Preplanned Studies: Pathogenic Surveillance of Foodborne Illness-Related Diarrhea — Beijing Municipality, China, 2013–2023

Foodborne illnesses caused by various microorganisms, such as viruses, bacteria, and parasites (1), pose a significant global public health threats, leading to widespread illness and mortality. A survey conducted on the burden of acute gastrointestinal infections (AGI) in China between 2014 and 2015 revealed a population prevalence of 2.3% with an annual incidence of 0.3 episodes per person (2). Over the decade from 2010 to 2020, there were 18,331 reported outbreaks in Chinese catering facilities, resulting in 206,718 illnesses, 68,561 hospitalizations, and 201 fatalities (3). The Beijing CDC initiated a foodborne disease surveillance system in 2013, gradually implementing city-wide population-based surveillance. This study examined data from 36 actively monitored hospitals (25 tertiary-level and 11 secondary-level hospitals) selected through probability proportional to size (PPS) sampling from all hospitals with enteric disease clinics across 16 districts in Beijing Municipality, China.

Patients included in the surveillance were those who visited the sentinel hospital with symptoms of suspected foodborne diarrhea, presenting with over three bowel movements in 24 hours and abnormal stool consistency (e.g., loose, liquid, mucous, or bloody stools), excluding cases linked to antibiotic use or chemical exposure. Surveillance was carried out year-round, with each district aiming to collect a minimum of 330 samples annually.

Fresh stool specimens were collected either in fecal containers or rectal swabs, which were then placed in Cary-Blair transport medium for testing within 24 hours at 4°C (samples forVibrio parahaemolyticuswere stored at room temperature). Virus detection samples not immediately sent were stored at −20 °C. Bacterial specimens were enriched, inoculated onto suitable media for culture, and then isolated. Virus detection was conducted through nucleic acid amplification using reverse transcription-polymerase chain reaction (RT-PCR).

The 36 sentinel hospitals were responsible for gathering demographic information, food history, clinical characteristics, and biological samples from individuals under surveillance. A total of 16 district CDC laboratories tested these specimens for major foodborne pathogens:Salmonella,Vibrio parahaemolyticus, diarrheagenicEscherichia coli(DEC),Shigella, and Norovirus. The detection results were then sent to higher authorities for confirmation. Surveillance data indicated that Norovirus and DEC were the predominant pathogens, individuals between 19 and 30 years old had the highest infection rates, and patients withSalmonellainfections were more likely to experience fever symptoms.

Summary statistics, including frequencies and proportions, were computed for categorical variables. The study period was stratified into two time frames, 2013–2017 and 2018–2023, due to changes in testing practices at sentinel hospitals in 2018. The chi-square test was employed to compare demographic characteristics (age, sex, area, and occupation) for four pathogens (Salmonella, Norovirus,Vibrio parahaemolyticus, and DEC);Shigellawas excluded due to a limited sample size. Fisher’s exact test was used if the conditions for the chi-square test were not met. All tests were two-sided, and a significance level ofP<0.05 was considered statistically significant.

From January 1, 2013, to December 31, 2023, a total of 60,223 patients were included in the Beijing Foodborne Disease Active Surveillance System. After excluding non-infectious diarrhea cases and those without biological samples, 57,021 specimens were analyzed forSalmonella,Shigella,Vibrio parahaemolyticus, DEC, and Norovirus. The detection rates forSalmonella,Shigella,Vibrio parahaemolyticus, and DEC were 3.96% (2,260/57,021), 0.18% (101/57,021), 3.18% (1,811/57,021), and 7.06% (4,024/57,021), respectively. Among the 23,506 specimens tested for Norovirus, the detection rate was 10.54% (2,478/23,506). The prevalentSalmonellaserotypes wereSalmonellaEnteritidis (864/2,260) andSalmonellaTyphimurium (340/2,260). EnteroaggregativeEscherichia coli(EAEC) and enterotoxigenicEscherichia coli(ETEC) were the most common DEC types, constituting 34.57% (1,391/4,026) and 34.09% (1,372/4,026) of the detected DEC, respectively.

The trends in detection rates of various pathogens from 2013 to 2023 were analyzed and shown inFigure 1. From 2013 to 2019,Salmonella, DEC, and Norovirus infection rates increased annually, whileShigellainfections decreased.Vibrio parahaemolyticuspeaked in 2016, followed by a yearly decline. During 2020–2022, all pathogen infection rates decreased due to the impact of the COVID-19 pandemic. However, in 2023, detection rates increased for all pathogens exceptShigella. Notably, pathogenic bacteria were most prevalent in summer, with Norovirus causing winter and spring outbreaks.

Figure 1.

Changes in the numbers of positive pathogens and detection rates by pathogen and season in Beijing Municipality, China, from 2013 to 2023. (A) Norovirus; (B) DEC; (C)Salmonella; (D)Vibrio parahaemolyticus; (E)Shigella.

Abbreviation: DEC=diarrhoeagenic Escherichia coli.

The study provides a summary and comparison of the epidemiological characteristics of pathogens over time (Table 1). A high percentage of patients, 28.38% (2,971/10,469), are aged 19–30 years. The incidence ofVibrio parahaemolyticusinfection is significantly lower in children under 3 years old compared to other age groups (P<0.001). Gender ratios of individuals infected with Norovirus, DEC,Salmonella, andVibrio parahaemolyticusshow no significant differences between the periods 2013–2017 and 2018–2023 (P>0.05). However, there are notable disparities in the regional distribution of Norovirus (P<0.001) andSalmonella(P=0.003) infections between these time frames, with urban infections being more common in 2018–2023 compared to 2013–2017. The majority of Norovirus cases involve official staff (25.00%–29.62%), while cases of DEC,Salmonella, andVibrio parahaemolyticuspredominantly affect unemployed individuals (17.00%–28.77%). Significant differences exist in the occupational distribution of infection rates for different pathogens in different periods (P<0.001).

Figure 2displays patient self-reports, indicating that meat products were the most commonly reported source of suspected exposure among the four pathogens analyzed, accounting for 22.64%. However, there were differences in the distribution of suspicious food categories across the various pathogens. ForVibrio parahaemolyticusinfections, aquatic products were the primary food source at 26.06%, followed by meat products at 24.74% and vegetables at 15.74%. Conversely, patients infected with Norovirus,Salmonella, and DEC identified meat products and vegetables as the top two food sources.

Figure 2.

Distribution of suspicious food categories by pathogens in Beijing Municipality, China, 2013–2023.

Abbreviation: DEC=diarrhoeagenic Escherichia coli.

The clinical characteristics of the pathogens are outlined (Table 2). Nausea was the most common symptom, affecting 51.05% (1,265/2,478) of Norovirus-infected patients. Abdominal cramps were prevalent in DEC,Salmonella, andVibrio parahaemolyticusinfections, ranging from 41.08% to 73.86%.Salmonellainfections showed a significantly higher fever prevalence (36.02%) compared to other pathogens during 2013–2023. Across eleven years, the proportion of abdominal cramps from these infections decreased notably from 53.72%–73.86% (2013–2017) to 40.63%–56.93% (2018–2023). The projected trend suggests an increase in patients experiencing loose stools due to pathogens, expected to rise from 16.75%–28.27% (2013–2017) to 35.32%–45.90% (2018–2023).