongside recalls	Yes, but split between FDA and CDC	Not consistently in recall feed	Often split between CFIA and public-health notices	Not consistently in recall listing	Varies by national notice
Consumer email alerts	Yes	Yes	Yes	Yes	Varies by system or country
RSS	FDA offers RSS services	Available through FSA services	Explicitly offered	Not prominent in reviewed pages	Varies
Clear state, province or territory field	Often in advisory text or maps	Regional detail varies	Often shown	Usually prominent	Country-level distribution prominent
UPC or barcode search	Data sometimes contains UPCs, but no unified household scanner	Product identifiers vary	Product tables frequently include identifiers	Product and date details shown	Identifier quality varies by notification
Restaurant versus retail distinction	Present inconsistently across advisories	Varies	Formal communication categories help	Distribution type included in recall process	Distribution actions shown
Integrated “is it still in my home?” assessment	No	No	No	No	No

The comparison suggests that America’s principal weakness is not the absence of data. It is the absence of integration.

The strongest new service would combine:

the product specificity of FDA advisories;
CDC’s case and symptom information;
openFDA’s machine-readable recall history;
Canada’s unified alert navigation;
the UK’s documented API approach;
Australia’s state and date presentation;
and RASFF’s cross-border distribution tracing.

What a genuinely useful American food-recall checker should contain

1. UPC, lot and plant-code lookup

A consumer should be able to type or scan:

a UPC;
a production or plant code;
a lot number;
a Julian date;
a sell-by date;
or a brand and product name.

The checker should return exact matches and near matches, while clearly distinguishing them.

For the egg recall, a UPC match alone would be insufficient. The consumer would also need the plant code and Julian date range.

For the blueberry recall, the system would need to understand that the action expanded from one lot to all lots of four products.

2. A time-sensitive household relevance score

The tool should distinguish between:

fresh products, likely to leave a home quickly;
refrigerated products, relevant until or shortly beyond their date marking;
frozen products, potentially relevant for years;
shelf-stable products, dependent on their best-before dates;
and ingredients, which may have been incorporated into other foods.

The score must be transparent. It should not imply medical risk that official agencies have not calculated.

A simple model could use:

days remaining to the latest date marking;
normal storage method;
whether the recall covers all lots;
whether consumers commonly decant or repackage the product;
retail versus food-service distribution;
and whether FDA still describes the investigation as ongoing.

3. Separate product distribution from illness geography

The egg investigation demonstrates why these should never be merged.

Confirmed product distribution covered a limited group of states, while illnesses appeared in 17 states. A map should present:

confirmed distribution;
possible further distribution;
states with reported illnesses;
and food-service locations where known.

4. A visible “source not yet identified” section

Unidentified-food investigations should not be hidden beneath conventional recalls.

They should be displayed separately with:

pathogen;
case count;
date first posted;
investigatory steps;
latest change;
whether the outbreak is growing;
and an explanation that no specific avoidance advice is available.

5. Version history

Recall information changes.

The blueberry action expanded. Cyclospora case counts rose. A preliminary lettuce test was reclassified as a false positive. CDC added a notice about two deaths without updating its headline total.

Every checker entry should show:

the information available at first publication;
subsequent expansions or corrections;
the timestamp of each source;
and the precise reason a household result changed.

6. Retail, restaurant and institutional exposure

A person cannot scan a package they never saw.

The checker should say whether the implicated food was supplied to:

supermarkets;
independent retailers;
restaurants;
schools;
hospitals;
care homes;
caterers;
or other institutions.

Restaurant exposure requires location and date search, not barcode search.

7. Symptoms and incubation periods without diagnosis

The tool should link pathogen-specific information to each outbreak.

For the current examples:

Salmonella commonly causes diarrhoea, fever and stomach cramps, with some groups at greater risk of severe illness.
Shiga toxin-producing E. coli commonly causes severe cramps, diarrhoea—often bloody—and vomiting; some patients develop haemolytic uraemic syndrome.
Cyclospora often causes watery diarrhoea, fatigue, appetite loss and weight loss, usually beginning about a week after exposure but potentially later.

The checker should not attempt to diagnose users. It should state when official guidance advises contacting a healthcare professional. (CDC)

The warning should last as long as the product

News coverage is organised around announcement dates.

Food risk is organised around storage.

A bag of frozen fruit dated for 2028 does not cease to matter because an outbreak article falls from a homepage.

An egg carton does not become irrelevant while its sell-by date remains in the future.

A restaurant-linked lettuce outbreak does not end for a patient whose symptoms emerge days after the final retail package expires.

And an unidentified-food investigation does not become reassuring merely because the public cannot yet name the product.

The American system already contains much of the information required to build a better warning service.

FDA publishes outbreak investigations, detailed product tables and weekly enforcement data. CDC supplies epidemiological totals and medical guidance. Companies provide UPCs, photographs and lot codes. State health departments add local detail. USDA maintains a parallel recall system for products outside FDA’s principal food jurisdiction.

What is missing is the connective tissue.

A consumer should not need to understand federal jurisdiction, traceback stages or the architecture of the Recall Enterprise System to decide whether to discard a carton of eggs.

The essential public-health question is straightforward:

Could this food still be in my home, and how do I identify it?

At present, the answer exists—but often only after the consumer assembles it from several different government pages.

That is the food-recall checker the United States still needs.

Consumer check: products covered by the three principal investigations

Shell eggs

Check for:

Country Morning, Cal-Maine Sunups, Brookshire’s, Simple Truth or Kroger products;
plant code P-1950 or 0840962;
Julian date 157–184;
sell-by or best-by date from 20 July to 17 August 2026.

Do not eat matching eggs. Follow FDA or CDC disposal, return and cleaning instructions.

GreenWise frozen berries

Check for:

GreenWise Organic Whole Blueberries, 10 oz — UPC 41415-06453;
GreenWise Organic Whole Blueberries, 48 oz — UPC 41415-12053;
GreenWise Organic Whole Mixed Berries, 10 oz — UPC 41415-06753;
GreenWise Organic Whole Mixed Berries, 48 oz — UPC 41415-12153.

The expanded recall covers all lots of these products.

The original 10-ounce blueberry recall included lot 60401 and best-by date 9 February 2028.

Taylor Farms iceberg lettuce

The recall covered all iceberg lettuce sourced by Taylor Farms de Mexico from central Mexico, including Marketside products with dates from 18 July to 3 August 2026 and wider food-service products.

Although the listed retail dates have passed, people who recently ate implicated lettuce and develop symptoms should consult current CDC guidance.

Methodology

This article reviewed information available on 4 August 2026 from:

the FDA CORE outbreak investigation table;
individual FDA outbreak advisories;
CDC outbreak notices;
FDA company recall announcements;
openFDA food-enforcement documentation;
the UK Food Standards Agency’s alerts API;
the Canadian Food Inspection Agency and Canada’s Recalls and Safety Alerts service;
Food Standards Australia New Zealand;
and the European Commission’s RASFF system.

“Household persistence” is an editorial classification, not a federal hazard rating. It considers storage method, product date, scope of recall, likely domestic retention and distribution channel.

Hospitalisation rates were calculated by dividing reported hospitalisations by reported cases. These figures describe detected cases in evolving outbreak investigations and do not estimate an individual consumer’s medical risk.

Case counts, recall scope and official findings can change. Each live database entry should therefore retain its source timestamp and update history.

Editorial fact-check note

The following points were independently rechecked against primary sources:

1,589,577 dozen eggs recalled;
egg dates extending through 17 August 2026;
egg case total of 98 and hospitalisation total of 26;
frozen-berry case total of 12 and hospitalisation total of four;
original blueberry best-by date of 9 February 2028;
expansion to all lots of four GreenWise frozen products;
iceberg-lettuce total of at least 1,947 cases and 98 hospitalisations;
invalidation of the preliminary positive Cyclospora lettuce sample;
continued FDA reliance on epidemiological and traceback evidence;
six recent FDA-table entries involving unidentified foods;
and the unresolved discrepancy between CDC’s notice of two Michigan deaths and the zero-death headline total.

No claim has been made that an unidentified current outbreak is connected to a particular brand, retailer, product category or geographic distribution unless an official source has publicly established that connection.

Tags: FDA

Trending Tags