Environmental Antecedents of Foodborne Illness Outbreaks, United States, 2017–2019

Foodborne outbreak investigations often provide data for public health officials to determine how the environment contributed to the outbreak and on how to prevent future outbreaks. State and local health departments are responsible for investigating foodborne illness outbreaks in their jurisdictions and reporting the data to national-level surveillance systems, including information from the environmental assessment. This assessment is designed to describe how the environment contributed to the outbreak and identifies factors that contributed to the outbreak and environmental antecedents to the outbreak. Environmental antecedents, also referred to as root causes, are specific reasons that allow biological or chemical agents to contaminate, survive, or grow in food. From 2017 to 2019, 24 jurisdictions reported 1,430 antecedents from 393 outbreaks to the National Environmental Assessment Reporting System. The most reported antecedents were lack of oversight of employees/enforcement of policies (89.1%), lack of training of employees on specific processes (74.0%), and lack of a food safety culture/attitude towards food safety (57.5%). These findings highlight the critical role that employees play in restaurant food safety and are heavily influenced by restaurant management, who can exercise active managerial control to manage these antecedents. Identifying antecedents during investigations is essential for understanding the outbreak’s root cause and implementing sustainable corrective actions to stop the immediate outbreak and future outbreaks.

The Centers for Disease Control and Prevention (CDC) estimates that around 48 million foodborne illnesses occur each year in the United States.This number includes 128,000 hospitalizations and 3,000 deaths (Scallan et al., 2011).Illnesses associated with outbreaks account for a small portion of the annual foodborne illness incidence (Dewey-Mattia et al., 2018).However, data from foodborne outbreak investigations can provide detailed information that helps determine how the environment contributes to outbreaks.That information can be used to prevent future outbreaks (Dewey-Mattia et al., 2018).The lessons we learn from outbreak investigations can also be used to prevent sporadic cases of foodborne illnesses.
State and local health departments are responsible for investigating foodborne illness outbreaks and reporting the data to national-level surveillance systems managed by the CDC.CDC aggregates these data to detect national trends and identify systematic food safety issues.The National Environmental Assessment Reporting System (NEARS) is a CDC surveillance system through which some state and local health departments voluntarily report data from their investigations of foodborne illness outbreaks at retail food establishments (Centers for Disease Control and Prevention, 2022b).Specifically, they report data from the environmental health component of the investigation, typically called the environmental assessment.The assessment is designed to describe how the outbreak environment contributed to the introduction or transmission of the agent that caused the outbreak (Centers for Disease Control and Prevention, 2022c).NEARS collects two key sets of environmental assessment data about the outbreak (Freeland et al., 2019;Moritz et al., 2023).
The first set is outbreak contributing factors-the practices that most likely led to contamination, proliferation, or survival of agents in the environment (Centers for Disease Control and Prevention, 2023).The second set is environmental antecedents, hereafter referred to as antecedents, of the outbreak.Antecedents are the circumstances and situations that led to the contributing factors.For example, an investigation may identify the contributing factor to a norovirus outbreak as contamination of food by an ill employee, and the antecedent as insufficient staffing, which led to the employee not being allowed to stay home while ill.
Identifying the contributing factors and antecedents is essential to understanding the outbreak's root cause and implementing sustainable corrective actions to stop the outbreak and future outbreaks (Firestone et al., 2018).Collecting and analyzing these data at a national level are essential for informing efforts to prevent future outbreaks in retail environments across the country.The data can be used to inform national food safety policy and training focused on reducing contributing factors and antecedents (Centers for Disease Control and Prevention, 2022a;Matis et al., 2017;Wittry et al., 2022).To contribute to these goals, the current study presents data on the contributing factors and antecedents for outbreaks reported to NEARS from 2017 to 2019.

Materials and Methods
Between 2017 and 2019, state and local health departments reported data to NEARS from 799 outbreak investigations with a suspected or confirmed foodborne agent.CDC defines a foodborne illness outbreak as an incident in which two or more people experience a similar illness resulting from the ingestion of a common food (Centers for Disease Control and Prevention, 2015).The dataset for this paper is limited to the 393 outbreaks that had at least one reported antecedent.The outbreaks occurred in Alaska, California, Connecticut, Delaware, Fairfax County (VA), Georgia, Harris County (TX), Indiana, Iowa, Jefferson County (CO), Maricopa County (AZ), Massachusetts, Michigan, Minnesota, New York City, New York State, North Carolina, Oregon, Rhode Island, South Carolina, the Southern Nevada Health District, Tennessee, Washington, and Wisconsin.
Investigators conducting environmental assessments for outbreaks in retail food establishments engage in a variety of data collection activities.These include kitchen observations, reviews of records, and staff interviews.Data from these activities, along with data obtained from epidemiological and laboratory investigations, are reported to NEARS (Lipcsei et al., 2019;Moritz et al., 2023).Additionally, investigators report any contributing factors or antecedents they identified during their outbreak investigation.The identification is based on their review of the totality of the data collected during the observation and their professional judgment and experience.Investigators select contributing factors from a predetermined list of 32 contributing factors.The list of factors is derived from an analysis of historical data from outbreak investigations (Bryan, 1978;Weingold et al., 1994).The investigators also select antecedents from a list of 17 potential antecedents and an option to write in an 'other' response.This list was developed by food safety experts using the food system environmental antecedent conceptual model (Selman & Guzewich, 2014).The model theorizes that there are five main groups of environmental antecedents that influence food safety in establishments.The groups are people (behaviors, characteristics, and attitudes of people working in the establishments); processes (characteristics of the processes used to prepare food and food preparation complexity); economics (costs and profit margins); equipment (the physical layout and equipment of establishments); and food (the inherent qualities of food prepared in establishments).Investigators can report multiple contributing factors and antecedents for each outbreak.
The top three antecedents fell into the people category: lack of oversight of employees (89.1%), lack of training of employees on specific processes (74.0%), and lack of food safety (57.5%)(Table 3).The next two most common antecedents fell into the process category: staff not following the facility's processes (24.9%) and insufficient process to mitigate hazard (23.9%).

Outbreak antecedents by etiologic agent.
The three most common etiologic agents were the virus norovirus, and the bacteria Salmonella and Clostridium perfringens.The three most common antecedents for all three types of outbreaks were people antecedents: lack of oversight of employees (78.8%, 60.4%, 48.0%, respectively), lack of training of employees (51.3%, 54.7%, and 56.0%, respectively), and lack of food safety culture (39.1%, 52.8%, and 36.0%,respectively) (Table 4).However, the antecedent of lack of oversight of employees was identified far more often for norovirus outbreaks than for the bacterial outbreaks (78.8% vs 60.4%, 48.0%).Additionally, lack of food safety culture was identified more often for Salmonella than for Clostridium perfringens (52.8% vs. 36.0%).
All three types of outbreaks had equipment antecedents identified, but again, the proportion differed by etiologic agent.More often, the two types of bacterial outbreaks, compared to norovirus outbreaks, had improperly used equipment identified as an antecedent (9.4% and 12.0% vs. 1.3%) (Table 4).Similarly, more often Clostridium perfringens outbreaks, compared to norovirus outbreaks, had insufficient equipment capacity identified as an antecedent (28.0% vs. 3.2%).
Lack of sick leave was identified as an antecedent for all three types of outbreaks but was more often identified for norovirus outbreaks than for the two types of bacterial outbreaks (26.3% vs. 9.4% and 4.0%) (Table 4).On the other hand, insufficient process to mitigate hazards was identified more often for the two types of bacterial outbreaks than for norovirus outbreaks (22.6% and 36.0%vs. 8.3%).

Outbreak antecedents by contributing factors.
The three most common contributing factors fell into the contamination categoryother mode of contamination by a suspected infectious worker, cross-contamination of ingredients, and bare-hand contact by a suspected infectious worker (Table 5).The three most common antecedents for the top three contributing factors were the people antecedents of lack of oversight of employees (79.2%, 71.4%, 76.8%, respectively), lack of training of employees (51.0%, 62.5%, and 66.1%), and lack of food safety culture (31.3%, 55.4%, 53.6%).However, the two antecedents of lack of employee training and lack of food safety culture were identified more often for the contributing factors of cross-contamination of ingredients and bare-hand contact by a suspected infectious worker than for other mode of contamination by a suspected infectious worker (62.5%, 66.1% vs. 51.0%;55.4% and 53.6% vs. 31.3%).
The equipment antecedent of poor facility layout was identified more often for crosscontamination of ingredients than for the two contributing factors associated with suspected infectious workers (12.5% vs. 4.2% and 1.8%) (Table 5).On the other hand, the economics antecedent of lack of sick leave was more often identified for the two contributing factors of suspected infectious worker than for cross-contamination of ingredients (30.2% and 25.0%, vs. 5.4%).Similarly, the people antecedent of low or insufficient staffing was identified more often for bare-hand contact by a suspected infectious worker than for cross contamination of ingredients (14.3% vs. 5.4%).

Discussion
People, lack of oversight.
The three most common outbreak antecedents identified by investigators were in the people category: lack of oversight of employees, lack of training of employees, and lack of food safety culture.Lack of a food safety culture is defined as the shared values, beliefs and norms that affect mind-set and behavior toward food safety in an organization (Yiannis, 2009).These findings highlight the critical role that employees play in restaurant food safety.The most common antecedents were observed in norovirus, Salmonella, and Clostridium perfringens outbreaks, indicating that these antecedents support both viral and bacterial outbreaks.However, the antecedent of lack of oversight of employees was identified much more often (> 25 percentage points) for norovirus outbreaks than for the two types of bacterial outbreaks.This suggests a particularly strong association between lack of oversight of employees and norovirus outbreaks.Foodborne norovirus outbreaks are more often associated with ill employees contaminating food than are bacterial outbreaks (Hall et al., 2012).Indeed, most foodborne norovirus outbreaks are associated with ill employees (Hall et al., 2012).Restaurant managers are responsible for preventing employees who are ill with foodborne illness symptoms from working (Food and Drug Administration, 2022).Our data suggest that managerial failure to exclude ill employees from working is a key antecedent to norovirus outbreaks.

People, lack of sick leave.
Similarly, lack of sick leave for employees was identified as an antecedent more often for norovirus outbreaks (> 18 percentage points) than for Salmonella and Clostridium perfringens outbreaks.This antecedent was also identified more often for the two suspected infectious worker contributing factors.Given that ill employees are more often associated with norovirus outbreaks than with bacterial outbreaks, lack of sick leave as an antecedent to norovirus outbreaks and to contributing factors related to ill employees is not surprising.Ill employees may not feel they can lose pay; thus, they work while ill and transmit their illness to customers through the food they contaminate.This finding is supported by other research showing that financial issues, such as lack of sick leave and loss of shifts, are cited by employees as reasons for working while ill (Carpenter et al., 2013;Sumner et al., 2011;Norton et al., 2015).Research also shows that mandated paid sick leave reduced instances of employees working while ill (Schneider, 2020).Restaurants can take other steps to manage ill employees.For example, they can create written policies, address reasons why employees work while ill, and create schedules for when an employee is ill (Centers for Disease Control and Prevention, 2022d).

Active managerial control.
The top three antecedents (lack of oversight of employees, lack of training of employees, lack of food safety culture) and other people antecedents, such as lack of sick leave, are heavily influenced by retail food establishment managers.Active managerial control is the purposeful incorporation of specific actions or procedures by industry management into the operation of their business to attain control over foodborne illness risk factors.Active managerial control can be used to encourage proper food safety practices (Food and Drug Administration, 2022).Some state and local health departments provide an active managerial control toolkit or resource center to assist food establishments in promoting and implementing this concept in their operations (Alexander-Leeder & Gzebb, 2023;Maricopa County Department of Public Health, 2023).State and local health departments have also encouraged food establishments to develop and adopt active managerial control in their operations through incentive programs.Regulators indicated this innovative strategy has been successful as they observed fewer violations in establishments resulting in fewer required inspections by the health department (Retail Food Safety Regulatory Association Collaborative, 2023).

Equipment.
We found that equipment-related antecedents were more common for bacterial outbreaks than norovirus outbreaks.Restaurants typically rely on equipment for temperature control (e.g., walk-in coolers for cold holding, bain-maries for hot holding) (McCabe-Sellers & Beattie, 2004).When this equipment fails or is used improperly, bacterial pathogens survive and proliferate in food.For example, Wittry et al. found that a common cause of Clostridium perfringens outbreaks is the use of hot-holding equipment that is not large enough for the establishment's operational demand (Wittry et al., 2022).Our finding highlights the importance of using the proper type of equipment and using the equipment properly to prevent bacterial outbreaks.

Poor facility layout.
The antecedent of poor facility layout was identified more often with the contributing factor of cross-contamination than with contributing factors related to ill employees.Poor facility layout, possibly due to a lack of space or poor design, can support ingredient cross-contamination.For example, a meat preparation area immediately adjacent to a salad preparation area could lead to raw meat juices contaminating the salad preparation area and ingredients.To prevent ingredient cross-contamination due to poor facility layout, the Food and Drug Administration (FDA) recommends that restaurants systematically identify their risks and develop plans for mitigating those risks (Food and Drug Administration, 2022).Ideally, these risks are identified during a construction plan review using hazard analysis critical control point (HACCP) principles as a tool to design safe facility layouts (Conference for Food Protection, 2016).HACCP plans are recommended by the FDA, endorsed by most state and local health departments, and are one of the best ways to mitigate risks in a food establishment (Center for Food Safety and Applied Nutrition, 2006).

Variations by year.
There are some variations of etiologic agents, contributing factors, and antecedents by year, which could be explained by advanced laboratory testing methods and a push for increased food safety.Norovirus outbreaks decreased while Salmonella outbreaks slightly increased.This change could be due to a wider use of whole genome sequencing (WGS).WGS provides a faster, more accurate detection of bacteria (e.g., Salmonella), which may have previously been labeled as unknown (Centers for Disease Control and Prevention, 2022e; Rounds, et al., 2020).Contributing factors related to ill food workers decreased from 2017 to 2019.The FDA Risk Factor Study data from 1998 to 2009 showed that compliance for no bare-hand contact with ready-to-eat foods increased over the 10-year study (Food and Drug Administration, 2010), which are actions that prevent food contamination.While this study shows similar results, more research is needed to determine what drives the decrease in bare-hand contact with ready-to-eat foods.Over the past decade, CDC and FDA research findings have focused on improving these food safety practices to prevent norovirus contamination in retail food establishments (Kambhampati et al., 2016;Hoover et al., 2020;Centers for Disease Control and Prevention, 2019).

Limitations.
There are a few limitations to this study.NEARS is a voluntary reporting system and although the reporting sites are geographically diverse, the data may not be representative of all foodborne outbreaks.The number of outbreaks at each site is likely underreported because health departments do not detect or investigate all outbreaks.Another limitation is the variability in investigation practices across health departments due to different jurisdictional policies and trainings.Identification of antecedents is based on the investigator's judgment and these judgments may vary.
The findings from this study provide valuable and novel information about antecedents to foodborne outbreaks.The most common antecedents to foodborne illness outbreaks we identified (lack of employee training, lack of employee oversight, and lack of food safety culture) are heavily influenced by restaurant management, who can exercise active managerial control to mitigate these antecedents.Establishment management can work to implement active managerial control in their establishments to help ensure effective oversight and training for employees.There are many resources, from FDA and various health departments across the United States, available to health departments to assist establishments in exercising active managerial control (Food and Drug Administration, 2022;Alexander-Leeder & Gzebb, 2023;Maricopa County Department of Public Health, 2023).Interventions to address common root causes of outbreaks could focus on people as root causes and managerial leadership to improve food safety in an establishment (Kramer et al., 2023;Lee et al., 2021).Identifying antecedents during investigations is essential for understanding the outbreak's root cause and implementing sustainable corrective actions to stop the immediate outbreak and future outbreaks.Health departments are encouraged to conduct a thorough environmental assessment and identify the root cause of outbreaks to prevent future outbreaks.The denominator is the number of outbreaks that reported the agent (confirmed or suspected).
J Food Prot.Author manuscript; available in PMC 2024 July 01.J Food Prot.Author manuscript; available in PMC 2024 July 01.

Foodborne
outbreak environmental antecedents by etiologic agent, National Environmental Assessment Reporting System, 2017more than 1 agent and environmental antecedent.b

Foodborne
outbreak environmental antecedents by contributing factor, National Environmental Assessment Reporting System, 2017more than 1 contributing factor and environmental antecedent.

Table 1 .
Foodborne outbreak etiologic agents-National Environmental Assessment Reporting System, 2017 -2019 b The denominator is the number of outbreaks that identified an environmental antecedent.J Food Prot.Author manuscript; available in PMC 2024 July 01.

Survival of foodborne illness agents after a process, such as cooking, that should have eliminated or reduced them a
J Food Prot.Author manuscript; available in PMC 2024 July 01.

Table 3 .
Foodborne outbreak environmental antecedents, National Environmental Assessment Reporting System, 2017-2019 The denominator is the number of environmental antecedents reported for all outbreaks that reported at least one environmental antecedent.c The denominator is the number of outbreaks that reported an environmental antecedent.
b J Food Prot.Author manuscript; available in PMC 2024 July 01.