Governments around the globe are observing the second annual World Food Safety Day on June 7 to promote collective and individual actions to ensure a safe food supply. To mark the occasion, The Pew Charitable Trusts spoke with three veteran public health officials who have encouraged the use of root cause analysis (RCA) to improve food safety. Their efforts helped inform a Pew guide, published in March, that details how food businesses, regulators, and other organizations can conduct effective RCAs.
Root cause analysis is an essential investigative method for food safety professionals, enabling them to understand why products have been contaminated and help businesses prevent problems from recurring. Some in the food industry, as well as local, state, and federal agencies, conduct RCAs of foodborne illness outbreaks or product contamination events routinely. But overall, the approach remains underused.
Jack Guzewich led the Food and Drug Administration’s outbreak investigation and response office from 1997 until 2011. He previously oversaw foodborne disease surveillance, food service establishment regulation, and staff training at the New York State Department of Health.
Steve Mandernach serves as executive director of the Association of Food and Drug Officials, a nonprofit membership group that advances public health regulations and laws at all levels of government. From 2010 to 2018, he headed the Food and Consumer Safety Bureau at the Iowa Department of Inspections and Appeals.
John Tilden works for Michigan State University’s online food safety program. He previously held food safety investigation, oversight, and training roles with the U.S. Department of Agriculture, the Centers for Disease Control and Prevention, and the Michigan Department of Agriculture and Rural Development.
Their responses have been edited for clarity and length.
Tilden: Despite progress in food safety, we all know that there are still areas of inadequately controlled risk in our food supply. Root cause analysis provides a framework for tying together the bits and pieces —behavioral, environmental, food—and finding sustainable paths to improve contamination prevention. What I love is that you can use it at any point in the farm-to-fork continuum. It works in food service businesses and food manufacturing and on the farm. You just have to pull in different disciplines. That’s powerful, the potential to bring in specialists in water and irrigation or food manufacturing and processing. To borrow terms from the emergency management field, RCA is both modular and scalable.
Mandernach: If we don’t go into a food safety situation looking to find the root cause, how can we expect to fix it? Finding root causes and developing meaningful interventions to stop similar problems from occurring multiplies the impact of other oversight activities, from routine inspections to outbreak investigations.
Guzewich: I have done a lot of training over the years, and my goal has always been to get people to think beyond inspections. Inspections tend to box one in—you just look at the regulation and consider if the situation is a violation, and if the findings will be used as evidence. I’ve encouraged people to use root cause processes in outbreaks, because if they do, they end up taking a broader thought process to their routine regulatory work too.
Mandernach: In Iowa, we saw a complete shift in how our inspection staff approached their work once we trained them on these techniques and made them part of their job evaluation criteria. They’re more actively engaged with food facility managers and employees and are working together to develop interventions to improve food safety. It’s been a real win-win: Businesses get to resolve safety problems, and our agency can reduce the potential for foodborne illness. We received very positive feedback from companies on our move to this more consultative approach.
Tilden: We’ve trained Michigan’s local health departments on root cause analysis for about 20 years, and building a deep bench of staff with these skills has paid off. We saw that in the way one local health department responded to a hepatitis A outbreak at a retail food service operation. The local investigation traced the illnesses to an employee who had worked while sick. The department provided extensive training to the business’s staff and offered vaccinations to protect against the virus. But less than a month later, another food worker at the same establishment worked while sick with hepatitis A. That highlighted how knowledge is not the same as behavioral change. So, we used resources from CDC, the food industry, and human behavioral research to create a workshop that gave employers and regulators evidence-based ways to change workplace culture and behavior. Actions like allowing people to make up hours after an illness and having staff on call to fill in for co-workers can be more effective than just telling people not to work if they’re sick.
Tilden: Sizing the root cause analysis to the investigation is important. Not everything needs to be a research project. The issue is prevention: How do we stop the current outbreak and prevent a recurrence? Sometimes you can do that with a smaller investigation. Sometimes you have to pull out all the stops when you’re dealing with an emerging pathogen and don’t know much about how it’s spread or the risk factors for transmission.
Guzewich: Even if you have an actionable finding, it may not carry forward into food safety practices if it challenges conventional wisdom. For example, in 2010, there was a non-O157 Shiga toxin-producing E. coli outbreak linked to romaine from Yuma, Arizona. The investigation and report were done by a team, including a microbiologist, a hydrologist, and an inspector, that had collected samples of irrigation canal water that tested negative for the pathogen. But samples of the canal’s sediment tested positive, suggesting harmful E. coli could have been transferred to fields when sediment was stirred up into irrigation water. The problem was getting people to address this important finding—that contamination may not necessarily be detected if only water is sampled and tested. In fact, the finding in 2010 didn’t prevent another outbreak eight years later with the same phenomenon and canal system. They again found the outbreak strain in the sediment.
Guzewich: Historically, we have not done enough to give people the tools they need, particularly to pursue the behavioral aspects of the analysis. You need a clear procedure, inspection forms, and checklists, because many people who do RCAs infrequently are not confident in their abilities. They need guidance on how to conduct interviews in a nonregulatory context and how to quickly think of the follow-up questions to ask based on the first answer.
Mandernach: Using common language is also key. Often, we’ve created our own terms of art for similar concepts that differ from the language used by industry and others. That just makes it more complex for everyone to understand what’s going on. If we can use the same terms, we’re going to be much better off.
Tilden: Leverage what already exists. There are a lot of really good resources out there. CDC’s environmental assessment training is one. I think the [Pew] guide can whet people’s appetite for RCAs and connect them to helpful resources. Also, try to keep it simple. As the guide says: The contributing factor is what went wrong, and the root cause is why it happened. That’s a potent way to distill and communicate the takeaways in ways that people who weren’t involved in the investigation can understand rapidly.
Mandernach: The American Society for Quality has done phenomenal work and training on root cause analysis, to the point where I’m trying to figure out if we can partner with them to focus their courses for the regulatory community. Their resources are a great way to begin to develop the skills.
Guzewich: There are people in the industry who are doing this well, and I recommend incorporating their learnings and approaches. For instance, training professionals from industry and regulatory agencies together in root cause analysis may deepen the experience for everyone involved and enhance communication between these groups.