A Conversation with Stic Harris, Conrad Choiniere, and Michael Rogers on Foodborne Illness and other Adverse Events 

Each year, the FDA receives approximately 9,600 reports of foodborne illness or other adverse events related to FDA-regulated human food and dietary supplement products submitted by individual consumers, public health professionals, or industry.[1] In addition, the FDA’s Coordinated Outbreak Response & Evaluation (CORE) Network evaluates an average of 75 incidents annually that may potentially constitute a national, multi-state outbreak. 

The FDA takes its responsibility to safeguard the food supply very seriously and is committed to being transparent about how FDA approaches its food safety mission, particularly about how the agency responds when it receives reports of foodborne illness and other adverse events.

That is what three foods program leaders will discuss in this conversation about challenges and opportunities they face every day in working to protect consumers from potential hazards in the complex, global food system.

Stic Harris, DVM, MPH, is the director of FDA’s Coordinated Outbreak Response & Evaluation (CORE) Network, which is on the frontline of foodborne illness outbreak investigations.

Conrad Choiniere, Ph.D., is the director of the Office of Analytics and Outreach at FDA’s Center for Food Safety and Applied Nutrition (CFSAN). His office evaluates consumer complaints and reports of potential hazards associated with FDA-regulated food, dietary supplements and cosmetic products.

Michael Rogers, MS, is the Assistant Commissioner for Human and Animal Food Operations in FDA’s Office of Regulatory Affairs (ORA), the inspection and enforcement arm of the FDA. In his role he works with ORA’s consumer complaint coordinators and investigators in the field, who are the agency’s boots on the ground in this work to keep foods safe.

These three leaders represent different aspects of the agency’s food safety work and, together, they will provide an overview of how FDA is doing its job and what it can, and will, do to continually improve in this space.

Q. How does the FDA typically receive consumer complaints and adverse event reports, and what’s the Agency’s process for responding to them?

Choiniere: CFSAN receives most complaints and adverse event reports through the CFSAN Adverse Event Reporting System (CAERS), which allows consumers, health care professionals, and industry to submit reports to voluntarily. FDA also receives consumer complaints through ORA’s consumer complaint coordinators located throughout the United States in the FDA’s field offices. While we can receive complaints in a few different ways, we all coordinate on the response efforts. Most complaints are made voluntarily, however dietary supplement manufacturers are required to submit serious adverse event reports they receive into the system, as well. CAERS captures any adverse events or complaints related to human foods, dietary supplements, or cosmetics. These can include minor to major medical events, but also complaints about off-taste or color of a product, defective packaging, and other non-medical issues.

I do think it’s important to note that because these reports are generally submitted voluntarily, they may not always include all the information we need about the product, the event, or whether the consumer also had an underlying medical condition. Sometimes we don’t even receive contact information with the report which can make it impossible for us to reach back out if we need more information. When important information is missing from a report, it can be challenging for the FDA to fully evaluate what may have happened, and what additional steps may be needed. That said, we’ll generally use the information that is submitted, to work with the relevant program offices, food safety experts, and medical officers within the agency to evaluate the level of severity of the complaint depending on if it involves an illness, hospitalization, or death or if it involves a vulnerable population. From there we’ll determine what next steps might be needed, which may include inspections, sampling, or other regulatory actions such as recalls or import alerts.

Rogers: As Conrad noted, we also receive consumer complaints through our field office consumer complaint coordinators. When we receive complaints, our consumer complaint coordinators gather detailed information regarding the nature of the complaint, the product involved with the complaint including lot codes, manufacturer information, and if the product caused illness or injury. For any food complaints involving illness, we collect a 72-hour food history from the complainant. This information is entered into a comprehensive, agency-wide, information system that provides automated support for the activities ORA conducts. This system helps us to coordinate across the agency and respond to complaints as they come in.

Q. Is the process different for outbreaks?

Harris: The process for responding to outbreaks can look very different from what Michael and Conrad described for consumer complaint reports. First, we typically learn about potential multi-state foodborne outbreaks - where more than one person is sickened by the same pathogen - from the Centers for Disease Control and Prevention (CDC) and state health departments. For them to learn about the illnesses, someone that is sick needs to have gone to the doctor, been tested, and have their test results reported to state public health authorities. If public health authorities find someone ill with a nationally notifiable disease, they will typically add the case to the CDC’s National Notifiable Diseases Surveillance System (NNDSS); however, the list of diseases that are required to be reported by doctors and laboratories varies by state. The Council of State and Territorial Epidemiologists (CSTE) has developed a list of recommended notifiable diseases that includes pathogens like E. coli, Listeria, and Salmonella – but notably there are other pathogens, such as Cronobacter sakazakii, that are not on the list.

CDC is able to look for genetic relatedness between samples from ill people for some diseases using PulseNet. This is done through Whole Genome Sequencing (WGS) -- DNA analysis that precisely identifies the relationships between different strains of bacteria. When pathogens are highly related, genetically, it indicates that people became sick from the same source. Importantly, not everyone goes to the doctor when they become sick, so there can be gaps in the information available to public health officials working to identify patterns and commonalities among the patients.  If it looks like a food may have been involved, the next step is to find out what food people ate and where they purchased it.

To do this, state and local public health authorities interview people who are sick to gather information about potential exposures including what they ate and where it may have come from. Generally, they will also try to use what they know about illness onset for different pathogens. For example, Listeria can cause illness in consumers up to two months after eating a contaminated product. As a result, if a consumer tests positive for listeriosis, public health officials might ask about the foods they ate in the weeks or months prior. CDC looks at this information to see, nationally, if those who were sick had eaten the same foods or had eaten in the same place, such as a restaurant.  

At this point, if an FDA-regulated product or products are identified as potential sources of the outbreak, the investigation comes to us in CORE. It’s important to note that this can be weeks after the cases of illness are first reported. As CORE gathers data on the supply chain of the product, we look for a common point in that chain where contamination may have occurred. Then we’ll work with Michael’s team in ORA to get boots on the ground in farms or facilities to take samples and review operational practices. Our ultimate goal is to identify the source of contamination and remove contaminated product from the market.

Q. Can you provide a few examples of how consumer complaints helped alert us to potential food safety issues?

Rogers: Probably the most high-profile example that I can point to right now is powdered infant formula. Between the months of September 2021 and February 2022, we received four consumer complaints of infants infected by Cronobacter sakazakii, which may have contributed to death in two patients. Each case reported consuming infant formula purchased throughout the country. We thoroughly reviewed each complaint individually, but as more complaints came in we were able to identify commonalities between them.  Although there was no common product consumed by the four infants, they all reported being fed formula produced at an Abbott Nutrition manufacturing facility in Sturgis, Michigan. At first glance this information didn’t seem to reveal too much considering that a substantial proportion of infant formula consumed in the U.S. was manufactured at this facility. Meaning, we likely would have expected more complaints if there was a more widespread issue. However, we take infant illnesses seriously, so we followed up with an inspection, which identified insanitary conditions at the facility leading to a recall to further protect public health.

Choiniere: I think we have discussed a lot about pathogen illnesses, but there are other dangerous hazards in our food system for which we rely on systems like CAERS and other consumer reports to flag for us. One example, this past summer we learned about edible products containing tetrahydrocannabinol (THC) that were being marketed like commonly consumed foods and popular brands of breakfast cereals, candy, and cookies. The FDA learned that because of the branding, the products were being accidently consumed, often by children, who are more vulnerable to serious effects from THC consumption. Based on adverse events reports, the FDA was able to explore the issue further and learned that America’s Poison Centers had received more than 10,000 exposure reports involving edible THC products, more than 77% of which were in consumers under the age of 19 and 65% of whom reported accidental exposure. In one instance a pediatric patient died. In June 2022, the FDA issued a warning to consumers about these products and coordinated with state public health partners to further monitor complaints and adverse events related to these types of products.

We’ve also experienced many situations where consumers have reported undeclared allergens or otherwise mislabeled products that have often resulted in recalls. For populations with food allergies, these mislabeled products can result in life-threatening adverse events.

Q. What about when we only receive one or two complaints?

Choiniere: We review every complaint we receive. At first, part of what we try to understand is if we’re dealing with a public health issue or if it’s maybe a more isolated issue, such a spoiled food or contamination that occurred in the home. We also try to assess whether the complainant might have a concurrent, underlying condition or even whether they ate a product, simply by chance, at around the same time they experienced an adverse event like a gastrointestinal illness. Sometimes, our review may end up taking us in a different direction and reveal a different food, or non-food, than what the original complaint reported.

The challenge here is that when we only have complaints from one or two consumers, the information we have may not be strong enough to indicate there is an issue with a specific product.

Q. What about when a risk is reported in the media? Does that tend to lead to an increase in complaints and reports to FDA?

Choiniere: While the adverse event reporting system is essential to our work, it is dependent on people taking the initiative to report a complaint or adverse event to the FDA. In many cases it’s likely that people do not associate an illness with a product, and even when the association is made, people are busy with many other demands on their time. So the number of events reported is likely a fraction of the total events that could have been reported. On the other hand, when people hear about a product-related risk in the media, the association can become more obvious and the media may include information about how to make a report to FDA. So yes, we typically receive many more reports when this happens. The uncertainty within such a system requires expert statistics and epidemiology in addition to the case-by-case investigation, so that we can differentiate signals from noise in the reporting system, especially when extensive media coverage is involved.

Q. You mentioned the powdered infant formula investigation. Can you talk more about that and what we learned?

Rogers: As I described earlier, our initial review of the consumer complaints pointed us to the Abbott facility in Sturgis, Michigan. During the inspections we collected environmental samples from the facility and found five different strains of Cronobacter; however, none of those samples genetically matched the available isolates from two case patients for whom WGS results were available or any other clinical isolates in the National Center for Biotechnology Informatics (NCBI) database. And the two case patients’ isolates didn’t match each other. While we aren’t able to say whether the infants became sick from the firm’s product or contamination by other means, these consumer complaints prompted a for-cause inspection that identified insanitary issues at the facility. These conditions led the company to voluntarily recall products made at this facility and close the facility to initiate corrective actions. In the months after, we worked extensively with the firm to ensure they were making acceptable corrective actions that could help them reopen and resume producing formula safely. We also worked with Conrad’s and Stic’s teams throughout our response efforts. Once we were aware of infant illnesses, their teams – and CDC – were on alert, monitoring for additional illnesses and complaints.

Q. Why was it difficult to link the infant illnesses to product?

Harris: There are a few points about this particular pathogen that added complexity to this investigation and the challenges we faced with linking the case patients to particular products or environmental samples. First, Cronobacter is not a nationally notifiable disease. The state of Minnesota is currently the only state that has decided Cronobacter is specifically reportable to the state. This means that even if we are dealing with a more widespread outbreak situation, we typically aren’t learning about the illnesses from public health partners but are instead reliant on consumer complaints. As Michael mentioned, in this case the four complaints we received involving Abbott infant formula came in over the course of roughly four months – not all at once – so that means it took some time for us to identify a possible pattern and better understand the situation we were facing. In addition, cases may also be reported to public health authorities as other illnesses, such a meningitis (which infants can develop if they get sick from Cronobacter), which make it more difficult than some other pathogens to really understand what may be happening.

Because Cronobacter is not a reportable disease in any states other than Minnesota and because severe infections are rare, WGS is rarely performed on cases, so we have less genomic data to work with to help us link cases to each other, to a product, or to a food production environment. For example, the NCBI database only has 1,086 isolates of Cronobacter compared to more than 480,000 isolates of Salmonella enterica, a nationally notifiable disease. So when we have illnesses of salmonellosis, we have compiled a lot of genomic data to work with right from the start.

The other important thing to know about Cronobacter is that it’s a natural pathogen commonly found in the environment. For most of us it’s harmless, but it can be life threatening to infants two months and younger, those who are born prematurely, and for infants that have weakened immune systems. Unless public health officials are able to genetically match strains found on the product or in the manufacturing environment to the strain that made a patient sick using WGS, it’s hard to say where exactly the contamination may have occurred. For example, contamination can occur in a facility if contaminated ingredients are used or if the product touches contaminated surfaces during production. However, contamination can also happen in the home. We’ve seen situations where Cronobacter infections have been linked to sinks, bottles, and breast-pump items. Contamination can also accidently happen during formula preparation if the lid or scoop is placed on a contaminated surface such as a kitchen counter or sink, and later touches the formula. Cronobacter can also get into formula if the formula is mixed with contaminated water or in a contaminated bottle.

Q. What is the agency doing to continue to improve processes for responding to foodborne illness and other complaints?

Choiniere: One foodborne illness or adverse event related to food is one too many, and so we are constantly reviewing cases, capturing lessons learned, and looking for ways to improve our processes. One thing we are always working on is informing industry and consumers about what they can do to keep food safe, as well as how they can inform the FDA of issues so that we can take prompt action. Another thing my team is working on is improving our collection of information on illnesses and product issues from consumers. Right now, we are developing a new online reporting form that the public will be able to use to submit information about their illness or other product complaints to CAERS. This new digital data collection approach will help us to streamline our processes while giving the public an easier way to submit complaints online. It will also allow for product photos to be submitted, which can provide us with a lot of data such as the specific name of the product, lot number, UPC code, and more. Consumers and public health officials can also submit reports online through FDA MedWatch or contact ORA’s consumer complaint coordinators. We actively work together to respond to complaints and are continuously working on improvements in how we communicate internally to ensure we are effectively responding to adverse events.

Harris: In CORE we’re also constantly looking to improve our outbreak response. This past December, we released the Foodborne Outbreak Response Improvement Plan (FORIP), which identifies steps that the FDA is taking to enhance the speed, effectiveness, coordination, and communication of outbreak investigations. Through the FORIP we’re building toward tech-enabled traceability, data-sharing, root-cause analyses, and operational improvements – all aimed at enhancing our response efforts.

Rogers: In ORA, we recently revised our internal consumer complaint procedure to strengthen the escalation process to better define when certain consumer complaints need to be escalated to senior officials, including when our field offices receive a complaint that has the potential to seriously impact public health or if it involves a whistleblower complaint. We’ve also streamlined the initial review process for when complaints come in so that they can be responded to more quickly.

We are also continually updating key trainings for our investigators. For instance, based on lessons learned from recent inspections, we updated our online investigator training and we are also planning to deliver an interactive workshop that will include CFSAN subject matter experts to address advances in industry manufacturing technology. We are also reviewing and sharing lessons learned from recent infant formula inspections among our infant formula investigators. These updates also include a review of infant formula regulations and further define what to look for among the different types of infant formula, including liquid, powder, milk, soy, and others- as each type of formulation has unique characteristics and manufacturing processes that our investigators need to understand. Similarly, we have updated trainings on conducting inspections of facilities that make dairy products, make low-acid canned foods, or use aseptic processing and packaging because it is critical for our investigators to understand how these manufacturing processes impact infant formula production.

We’ve also provided our investigators with updated information about reviewing complaints and adverse event reports prior to and during an inspection to better inform what they are observing.

Q. Anything else you want to add?

Rogers: We take our public health responsibility very seriously. As consumers ourselves, we have seen the effects of foodborne illness and other adverse events and we know that the FDA is on the front lines of the effort to respond to and prevent these issues. Data can make all the difference when trying to figure out what foods made people sick and why, so we’ll continue to explore ways to enhance our data streams so that we can streamline our systems, focus our resources, and improve our overall response efforts.