As the familiar saying goes, “Insanity is doing the same thing over and over again and expecting different results.”
This is one of my favorite quotes, particularly when specific outcomes are not what we desire. It may seem self-evident not to repeat mistakes, and yet, we’re all guilty of doing this very thing on more occasions than we’d like to admit. Similar to habitual behavior in our personal lives, this maxim also captures an important challenge in how we think about foodborne illness investigations today.
The food-production landscape is continually changing, including the risks we face and the tools available to identify and manage them. Global supply chains are increasingly complex, environmental conditions are shifting, pathogens and hazards are evolving and detection tools are becoming more advanced. Outbreak investigations must evolve as well so we can prevent, detect and contain foodborne illness more quickly and effectively.
Much of today’s conversation around outbreak investigations understandably focuses on traceability and how quickly can we identify a potential source or point of contamination. More rapid and detailed traceability is essential to outbreak intervention. But we must assess the broader system that encompasses outbreak investigations. The broader system, as defined by the U.S. Food and Drug Administration (FDA) and the Centers for Disease Control and Prevention (CDC), recognizes that foodborne outbreak investigations comprise three interconnected, fundamental streams of evidence: epidemiology, laboratory testing and traceability—components that must not only be robust on their own but also collectively.
The system is not sustainable if we only focus on strengthening one component and neglecting the others. Like a three-legged stool, the system depends on the strength and stability of all three.
Start with epidemiology. Before investigators can trace a product, they first need to determine which product they are looking for. That sounds simple enough, but it often is not. It begins with reconstructing what infected people who report their illness ate over a specified time period, including where they purchased it and what other exposures they may have had. As time passes, memories fade, receipts disappear, packaging is discarded and seemingly simple questions—What brand was it? Where did you buy it? When did you eat it?— become more difficult to answer. Collecting detailed, structured exposure information closer to the time of illness can improve patient recall, strengthen hypothesis generation and ultimately support more effective outbreak investigations. The CDC itself recognizes that investigators sometimes lack enough detail about exposures to even begin traceback.
That should prompt us to ask whether improvements in traceability need to be accompanied by improvements in how—and how quickly—we collect epidemiological information. Can standardized questionnaires be used sooner (e.g., in the doctor/clinic office)? How often are questionnaires evaluated for improvement or expansion (e.g., other possible exposure pathways)? Can technology help capture information before memories fade or disappear (e.g., streamline access to electronic purchase data from retail/food service)? Can local and state entities play a more active role in capturing relevant information earlier (e.g., use of standardized surveillance system or database)? These questions relate to the broader information-gathering system and deserve the same level of attention as traceability.
Laboratory testing brings a different set of challenges. While extremely valuable, laboratory testing has limitations in its application and relevancy. Testing is more than generating a positive or negative result. A recent paper on food safety testing highlights the importance of understanding the capabilities and limitations of sampling and testing systems so results can be appropriately interpreted and used to inform food safety decisions. First, contamination in food and the production environment is typically unevenly distributed. Second, products infected people report consuming may no longer be available when an investigation begins. Because of its relatively short shelf life, this is especially pertinent to fresh produce. In addition, only a small portion of a food, if it’s available, or the production environment can realistically be sampled. Also, as FDA has noted, a negative result does not necessarily mean contamination was absent.
Aside from testing challenges, but relevant to testing programs, research also indicates that a changing environment brings new challenges. For example, pathogen occurrence and persistence are influenced by slow, permanent changes in climate over time and transient changes such as extreme weather events (e.g., flooding, drought, wildfires, high winds). Weather-related climate changes may vary season to season or change gradually over time. Innovation in microbial detection methods and other emerging technologies can offer opportunities for sensitive and faster detection, but new technologies also require substantial research, capital investment and validation, prior to integration into surveillance systems.
So perhaps the testing conversation also needs to evolve—from simply asking, “Are we testing enough?” to asking, “Are we testing effectively and developing the science, technologies and surveillance capabilities necessary to anticipate and detect the emerging hazards we will face tomorrow?”
And then there is traceability. With the publishing of FDA’s FSMA Rule 204 on traceability, the promise of enhanced, more efficient traceability is significant. However, no grower, shipper, processor, distributor, broker, retailer or restaurant can create an effective traceback system independently. Information must remain consistent and connected to food as it moves through the supply chain. This need for continuity marks traceability as a shared supply-chain responsibility and not simply a regulatory obligation assigned to individual businesses or sectors of a supply chain.
Traceability, too, presents its own systems challenges. Companies are at different stages of implementation and have invested in different technological solutions and approaches to comply with FSMA Rule 204. As companies make decisions about which system to use and service providers build various tools, the focus, as a whole, should be on system interoperability, data quality and best practices that allow information to move safely and efficiently across the supply chain. Ultimately, no single supply chain segment can make traceability successful on its own; its value depends on shared responsibility and a coordinated farm-to-fork effort.
The real opportunity before us is to strengthen all three legs of the stool and how they work together.
After every outbreak, rather than focusing on traceability improvements or immediately looking for someone to blame, we should be asking where information was lost, where time was lost, if investigations were driven by the data or by assumptions and how the connections among epidemiology, laboratory testing and traceability can be strengthened.
Finding the source of an implicated food is not the same as solving an outbreak. Traceability may help investigators follow the trail, and recalls may help contain the risk, but neither can stand alone. Effective outbreak investigations require epidemiology, laboratory evidence and traceability to work together—not only to identify the problem, but also to understand it, contain it and prevent similar issues in the future.
Earlier and more consistent exposure-data collection could improve the accuracy and completeness of the information available to investigators, helping them develop stronger hypotheses. At the same time, the testing conversation must evolve and focus on advancing the science, knowledge, technologies and surveillance capabilities needed to maximize its value. Combined with better traceability, these efforts could accelerate source identification and improve outbreak response.
The goal for both industry and regulators should not simply be faster traceback. Traceback speed matters, but the larger goal should be a faster-learning, more efficient and effective food safety system—one capable of detecting potential signals sooner, forming better hypotheses and testing them with the best available science; tracing products rapidly through an interconnected and often complex supply chain; taking proportionate action to protect consumers; and, most importantly, using what is learned to prevent future foodborne illness outbreaks.