Finalized outbreak reports were extracted on 24 February 2021 from NORS for outbreaks with first illness onset occurring during 2009–2019. Outbreaks in NORS are defined as 2 or more cases of similar illness associated with a common exposure. NORS captures outbreaks with exposures in US states, the District of Columbia, and US territories and Freely Associated States (referred to hereafter as “reporting sites”). Annual population estimates from 2009–2019 were available from the US Census Bureau for the 50 US states, the District of Columbia, and Puerto Rico and were used to calculate average annual outbreak reporting rates, expressed per 1 million population [7, 8]. Outbreaks in which exposure occurred in more than 1 reporting site (“multisite outbreaks”) were assigned as an outbreak to each site involved for the purpose of rate calculations. Primary mode of transmission was determined by each reporting site. Outbreak and case definitions might not be consistent across all sites. Medians were compared using the Wilcoxon rank-sum test. All analyses were conducted in SAS version 9.4 (SAS Institute).

The setting of exposure was the setting where patients were exposed to the outbreak etiology; for foodborne outbreaks, this was the location(s) of food preparation. Four of the most commonly reported symptoms (diarrhea, vomiting, bloody stools, and fever) were used to assess the symptom profile of patients. Not all symptoms or patient outcomes were reported for all outbreaks, and data were not available for all patients on specific symptoms or outcomes. Therefore, symptoms and outcomes are presented as the percentage of patients reporting that symptom or outcome among all patients with that information available.

Outbreak reports with only a reported nonenteric etiology were excluded from analysis (see Supplementary Methods and Supplementary Table 1). If no specific etiology was reported for an outbreak, the report was included if any patients reported symptoms of diarrhea, vomiting, bloody stools, or unspecified acute gastroenteritis. Outbreaks with no reported etiology and no symptom information were excluded from analysis. Outbreak etiologies were considered laboratory-confirmed if defined as such by the reporting site or if 2 or more laboratory-confirmed cases were identified. Unless otherwise specified, all figures and analyses include both laboratory-confirmed and suspected etiologies.

Outbreak reports form the basis for attribution estimations when considering policies for prevention, and robust outbreak surveillance is vital in guiding appropriate, targeted interventions [9, 10]. The CDC conducted foodborne and waterborne disease outbreak surveillance prior to creating NORS, but these 2 modes of transmission do not account for all enteric illness outbreaks. The number of enteric illness outbreaks reported through NORS more than doubled from 2009 to 2019, largely due to inclusion of person-to-person transmission starting in 2009. Person-to-person contact was the primary mode of transmission in nearly two-thirds of enteric illness outbreaks, and it is likely that sites needed additional time to implement reporting of these outbreaks. The increased reporting over time may reflect this additional time needed rather than a true increase in the number of outbreaks occurring.

Outbreaks were reported to NORS with exposures occurring in all 50 US states; Washington, DC; Puerto Rico; and American Samoa. However, reporting rates varied widely by reporting site, with the highest reporting rate more than 20 times higher than the lowest reporting rate. The 3 most populous US states—California, Texas, and Florida—all had outbreak reporting rates in the lowest quartile. Reporting also varied by mode, with fewer sites reporting person-to-person outbreaks than foodborne outbreaks even during recent years of surveillance. Differences in reporting rates by site and mode may thus be a function of resource availability and prioritization for outbreak investigation rather than a true reflection of outbreak incidence.

Norovirus was the most common suspected or laboratory-confirmed outbreak etiology, identified in over half (59%) of all reported enteric illness outbreaks. Norovirus spreads easily by many modes of transmission, due to a low infectious dose and ability to survive for long periods on surfaces [11]. Typically, norovirus causes a self-limited illness characterized by diarrhea and vomiting, although it can cause severe outcomes in the elderly and immunocompromised [12]. Norovirus outbreaks in this study exhibited a strong winter seasonality and were most often reported in LTCFs, consistent with other studies [13, 14]. Norovirus is also the leading cause of foodborne enteric illness outbreaks, identified here in 44% of foodborne outbreaks with a reported etiology, consistent with previous studies [15]. Because norovirus outbreaks are so prevalent and cause significantly larger outbreaks than other common etiologies, norovirus is also the leading cause of enteric illness outbreak–associated hospitalizations and deaths, despite being associated with comparatively low case fatality and hospitalization rates. Outbreaks with no reported etiology also exhibited a strong winter seasonality, were most often spread by person-to-person transmission, and were most often reported in LTCFs, similar to norovirus and suggesting norovirus as a possible etiology. Salmonella was identified as a leading cause of foodborne enteric outbreaks, second only to norovirus, in this study as in other studies [15]. Salmonella outbreaks were responsible for nearly two-thirds of multisite outbreaks. Salmonella has long been recognized as the predominant cause of multisite outbreaks, which have been on the rise as food distribution, consumption habits, and laboratory testing methods have evolved [15, 16]. PulseNet, a laboratory-based subtyping network, and the use of whole-genome sequencing (WGS) aid in the detection of more clusters of illness more quickly, linking cases to one another, and traceback to a source [17, 18]. Multisite outbreaks generally cause more illnesses than single-site outbreaks and, when combined with the high hospitalization rate of Salmonella outbreaks (20.9% in this study), can contribute disproportionately to more illnesses and hospitalizations [16].

Shigella was also identified as a leading cause of enteric illness outbreaks in this study. Similar to norovirus outbreaks, Shigella outbreaks also mostly spread through direct person-to-person contact but were more often identified in school and daycare settings. Shigella is a well-recognized cause of enteric illness outbreaks in children [19, 20]. While Shigella is considered a year-round enteric pathogen, these findings align with other studies that have found a slight increase in cases in summer months [21, 22]. Shigella often causes a mild, self-limiting diarrheal illness, as illustrated by the very few Shigella outbreak–related deaths reported to NORS [23]. Previous studies have shown that patients with bloody diarrhea and concurrent fever, symptoms commonly reported in Shigella outbreaks, are more likely to seek medical care than those without, potentially contributing to the high hospitalization rate seen in this study [24].

The other leading bacterial enteric pathogens—Salmonella, E. coli, and Campylobacter—were all most commonly associated with foodborne outbreaks. Outbreaks of these pathogens were identified year-round, but this study found a marked increase in the summer months, as has been noted elsewhere [21, 22]. While more than half of norovirus and Shigella outbreaks occurred in a single identified setting (LTCFs and schools or daycare centers, respectively), outbreaks caused by Salmonella, Campylobacter, and E. coli did not have a predominant single setting. This may be due in part to upstream contamination of food products that are then prepared and consumed in a variety of settings, rather than point-source contamination of food by an infected foodworker, as is often the case in foodborne norovirus outbreaks [25].

Although Salmonella outbreaks contributed the most absolute numbers of deaths and hospitalizations after norovirus, the highest case fatality and hospitalization rates in this study were attributed to E. coli outbreaks. While many strains of E. coli are relatively harmless, other strains such as Shiga toxin–producing E. coli are a well-known cause of severe and often bloody diarrhea and can lead to serious outcomes like hemolytic uremic syndrome and death [26]. Campylobacter infection typically causes a mild illness but can lead to severe complications and is the leading bacterial cause of diarrheal illness in the United States [27]. In this study, only 1 death was attributed to a Campylobacter outbreak, but nearly 9% of patients were hospitalized. Like Shigella and other bacterial pathogens, multidrug resistance of Salmonella and Campylobacter is of significant concern and may, in turn, lead to more severe outcomes [28, 29]. Prevention efforts to reduce the overall burden of Salmonella, Campylobacter, and E. coli outbreaks must address multidrug resistance, multiple modes of transmission, contamination throughout the food chain, and the diversity of settings in which these outbreaks have been reported.

There are multiple limitations to this study. NORS is a voluntary reporting system that relies on state, local, and territorial public health agencies to detect, investigate, and report outbreaks. Therefore, data captured by NORS may not be generalizable to all outbreaks or illness associated with the etiologies reported here. Persons with an enteric illness often do not seek medical care, and among those who do, few provide a stool specimen for etiologic testing [24]. Variability in reporting by site and mode of transmission may introduce bias in which outbreaks are reported via NORS. NORS also collects data on foodborne and waterborne outbreaks of nonenteric illness; our inclusion criteria may have excluded some enteric illness outbreaks that did not include either an etiology or symptom information in the outbreak report, or else included outbreaks of nonenteric illness due to a pathogen that can cause both enteric and nonenteric illness. Nearly 60% of outbreak reports did not have a laboratory-confirmed etiology, and 19% did not report any suspected or confirmed etiology.

This study reviewed 11 years of data on enteric illness outbreaks for all reported etiologies and modes of transmission, building on the knowledge gained through decades of foodborne and waterborne outbreak surveillance. Norovirus is the leading cause of enteric illness outbreaks as well as outbreak-related deaths and hospitalizations in the United States. However, outbreaks due to common bacterial etiologies like Shigella and Salmonella continue to contribute significantly to morbidity and mortality. The continued efforts of local public health agencies to investigate and report enteric illness outbreaks to CDC and collect and test specimens are essential in understanding the national burden of enteric illness outbreaks. Expanded enteric illness outbreak reporting through NORS has provided a more comprehensive picture of the relative contribution of each outbreak etiology and mode of transmission and the characteristics of those outbreaks. Robust NORS data can, in turn, help track trends over time and guide targeted interventions by identifying settings, sources, and patients most at risk for enteric illness.