September 14, 2018

The Anatomy of the Outbreak Investigation

As a cog in the machine of the Leafy Greens Task Force (a coalition of produce companies, trade associations and others committed to understanding the recent outbreaks associated with romaine and developing steps to prevent further occurrence), I wanted to share my perspective on some of the difficult issues that have been focal points of discussion for that group with regard to the recent E.coli outbreak.

The first thing on my list is how the Centers for Disease Control (CDC), the Food and Drug Administration (FDA) and industry work together on outbreaks, and the resolutions that stem from that work. Make no mistake, CDC and FDA staff are smart, committed and are accessible to industry during these events. It is their job to discover an outbreak, identify the commodity, identify the source and stop the distribution of contaminated products. Their bottom line is to protect public health and their immense challenge is the need to be both “fast and right.”

Typically, though not always, these events start with CDC who uses PulseNet to stitch together reports from discrete state departments of health to confirm that ill people in varying geographies are infected by a common source. Next they interview victims and healthy consumers to determine the likely food vehicle (source). Occasionally, they have product available to genetically confirm the source.

To “be right” they strive to have multiple lines of confirmation, including the epidemiologic information, traceback information and microbiological information, but to “be fast” they don’t always have the luxury of confirming all three lines. It is a difficult position for them. CDC and FDA try to keep industry representatives informed as they progress their thinking, and industry representatives regularly make themselves available to answer any questions that will help the agencies progress including offering information about products, seasons, distribution, etc. Industry representatives also challenge the agencies when their hypotheses don’t fit industry patterns. But in reality, there is little that trade organizations can do at this point of an investigation.

Once a food vehicle has been identified, the traceback, led by FDA, begins and it is at this point that trade organizations, who understand when, where and how products are grown, sold, transported, and marketed are much better positioned to help. This, however, is when individual companies begin to be contacted by FDA and FDA will not (nor should not) discuss these lines of investigation at this point because of the reputational and economic implications of naming names. Again the challenge is to be “fast and right” but traceback is a tedious step-by-step process that requires accurate documentation of each node within a supply chain to ensure that the source is identified correctly.

Ideally, FDA identifies a clear single source and the collateral impacts to other producers are minimal. In the most recent outbreak and traceback investigation, many cast aspersions on grower and handler traceability programs but delays were mostly within the first step back from the point of sale. While trade organizations can speak generally to supply lines, they cannot connect a point of sale to the distribution center that delivered the product to the store. This is an outbreak in which there are many lines of traceback and no convergence to identify a single source.

So because of the seasonality of supplies and multiple lines of investigation leading to a general area, FDA identifies a single region as the source and begins environmental assessments in that region in its attempts to more narrowly define the source. In addition, industry hopes the environmental investigation will help identify potential reservoirs for the pathogen, sources of contamination and mechanisms for introducing contamination into a field. Insight into these factors helps to understand vulnerabilities, what might have gone wrong and how we may be able to prevent future events. While FDA has indicated it is incumbent on industry to “resolve” the issues, and industry embraces this responsibility, this cannot be accomplished without full and transparent access to all the results of FDA’s environmental assessment. In this particular case, it has been an industry imperative to act in advance of the next growing season but we are hindered by FDA’s lack of detail on the results of the environmental assessment—which as of this writing in August may still be underway.

What We Know

So we are left with what we know and our own hypotheses. This includes knowledge that matching samples to outbreak strains have been identified in canal water within the area of investigation, that there were anomalous weather events near the region leading up to the point when product may have been harvested and that a large concentrated animal feeding operation (CAFO) is also present in the area of investigation. The Leafy Greens Task Force has taken on the task of using this sparse information to develop recommended changes to industry’s best food safety practices.

CAFOs are a recognized microbiological hazard when closely situated in areas of fresh produce production and may present varying levels of risk depending on many intrinsic factors of confined domesticated animal production as well as extrinsic environmental, topographical, hydrogeological, and crop management factors. The Leafy Greens Task Force organized a “working group” comprised of industry, academia, and other interested parties to discuss these factors, the risk they may present, common mitigation strategies to minimize risk, and to make any necessary recommendations to industry on how to further minimize the potential for microbial contamination from CAFOs. Additionally, the group is tasked with identifying priority knowledge gaps as a prelude to recommending key solutions-based research opportunities.

While this effort was born out of a rapid industry response to the severe and recent outbreak of E. coli O157:H7 linked to romaine from the desert growing region where a large CAFO is intimately proximate with fresh produce operations, it is important to emphasize that the cause of the outbreak is unknown and that this CAFO has not been specifically implicated. One important observation from this CAFO working group effort is that while the CAFO may be a reservoir for E. coli 0157:H7, there are many factors that influence the potential discharge of pathogenic bacteria from a CAFO into the surrounding area and many more factors that may determine or influence potential contamination in a field(s). These include such things as geography, weather and, perhaps most significantly, the management and best practices within. The natural tendency of our industry is to gravitate toward “buffer zones”—defined as areas set aside to account for offsite movement and deposition of, in this case pathogens, where there will be no deleterious effects. In fact, many in the supply chain are already requiring buffer zones of varying distance (most are around 1-2 miles) that drastically restrict where leafy greens (not just romaine) may be grown.

Where is the science to support the establishment of these distances? To date, the research on the offsite movement of pathogens and the subsequent deposition and survival in leafy greens fields has been sparse. The CAFO working group recommended a distance of 1,200 feet from the edge of a concentrated animal feeding operation larger than 1,000 head as a proximate safe distance and further encouraged growers to perform an aggressive hazard and risk assessment in each area where a CAFO is near fresh produce fields. Growers need to make site specific determinations based on numerous factors (some named above) as to how far away from an animal operation fields and surface water sources should be.

Set distances impact property rights and economic livelihoods—but no set distance can guarantee safety. Instead, careful analysis of hazard, risk and implementation of controls should drive decisions. If approached correctly at the outset, growers can minimize the potential for contamination and prevent outbreaks, which in turn minimizes the collateral impacts associated with a CDC and FDA investigation and traceback.