Food companies that are voluntarily recalling a product usually use the language suggested by regulatory agencies that is then distributed electronically via the federal government websites. Recalls not connected to outbreaks almost always contain a line saying there are no reported illnesses associated with the recalled product.
William E. Keene, Oregon’s senior state epidemiologist, thinks it’s time for reporters and editors to “not let themselves be duped into amplifying press release boilerplate.”
Both the U.S. Food and Drug Administration (FDA) and USDA’s Food Safety and Inspection Service (FSIS) have templates for companies to use when writing a recall announcement. They are:
FDA: ”No illnesses have been reported to date.”
FSIS: “There have been no reports of illnesses associated with consumption of these products.”
“Until they have the isolate(s) serotyped, have PFGE’d (conducted a pulsed-field gel electrophoresis analysis), and compared to PulseNet or analogous data,” says Keene, “they are just blowing smoke.” He says that without subtyping data, the media should not fall for it. The only accurate report is that there is no way of knowing (at this time) if there are any associated illnesses.
Dr. David Acheson, who previously held top posts at both FSIS and FDA before becoming a partner at Leavitt Partners, says Keene “is correct, but nobody other than a microbiologist would know what the heck that means and the basic statement from the regulators is still correct.”
Acheson says the point of recall announcements is to answer the public’s “what should I do” questions, and he prefers keeping it “simple and specific.” Acheson says he might add the word “yet” to statements about no illnesses being reported.
Phyllis Entis, the food safety microbiologist and author, who publishes the highly respected eFoodAlert, covering food recalls worldwide, goes a step further than Keene.
“I think that an absence of a ‘no illnesses’ statement would be preferable to the current boilerplate,” she says. “I agree that we should be asking the question, and I have queried CDC in some cases. At times, I even get an answer!”
And Entis notes that boilerplate is used worldwide. The XL Foods recall, the largest beef recall in Canada’s history, originally came out with this line from the template: “There have been no reported illnesses associated with the consumption of these products.”
A short time later the Public Health Agency of Canada (PHAC) was confirming 18 cases of E coli O157:H7 from XL’s contaminated beef.
Keene, who has solved many a multistate outbreak, admits he is motivated by a desire to “light a fire” under food companies to get them to turn over isolates to public health labs so they can be subtyped. Only then can the food isolates be matched to any human cases that might exist.
“As you can imagine,” Keene says, “this is not a prospect that excites many food processors, as it literally asking for trouble. I think making it clear that we don’t know if there might be illnesses will prompt people to ask more of the right questions.”
Keene notes that ill people who are matched to contaminated food are often never asked (or told) if they ate it. He says the system is complicated, with the left hand often not knowing what the right hand is doing.
Carol L. Tucker-Foreman, distinguished food policy fellow at the Consumer Federation of America and USDA’s former Assistant Secretary for Food and Consumer Services, says where there is a recall, there is risk.
“The current FSIS language suggests that there are no reported illnesses as yet and therefore no risk to public health,” she says. “But if that were the case there would be no recall. The recall means the company and FSIS know some meat from the recalled lot is available to the public. That means there is some continuing risk and the agency should say so.”
With her experience in and out of the federal government, Tucker-Foreman says both FDA and FSIS “are eager not to alarm the public unnecessarily or create to much of a problem for the company where there’s no reported illnesses.”
She says food companies “particularly hate the fact that a press release and recall occur based only on plant testing, However, there is no recall unless FSIS believes some contaminated product left the plant and has not yet been recovered.”
For his part, Keene is the first to acknowledge that most recalls are not associated with illnesses. However, he says, that is only partially true because nobody is asking. If isolates were turned over to public health labs and searches for matches conducted, then we’d know more about whether or not there were illnesses.
Keene says even then if the test produced a common pattern and there was no spiking of the pattern, its unlikely CDC would bother getting back to “state epi people” and the matches might not be contemporaneous enough to make it worthwhile to health officials to get back in touch with the ill people.
Keene was last in the national news in September 2011, when Del Monte Fresh Produce dropped a lawsuit against the epidemiologist that some said sought to put the science of epidemiology on trial. Del Monte was upset because some of its Central American grown cantaloupe were subjected to an import alert.© Food Safety News