Among the list of food recalls in the Food and Drug Administration’s weekly enforcement report from February 8, the agency published information on a Class I, farm-initiated recall of 228,360 pounds of curly leaf spinach due to samples testing positive for E. coli O157:H7.
While notable for its size alone, the 114-ton recall came under scrutiny from some food safety experts for an entirely different reason: More than two months passed between the time the recall occurred on December 31 and the time it was first publicized in the February 8 report.
On the surface, it seemed that a Class I recall — those involving a product that could cause serious injury or death — would have been made public as soon as it happened. Instead, this one appeared downplayed and — to some — downright stealthy.
In the weeks that have followed, the spinach incident has sparked a discussion about the circumstances under which recalls should or should not be made public. It’s a decision that pits the benefits of public knowledge against the perils of public overreaction.
Disclosure or Discretion?
The recall in question, voluntarily issued by Texas-based Tiro Tres Farms, was never linked to any E. coli infections. Tiro Tres does not package spinach for consumers, instead distributing it in bulk to processors who repackage it as their own, and so only the processors were notified of the recall on December 31.
Microbiologist Phyllis Entis first brought the situation to attention on her website, eFoodAlert, the day after the FDA published its information. She criticized the FDA, saying she could not understand the rationale behind not notifying the public of a Class I recall. The next day, Food Safety News ran a story of its own, “Big Spinach Recall with No Public Notice.”
It’s unclear how many processors received the spinach, or how much of it made it to the supermarket and restaurant kitchens. What is known, however, is that it had passed its “best-by” date before Tiro Tres’ sample tests came back positive and processors were notified. Tiro Tres had distributed this batch to processors in Colorado, Kentucky, Massachusetts, Pennsylvania, Ontario and Quebec.
One processor in Pennsylvania already knew there was a problem. Avon Heights Mushroom had issued an earlier recall to customers based on its own testing. That recall was announced by the FDA on December 23. That spinach had been sold as fresh packaged spinach with a “best-by” date of December 16 under the brand names Krisp Pak, Better Brands” and Avon Heights Select.
Given that Avon Heights’ recall occurred a full eight days before Tiro Tres announced a voluntary recall to other processors, it’s possible that Avon’s tests prompted Tiro Tres to test.
In an interview with Fresh Cut Magazine, the owner of Tiro Tres Farms said that Avon’s contaminated spinach was linked to a 27,600-pound batch, while the remaining spinach appeared to have been recalled out of caution. After inspecting Avon Heights’ facility and spending three days at Tiro Tres, FDA officials could not find a source for the contamination.
No Public Benefit?
When asked why the FDA didn’t publicize the Tiro Tres recall, the agency told Food Safety News that it would not have benefitted the public, considering that the spinach was well past its expiration date and was only sent to processors. It would have been up to the processors such as Avon Heights, it seems, to issue a recall to consumers if they thought it was necessary.
According to the FDA’s Regulatory Procedures Manual, “It is FDA’s policy that press releases are issued for Class I recalls unless specific circumstances indicate that a press release would not be beneficial to the public.”
Entis didn’t buy it.
“[The U.S. Department of Agriculture], to its credit, notifies the public of any Class I recall [in their jurisdiction], even when the food just goes to food service or institutions. They don’t use that excuse,” she told Food Safety News. “I don’t understand why the FDA does not do the same.”
Entis added that with most contemporary pathogens tests returning results in 24 to 48 hours, she does not understand why a farm would be running tests that would not produce results until after their product had shipped out to processors, let alone expired.
Produce industry expert Jim Prevor disagreed with Entis, saying the FDA’s move was justified. In a response to the situation on his website, Perishable Pundit, he argued against announcing recalls of products that don’t go directly to consumers, saying nondisclosure was the responsible path. Based on the information available at the time, Prevor said that one could assume that the spinach did not reach any customers.
“When should consumers be notified? Generally speaking, it comes down to one thing and one thing only: Is the product in the hands of a consumer? If it is, then they should issue a public recall,” he told Food Safety News.
Prevor noted that recalls affect sales for the entire industry of the food in question, not just the one producer with a contaminated shipment. When the FDA warned the public specifically not to eat Rocky Ford-brand cantaloupes from Colorado in the 2011 cantaloupe Listeria outbreak, cantaloupe farmers’ profits suffered across the board. Announcing a big spinach recall, however specific, can hurt sales for all spinach growers.
Even then, for any spinach that might have made it to the store, the fact that the contaminated batch was already well-past its expiration date made slim prospects for catching any leftover product.
“Sometimes things are learned very late in the process, and as a result, the product is so old that it couldn’t possibly still be out there to be eaten,” Prevor said. “In those situations, the FDA will decide there’s nothing to be gained by notifying the public. It’s not that it’s not announced — all these things are announced to processors — they’re just not promoted to consumers.”
Testing to Check Practices, Not to Protect Consumers
But why test for pathogens if the results don’t come back until after the expiration date?
As Prevor explained, the primary purpose of pathogen testing on farms is not to protect consumers, but to make sure that the farm’s practices are generally safe over time. Contamination events can be so localized that random testing will never catch every bug, he said, and anything more than random testing becomes too costly for most operations.
“Let’s say we have a field of spinach with 200 million spinach leaves. If we went through and randomly picked a few to test, we might get negative tests on all, but there might be a thousand leaves in the field that are positive. To test them all, we’d have to charge $200 for a bag of spinach,” he said. “There’s no way to guarantee everything is 100 percent safe. If you think that’s unacceptable, you’re asking that we only sell cooked spinach.”
Entis agreed that absolute guarantees for food safety were impossible and pathogens tests on farms were largely meant to monitor consistency in food safety, but she emphasized that a positive test result is never something to ignore.
“Speaking as a microbiologist, finding a pathogen via random testing is so difficult that when you do find one, you know you’ve got a problem,” Entis said. “It’s not a theoretical p
roblem. It doesn’t matter if someone got sick or not. If the problem is not corrected, it’s only going to get worse. We’ve seen this happen over and over again.”
While she also agreed that not all recalls needed publicizing, Entis argued in favor of greater public disclosure for the sake of better scrutiny. If companies have repeated contaminations and recalls, for example, people should have the right to know they could be dealing with an unreliable product.
“I think consumers have a right to know if a company has processed and shipped a product that turned out to be contaminated with a deadly pathogen,” she said. “Try to protect the consumer so they can choose if they want to support that product in the future. How else are we supposed to become informed consumers who make informed decisions?”
Following the attention the big spinach recall received, Entis decided to look back through the last four months of FDA enforcement reports for any other unannounced Class I recalls. She found six such recalls described in enforcement reports since November 2011.
That revelation did not surprise Prevor, who said the FDA takes a number of precautions into account before announcing recalls.
“There are a lot of different values here that have to be balanced,” Prevor said. “One value — without a doubt — is safe food. But we also don’t want broke farmers and people in a panic when they’re not at risk, either.”
Test and Hold to be Safe
Some farms, however, have taken it upon themselves to test and hold produce before shipping anything out.
Since transforming its food safety program in 2006, California-based Earthbound Farms has become an industry leader in foodborne illness prevention. Earthbound was involved in the 2006 E. coli O157:H7 outbreak in spinach that sickened over 200 people across 26 states and Canada and killed five.
Following the 2006 outbreak, Earthbound decided to adopt a preventative strategy for spinach and other higher-risk produce. Instead of performing random testing, they now test a sample from each pallet they ship, holding the product for 12 to 16 hours until they receive a result.
According to Will Daniels, Earthbound’s senior vice president of food safety, the farm’s detection system will catch 99.99 percent of E. coli, Salmonella or Shigella passing through its various testing hurdles. Simply put, they have refused to be involved with any more illnesses.
Daniels said paying for their proprietary testing system, holding space, upgraded equipment and a scientific advisory panel ends up costing roughly three additional cents per bag of spinach. The bottom line, however, outweighs any cuts to profits:
“After 2006, we knew we needed to do more,” he said. “We’re in the business of healthy food and healthy food doesn’t carry pathogens that make people sick. This is just the right thing to do.”
Like Prevor and Entis, Daniels agreed that random sampling cannot be trusted to prevent foodborne illness outbreaks. But no other farms have yet adopted the test-and-hold method to the extent of Earthbound. Implementing such an extensive system takes time and money, and beyond that, doggedness.
“Contamination is highly sporadic — it’s in one little tiny section of a field when the rest is clean,” Daniels said. “When we started doing this, the criticism was, ‘You’re looking for a needle in a haystack.’ But here we are, five years into it, and we are catching contamination and preventing it from going into our processing stream.”
Earthbound detects contamination on approximately 0.15 percent of its produce — 3,000 pounds out of the 2 million it ships each week. Whenever a contamination is detected, not only is the produce saved from potentially harming customers, but the situation gives Earthbound technicians an opportunity to study how the contamination occurred and remedy it. They have amassed a lot of their field lessons into a database of information on how pathogens behave on produce.
“Those people who criticized our program for its vigor back in 2006 are now asking me for the data,” Daniels said. “From our perspective, we have lots of great evidence that suggests our program is viable. We’re committed to it. We won’t back away from it.”
As for when recalls should be made public, Daniels kept his viewpoint uncharacteristically simple:
“If the consuming public is at risk of illness,” he said, “they should be made aware.”
Thanks to Daniel B. Cohen for contributing his knowledge of the produce industry. Photo of Will Daniels courtesy Earthbound Farms.