Epidemiology versus Traceback
Twenty months after watershed events in the lettuce and leafy greens industry, the current tomato foodborne illness outbreak scenario is ominously reminiscent of the September 2006 event, with epidemiological data (to determine the most likely food and source of that food) lagging far behind reported/confirmed illnesses cases. The produce supply chain has been again stopped in mid-motion due to a foodborne illness outbreak, and there is deep and pervasive uncertainty in the marketplace. Everyone in the tomato industry is now suffering from being labeled with a broad brush as the potentially responsible party, and it affects those responsible for these illnesses as well as the vast majority of completely innocent parties.
Since September 2006, many persons have advocated for expenditure of more resources to enhance produce traceability, as they believe that enhanced traceability is the most effective means of avoiding the complete shutdown of a produce industry sector. Well, it has happened again and traceability was never the cause of these industrywide shut downs, and it is unlikely that enhanced traceability will ever prevent a future industrywide shutdown.
First and foremost, continued foodborne illness outbreaks being associated with consumption of certain produce items is the real issue here. Continued focus on preventing produce contamination in the first place throughout the supply chain is the most effective means of preventing any and all future produce industrywide supply chain shutdowns.
Secondly, painstakingly slow epidemiological investigations carried out by local, state and federal public health officials are the real cause of these industrywide shutdowns. Epidemiological investigations aim at identifying the most likely root cause of a foodborne illnesses outbreak. This occurs by first identifying clusters of illnesses, making sure the disease-causing agent is exactly the same among ill individuals, determining what common food ill individuals ate and where that food came from. Determining in the epidemiological investigation where the tainted food came from (which store, then which distributor, then which grower/shipper/packer) by identifying commonalities among where ill individuals purchased their food is critical to ultimately identifying the most likely root cause of contamination.
Importantly, epidemiologists investigating the foodborne illness outbreak must be EXTREMELY careful in assuring that the link between what ill persons ate and where they purchased that food item is factually correct. If a false assumption is made early in the investigative process (i.e., wrong food product or wrong point of purchase), it leads the epidemiological investigation down the wrong distribution chain, thus wasting time, resources and ultimately delaying the identification of the true cause of the illnesses.
The real issue regarding industrywide shutdowns is not about produce traceability (although good traceability does help), but it is about epidemiological investigations that are slow, laborious, time-consuming and resource-intensive affairs. Because multiple federal, state and local public health authorities must collaborate in these investigations, it increases the complexity of the task. This is not a criticism of public health officials trying to protect public health but merely a statement of facts regarding these investigations.
Public health agencies are currently fragmented and under-resourced to effectively monitor and respond rapidly to developing public health issues. Why did it take upwards of 6 weeks for public health officials to identify this cluster of illnesses… and then only due to the diligence of state public health officials in New Mexico?
Food industry, government and the public should be concerned about this lagging response time and its subsequent effects on public health and commerce. Since September 2006, each and every foodborne illness outbreak investigation should have been viewed as an opportunity to enhance investigation team response times to assure public health and well being. Simply put, more public health resources are needed to adequately protect public health and commerce in a timely manner or we will continue to see industrywide shutdowns in the produce industry.
We followed up with a quick question: “When New Mexico made its release, it went down to announcing three specific supermarkets that were the issue. Obviously the FDA is looking more broadly in other states, etc., but it seems as if the FDA is rejecting the epidemiology that New Mexico did to trace everything back to three supermarkets.
Is that the way it appears to you?”
And Jim provided a swift reply:
Good question, you’ve hit it right on. FDA is simply taking a much broader perspective. They mostly agree with New Mexico public health officials that ill individuals most likely purchased product from the named New Mexico retail establishments, but FDA is looking at it in a broader context, trying to connect the dots via the epidemiology investigations.
FDA most likely now knows exactly whom the three implicated New Mexico retail establishments purchased tomatoes from, and FDA is now working with regulators in Texas to find the most probable retailer(s) from whom product was purchased by ill individuals. They are looking at lists of each Texas retailer’s suppliers to see if there are any commonalities with the New Mexico retailers’ supplier list.
This points out it is a lot more complex then just traceback. You have to have the traceability, but then you have to be able to do the epidemiology very quickly, sorting through lots of data to find matching common suppliers and making sure you’ve chosen the right retailer to do the traceback.
The slowness in FDA’s response may possibly be that the agency can’t find any commonalities between the New Mexico and Texas suppliers, so they can’t identify the potential responsible party or even a growing region. The FDA is also very careful when making any announcements as they truly do understand the implications. Silence by FDA is currently very painful, but it is better than being falsely accused, as is exemplified by what happened to California Strawberry Growers in 1995.
Jim Gorny has noted something very important with regard to fresh produce. Traceability, important as it may be, is not the primary cause of a delay in identifying the specific source in foodborne illness outbreaks.
When Bruce Peterson launched — along with Michael McCartney — an industry traceability initiative, he identified the problem this way:
As Senior VP Perishables at Wal-Mart, a day didn’t go by… ok, Maybe not every day, but some 250 times a year… where we had a product recall because of a wrong label, wrong code, a date-quality issue, it didn’t taste good, to more serious problems of E. coli, salmonella or lysteria, and most of these recalls were with products you’d know and trust.
Why aren’t these consumer products companies going out of business? There’s a really simple answer: What CPG companies like Proctor & Gamble, Sara Lee, and Kraft have that the produce industry doesn’t have is effective recall.
A CPG company says, “We have a problem with this lot of lunch meat,” then separates it and communicates that to the public in two ways. The company issues a statement, and secondarily FDA issues a statement. All these three things happen within hours, sometimes days, but usually within 24 hours.
Now Jim Gorny is pointing out that as important as effective traceback may be, traceback, by its nature, can’t start until effective epidemiology has identified what the product is and where it was purchased.
Perhaps if Campbell’s Soup gets consumers sick, the sick people tend to have cans of it in the garbage or relatively good recall of the brand, so epidemiology gets to the proper source quickly. If it was a tomato, few will remember the brand. The tomato might have been on a sandwich from a restaurant or bulk product from a supermarket. Consumers may not be able to state clearly what variety of tomato they purchased.
The important question that Jim asks for the whole industry is this: Why did it take upwards of 6 weeks for public health officials to identify this cluster of illnesses… and then only due to the diligence of state public health officials in New Mexico?
Jim is undoubtedly correct. If it is going to be weeks and months to identify what product is responsible for an outbreak, we will have two big problems: first, people will unnecessarily get sick. Swift epidemiology can save lives and prevent illness. Second, if we have partial indications of things for weeks and months, the FDA will feel a need to act on limited information and will thus impose excessively broad restrictions on sales.
Of course, the great advantage for the industry of focusing on traceability is it is the part of the puzzle we can actually do something about.
We can all agree with Jim that, as he wrote: “First and foremost, continued foodborne illness outbreaks being associated with consumption of certain produce items is the real issue here. Continued focus on preventing produce contamination in the first place throughout the supply chain is the most effective means of preventing any and all future produce industrywide supply chain shutdowns.”
Yet we must recognize that we don’t have a failsafe way to achieve this yet. That is why we are investing, as an industry, in things such as The Center for Produce Safety. We certainly hope to one day rid ourselves and the world of the scourge of foodborne illness but, in the meantime, we have to manage the problem.
In addition, as Jim states, we may all be able to acknowledge that “Public health agencies are currently fragmented and under-resourced to effectively monitor and respond rapidly to developing public health issues.”
Certainly the industry can lobby to get more resources for agencies that hold our future in their hands. But whereas we can decide to implement industry traceability and make it happen, we can only hope to influence public policy discussion on the relative merit of funding different aspects of the public health authorities.
Yet Jim Gorny’s letter is of utmost importance because we have to understand the problem to begin solving it. Jim is right… this is a teachable moment. What we learn just by looking at the tomato/salmonella situation is that we need more effective, more rapidly deployed, more insightful epidemiology.
This is not an easy problem to solve. Effective epidemiology requires a rare combination of art and science, half Sherlock Holmes and half Louis Pasteur. Many of the best leave government service for more lucrative opportunities in the private sector. State departments of health are wildly irregular in their competency and even within FDA, the competency levels vary wildly.
Can the Center for Produce Safety look to develop better tools for epidemiology in produce? Can the government relations pros at PMA, United, WGA, FFVA, etc., get budget allocations to beef up staffing and pay people in a way likely to increase competency?
Many growers are already suffering as a result of the consumer warning not to eat certain types of tomatoes. If we use this moment as a turning point to focus industry efforts on this critical area of epidemiology, perhaps some of this hardship will not be incurred in vain.