We’ve frequently written on traceability, including these three recent pieces:
After reviewing the articles, he wrote a piece titled, A Traceability Story: Request for Comments, which asked five important questions. Since he asked for comments, we thought we would give him some. We list his question first and our comment beneath each question:
Is the implication of Jim’s article right, that spending money on good tracing systems may be futile because consumers and regulators will never trust the system?
We wouldn’t use the word futile. In the first case, there seems to be some evidence that excellent traceability systems provide a kind of information infrastructure that can pay for itself. We dealt with some of these possibilities here. So traceability may be justified for reasons beyond food safety.
In addition, excellent traceability can certainly help limit the scope of a recall — if a problem can be identified. In other words, if product was traced back to a plant and the FDA quickly noted that there was a rat infestation and that this was causing the problem, great traceability would probably stop the recalls right at that point.
Equally during certain times of the year, excellent traceability could help by limiting the region that is possibly at fault. So if the pathogen is found in a bag made from Yuma-sourced product but by the time it is discovered the trade has shifted to Salinas, a traceability system might save the industry as the FDA, which might have otherwise done an industry-wide ban, instead sees the futility of such a declaration.
So good traceability is not futile. The outbreak and the behavior of the trade and the FDA does, however, confirm the validity of some concerns expressed about the issue of traceability.
One concern is whether the focus on traceability is simply misguided. Certainly the most prescient voice in this discussion, a voice whose point has been confirmed by every subsequent outbreak, was that of Jim Gorny. He wrote a piece we ran under the title, Tomato/Salmonella Situation Cries For Improved Epidemiology. Though Jim was at the time a respected academic, holding the position of Executive Director, Postharvest Technology Research & Information Center at University of California, Davis, he had previously worked as Senior Vice President, Food Safety and Technology at United Fresh and prior to that he was a founding principal at Davis Fresh Technology. Now, of course, he holds the important position of Senior Advisor, Office of Food Safety at the FDA’s Center for Food Safety and Applied Nutrition (CFSAN).
His point that the problem is typically not traceability but epidemiology is certainly true. Indeed one suspects that all the attention being paid to traceability can be explained by two simple facts: First, it is something tangible that government can mandate. After all, the government can’t pass a law requiring epidemiologists get it right each time. Second, traceability is a private sector function. So government can pass a law or make a regulation, and it functions as an unfunded mandate on business. Epidemiology, in contrast, is typically a government function. So if the decision was made to improve epidemiology, Congress would have to appropriate money — money the government doesn’t have.
The second concern is that the nature of most produce problems is such that case-specific traceability doesn’t typically provide the government with a limiting factor — if we assume the policy is a zero tolerance for pathogens.
Traceability to a specific factory that makes soup limits the scope of recalls because the critical control point is typically the cooking in the factory. So it is assumed that chicken may have salmonella or meatballs a pathogen — but the cooking should prevent a problem. That it didn’t tells regulators that a critical control point in that particular plant was not working. The regulators would have no reason to go back down the supply chain or to ban all soup from everywhere.
Produce is just different. Although we have run pieces such as this one in which some have claimed that the expectation is that the farmer will deliver “dirty” produce and the fresh-cut plant will be responsible for making it clean, fresh-cut facilities don’t actually have the kind of tools to reliably do that 100% of the time. Even chlorine, for example, is not approved for the purpose of cleaning food; it is only approved to clean the water!
So when a pathogen shows up, the bottom line is unless there is some obvious unsanitary problem in a plant, the FDA is going to look to the fields. And once it does, there is no logical limitation on a recall or a recommendation not to consume that comes from knowing the case. Think of it this way: Imagine our traceability was so perfect that we had a video camera on each individual head of Romaine. Now when it is implicated, we play back the video from the moment it was planted and we eventually see a wild boar defecating on the plant. Now the FDA goes down, captures Wilbur, kills him and confirms he was the source of the pathogen. That is very specific traceability.
Yet, what should the FDA do? Assume that this one boar — and only this one boar — has a pathogen? Assume the boar stayed on that one farm? No, if the rule is that it is unacceptable to have pathogens on the produce — and certainly no FDA will accept the possibility of someone being sickened when the FDA knows this causation, the FDA — especially once it has mandatory recall authority — is likely to recall everything that could have come in contact with this wild boar or any other wild animal in the region.
The third concern is that buyers have two issues: A) They have the limitations of their own people and systems. They simply can’t be sure they will completely dispose of a particular lot number, so they will always err on the side of caution and throw out more than the minimum recalled. B) They have to communicate with consumers and consumer communication requires simplicity. If it is not a very powerful brand, the easiest communication is to say “We’ve eliminated anything to do with that company from our facility and are no longer buying from them.”
This is especially true in foodservice, where business depends on the consumer trusting the operator to put only healthful product in the food served.
So, yes, buyers probably will never accept anyone telling them that the contamination is limited to these 78 cases and not one more. So in this sense, case-specific traceability won’t offer the returns in reduced scope of recall that many had hoped.
What kind of public education might work to improve the acceptance of traceability?
This is an interesting question and one we haven’t seen addressed. Very possibly a great TV commercial showing the ease with which the industry knows how to trace things back would build public confidence.
But we don’t think the problem is ignorance. The problem is fundamental. If we have a problem with melons from Honduras and the FDA believes the problem is river water, then traceability to a specific melon in the field would not serve to limit the scope of concern to that one melon, or that one row, or that field.
If you go back to our interview with Michael McCartney, which we ran here, you see the key point: Traceability is only useful if it traces back to relevant distinctions. The classic example is artificially small lots off a production line. If you are just running the line, simply declaring a new “lot” every 100 boxes doesn’t limit the scope of concern. You need to do something between lots to make the distinction meaningful — say sanitize the machinery or switch raw product suppliers, etc.
As long as the problem is reality — not ignorance — education is of limited use.
As a legal matter, it’s unlikely that the buyer of the non-recalled products has any recourse against the seller; in the real world, however, is the seller likely to make good in order to assure future sales?
We are not going to parse cases with Richard Goldfarb, but we would add a caveat. Yes, legally speaking, a buyer of non-recalled product doesn’t typically have claim against a vendor — but that is in the absence of a contractual agreement saying otherwise. Yet to get a vendor number at most big buyers, one has to sign an agreement that often says the buyer can clip the bill. Put in layman’s terms, the buyers may pay less than the invoice whenever they want to for whatever reason they may have.
Of course, in most cases this isn’t necessary. The retailers will take the money and the shippers won’t sue to get it back because the lifetime of Purchase Orders is much more valuable than the one-time loss.
Ken [Odza] recently wrote about a similar insurance issue; is there any kind of insurance for something taken off the shelves because of an abundance of caution when the supplier says only to recall specific items?
The “Ken” that Richard Goldfarb is writing about is Ken Odza, a partner at Stoel Rives, who is prominent on defense work in the food safety arena. The article he references is an interesting piece on a case we would hold to have been wrongfully decided. The piece is called, Defective Cans of Fruit Deemed ‘Usable’ by Insurer…but for What?
The essence of the piece was that Del Monte Foods bought a bunch of cans with pull-tabs. The pull-tabs didn’t work, but the court ruled that since the fruit inside was still good the technical requirement of the insurance for “loss of use” wasn’t realized as the fruit was still “usable.”
The court was apparently oblivious to the fact that the fruit was not, in a commercial sense, useable as the instant you open the cans, you would have to reprocess the fruit to maintain food safety.
The lesson was two-fold: A) Get yourself a good lawyer in insurance matters, and B) Sometimes it is hard to know what insurance will actually cover.
We’ve never heard of “abundance of caution” insurance, but we will appeal to the Pundit readers to fill us in if they know who sells such a product.
In his updates, Jim suggests that the real issue is that perhaps we are providing more traceability than the market demands and others suggest that the issue is that upon discovery of an outbreak, the FDA doesn’t either adequately communicate the perceived cause of the outbreak or ever issue an “all clear” after it is over. Is either step either (a) practical when things are moving in real time, or (b) really the FDA’s responsibility or even power under current laws and regulations?
The whole point of the Produce Traceability Initiative is to move the industry to provide more traceability than the market was doing. One of the strongest motivations for PTI was to supersede government action on the matter. At the same time PTI, by providing the industry with a common “language,” may create unknown benefits above and beyond PTI.
It would be very good for FDA to communicate the “cause” of an outbreak but FDA rarely knows the cause in real time — as the Jim Gorny piece we mentioned above indicates.
“All Clears” would also be desirable, but outbreaks are really unpredictable events. And the FDA would look silly if it went around declaring “all clear” and there was another outbreak the next day.
What would be useful is for the FDA to place in perspective its legitimate concerns. During the Salmonella Santpaul Tomato cumm chili pepper crisis and the spinach crisis, we often had some fun figuring out how great the risk really was based on the number of servings of these products produced.
It was typically so small that it seemed certain the FDA would be embarrassed to actually stand out there and say, “There is a risk to the fifth decimal place that you might get ill by eating this product, so we are banning it.”
It does seem to us that if the FDA is going to act by, say, recommending that consumers not consume and retailers not sell a particular commodity, it does have a responsibility to explain the standard by which it imposes such a draconian plan and, by the same measure, the standard by which it lifts such a ban.
In many cases, it appears that there is a subjective highly risk-averse culture at FDA and there is simply no indication that the citizenry is as risk-averse as the FDA is.
We thank Richard Goldfarb and Stoel Reeves for helping us think through such important issues.
Richard ended his piece with a little joke and we will steal it:
We say it is a joke because, obviously, if you come across a virulent pathogen that can cause serious illness or death — you throw it out and sanitize anything that came near it.
The FDA doesn’t want consumers to wash pre-washed salads — although Costco requires vendors to put a “wash before using” sign on its spinach. The FDA takes its position because the likelihood of cross-contaminating your kitchen is far greater than the likelihood of washing it off.