Agent detection to identify contamination of food products is required in food safety and defense programs. Detection typically involves laboratory methods or technologies, such as biosensors, that are used in close physical contact with food products. While the field of food protection has benefited from the development of novel agent detection methods in recent years, the challenge of determining which food products to test remains. The sheer volume of food produced within and traded across U.S. borders makes agent detection a daunting, time-consuming and expensive task. The decision of when to utilize detection methods depends on the risk of a particular product being contaminated. Contamination may be unintentional or intentional, including economically motivated adulteration (EMA).

The risk of contamination fluctuates over time and is a function of several factors. Risk depends on the biochemical makeup of the product, supply chain characteristics such as complexity and transport distance, and a wide range of natural or manmade events that may disrupt supply and potentially incentivize intentional adulteration. This is particularly true in the case of EMA. Events include but are not limited to natural disasters that destroy or reduce the usual supply of an ingredient, political instability that disrupts usual trade patterns, interruptions of routine food safety inspections, and market fluctuations that impact global prices. While data exists to monitor these risk factors of contamination, optimal use of this information by government and private industry is hindered by several challenges. For example, valuable data often exists across multiple data systems with data across systems appearing in inconsistent formats. In addition, the amount of data that must be reviewed to find a signal within the noise is frequently overwhelming.

To address finding signals within vast quantities of data sources and systems, the Food Protection and Defense Institute (FPDI) developed technology to curate and help make sense of this data. With support from both the FDA and the Department of Homeland Security, FPDI developed FIDES or Focused Integration of Data for Early Signals to perform “horizon scanning” of food system disruptions in support of food protection efforts, including agent detection. FIDES was designed to help users forecast, monitor and identify food system risk factors and adverse food events. The FIDES web application fuses multiple streams of data from disparate sources and displays information in the form of an online dashboard where users browse, search and layer both dynamic and reference data sets related to food system disruption events. Examples of data currently included in FIDES are import refusals, global disasters, animal health alerts, food defense incidents, historical food safety incidents, import data, price alerts and reference data on food production worldwide.

Events in recent years illustrate the value of gathering intelligence and utilizing data related to food system risks to inform decisions regarding product targeting. Tsunamis, crop failures and disease outbreaks in humans and animals around the globe have threatened supply of products such as shrimp, spices, cocoa and eggs. When supply is disrupted, companies are often forced to quickly identify new and sometimes previously unvetted suppliers, including spot market purchasing. Likewise, supply disruptions often lead to price increases. As prices increase in the absence of adequate supply, concerns about EMA also increase. In both of these instances, the risk of product contamination—both unintentional and intentional—may rise and an increase in product screening or a change in agent detection methods may be appropriate.

For example, the 2014–2016 Ebola outbreak had a significant impact on West Africa, the primary production region for the world’s cocoa supply. Disruptions from the outbreak, including border closures and other trade interference, led to uncertainty about supply availability and prices. This raised concern for EMA, particularly given that many cocoa products are sold as powders, butters and liquors— forms that are more vulnerable to EMA than raw ingredients. As a test case, FPDI reviewed FIDES data streams during the peak of the outbreak. Real-time data on the outbreak was layered with data on global cocoa production and import patterns. Import refusal data from multiple global systems was assessed to identify any concerning patterns. Historical food defense and food safety incidents were also reviewed to determine which cocoa products had been previously contaminated. A similar approach could be used by the food and agriculture sector to guide decisions about targeted inspections—which product(s) and region(s) to monitor, which method(s) to use and which contaminant(s) to test. FIDES could support targeted screening and enhanced awareness of product risk profile that would allow the food industry to assure continued supply of authentic and quality products.