Epidemiology of Norovirus Outbreaks in Kindergartens, Primary Schools, and Junior High Schools in Xi’an, China, From October 2020 to February 2023

Introduction

Norovirus is the only highly infectious pathogen currently recognized as having the potential to cause a global pandemic of acute gastroenteritis (AGE) [1]. Globally, norovirus is responsible for an estimated 685 million cases of illness annually, including approximately 200 million cases in children aged 5 years or younger [2]. This burden results in around 200 000 deaths each year, primarily in developing countries, and is associated with global economic losses estimated at $58.5 billion [3]. In addition, norovirus causes at least 50% of AGE outbreaks and is the primary cause of foodborne illness worldwide [4]. In China, norovirus-related AGE outbreaks have increased rapidly since their first report in 2006, drawing attention from public health agencies and government authorities [5]. Between 2017 and 2021, norovirus caused 89.92% of all infectious diarrhea-related public health emergencies nationwide [6]. Notably, over 65% of norovirus outbreaks occur in educational settings [7], such as schools and kindergartens, which causes significant disruptions to teaching activities and broader socio-economic operations, highlighting the urgent need for more effective prevention and control measures.

Currently, the absence of specific therapeutic drugs and commercially available vaccines makes prevention and control measures for norovirus outbreaks notably limited, relying predominantly on non-pharmaceutical interventions (NPIs) [3,8]. However, the implementation of NPIs in emergency responses to norovirus outbreaks in China frequently encounters several critical challenges. The first challenge lies in determining whether an acute gastroenteritis outbreak is caused by norovirus, based on clinical symptoms such as vomiting and diarrhea. In practice, the Kaplan criteria, which define a suspected norovirus outbreak as one in which more than 50% of cases involve vomiting [9], and the “Guidelines on Outbreak Investigation, Prevention, and Control of Norovirus Infection (2015)” of China (hereafter referred to as the national guidelines), which specify a diarrhea rate exceeding 52% [8], are often insufficiently accurate. The second challenge is assessing the scale of a norovirus outbreak. So far, limited research has focused on the attack rate of norovirus outbreaks [10,11]. The third challenge is to determine the appropriate policies for class closures during norovirus outbreaks. In China, the typical response to norovirus outbreaks in schools and kindergartens involves temporarily suspending symptomatic students from attending school, followed by isolating them at home [12]. Asymptomatic infected students and teachers are often overlooked and continue to attend school [12]. However, fecal shedding from asymptomatic infected individuals can still cause new AGE outbreaks and prolong their duration [13–15]. The final challenge involves determining the appropriate sampling volume for AGE cases during norovirus outbreaks to optimize the use of limited healthcare resources. Therefore, this study aimed to evaluate several intriguing epidemiological characteristics – vomiting rate, diarrhea rate, attack rate, asymptomatic prevalence, and norovirus detection rate – of norovirus outbreaks reported to the Xi’an Center for Disease Control and Prevention (CDC) in kindergartens, primary schools, and junior high schools between October 2020 and February 2023, to inform policy adjustments and enhance response strategies.

Material and Methods

ETHICAL STATEMENT:

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Xi’an CDC, China (Approval No. 2020012). Verbal informed consent was obtained from all participants or their legal guardians. According to Article 36 of China’s “Regulations on Emergency Response to Public Health Emergencies”, all entities and individuals within the territory of the People’s Republic of China are obligated to support infectious disease prevention and control efforts and to cooperate with legally mandated measures, including investigation, specimen collection, testing, isolation and treatment, and medical observation, aimed at preventing and mitigating the adverse impacts of infectious diseases [16]. In addition, the rectal swab specimen sampling procedure was non-invasive and posed no risk to the participants. Given that this study was part of an emergency response to these norovirus outbreaks and the harmlessness of sampling, the requirement for written informed consent was waived in compliance with the regulations.

INTRODUCTION TO THE XI’AN CDC:

Xi’an CDC (http://www.xiancdc.com/) is a public welfare institution established by the municipal government of Xi’an, Shaanxi Province, China. It is responsible for providing essential public health services to the residents of Xi’an, a city located in the northwest region of the country. Its core duties include the prevention and control of infectious, parasitic, and non-communicable diseases; responding to public health emergencies and disasters; managing epidemic-related information and health risk factors; implementing interventions to mitigate health risks; identifying pathogenic organisms; detecting and evaluating physical and chemical health hazards; promoting health education and awareness; and offering technical guidance and supporting research in the field of disease prevention and control.

STUDY DESIGN AND SUBJECTS:

The study design was cross-sectional. The population affected by norovirus outbreaks in all kindergartens, primary schools, and junior high schools in Xi’an city from October 2020 to February 2023 was enrolled in this prospective study. According to the Compulsory Education Law of the People’s Republic of China, children aged 3–5.9 years attend preschool (kindergarten), students aged 6–11.9 are enrolled in primary schools, and students aged 12–15 are enrolled in junior high schools. Due to the unpredictable nature of outbreak emergence, asymptomatic individuals were selected from the exposed population from each outbreak using simple random sampling and were subsequently tested. Rectal swab specimens were collected and tested using real-time reverse transcription-polymerase chain reaction (RT-PCR) in strict adherence to the national guidelines [8].

RELATED DEFINITIONS:

A norovirus outbreak was defined as the occurrence of 5 or more epidemiologically linked AGE cases in public settings within a 3-day period, with at least 2 specimens (whole stool, rectal swab, or vomitus) testing positive for norovirus [10]. The other relevant definitions used in this study are shown in Tables 1 and 2. In this study, cases included both clinically confirmed and laboratory-confirmed cases, excluding individuals who were infected but remained asymptomatic.

DATA SOURCE:

Data for this study were manually collected from norovirus outbreaks in all schools as well as kindergartens under the unified surveillance and response of the Xi’an CDC, between October 2020 and February 2023. The raw data on norovirus outbreaks are not available in the form of structured datasets, but are instead documented in epidemiological investigation reports, case questionnaires, outbreak reports received, emergency response records, laboratory test result reports, and other relevant materials.

QUALITY ASSESSMENT:

The quality of outbreak investigations was assessed using well-established standards derived from the literature [19]. The aggregate score was 5. Scoring was 1 point for each “yes” answer, 0 for “no” or “unclear”. Quality rating was 3–5=high, 1–2=moderate, 0=low. Outbreak-related data of low quality were deemed unreliable and were thus excluded.

DATA EXTRACTION:

Specific data extracted from the remaining norovirus outbreaks included: the number of cases, the number of exposed individuals, the number of cases exhibiting vomiting or diarrhea symptoms, the number of AGE cases tested, the number of norovirus-positive AGE cases, the number of asymptomatic individuals tested, the number of norovirus-positive asymptomatic individuals, sample types, and norovirus genome types. Data were entered into the EpiData 3.0 database by 2 independent researchers, followed by consistency checks. In the event of data inconsistencies, the lead investigators involved in the outbreaks conducted a review of the original data to reach final determinations. Full data regarding these outbreaks were collected in March 2023, and the processes of data extraction, correction, and collation were finalized in November.

STATISTICAL ANALYSIS:

Descriptive statistics were used to analyze the epidemiological characteristics of these norovirus outbreaks. Categorical variables were expressed as numbers and proportions, while continuous variables were reported as medians and interquartile ranges (M[P25, P75]) due to the non-normal distribution of the data; therefore, differences in rates among subgroups were analyzed using the Wilcoxon signed-rank test and the Kruskal-Wallis H test [19,20]. A single-group rate meta-analysis was applied to calculate the vomiting rate, diarrhea rate, attack rate, asymptomatic prevalence, and norovirus detection rate of these norovirus outbreaks [17,21–23]. To address instances where the number of norovirus-positive asymptomatic individuals was zero, 0.5 was added to the count for calculation purposes [20,23]. The I2 test was used to assess heterogeneity among outbreaks. Due to the expected high heterogeneity, random-effect models were utilized for the analysis. The results of the single-group rate meta-analysis are descriptive and not indicative of comparative differences. Given the high potential for publication bias (as results are almost always “positive”) and high stability when removing any individual study in single-group rate meta-analyses, publication bias tests and sensitivity analyses, while potentially useful, were not deemed essential in this context [17,23]. The single-group rate meta-analysis and statistical tests were performed using R software version 4.0.2 (R Development Core Team, Vienna, Austria). All p values were calculated using two-sided tests, with statistical significance defined as p<0.05.

Results

BASELINE CHARACTERISTICS OF NOROVIRUS OUTBREAKS:

Between October 2020 and February 2023, the Xi’an CDC actively monitored and responded to 61 norovirus outbreaks. Of these, 44 outbreaks were included in this study due to the availability of comprehensive data and high-quality information obtained during field investigations. All outbreaks were attributed to human-to-human transmission. According to the quality evaluation standards for outbreak investigations, all 44 outbreaks scored 4 to 5 points, indicating high-quality survey results. Among the 44 outbreaks, 33 occurred in kindergartens, 9 in primary schools, and 2 in junior high schools, involving 768 cases (438 males and 330 females; 295 laboratory-confirmed cases and 473 clinically confirmed cases) and 7945 exposed individuals. Of the 768 cases, 616 experienced vomiting, 273 reported abdominal pain, 176 developed fever, and 147 exhibited diarrhea. Of note, some individuals may have experienced multiple symptoms. Specifically, 71 cases were reported among 2431 junior high school students, 171 cases among 3562 primary school students, 525 cases among 1949 kindergarten children, and 1 case among 3 kindergarten staff members. The highest number of outbreaks occurred in spring (21, 47.73%), followed by winter (15, 34.09%), autumn (5, 11.36%), and summer (3, 6.82%). Outbreaks in the spring semester accounted for 81.82% (36/44), while the autumn semester accounted for 18.18% (8/44). Among the 44 outbreaks, 3 educational institutions – Kindergartens 2 and 7 and Primary School 3 – reported 1 case each that tested positive for the enzyme-linked immunosorbent assay (ELISA) antigen at medical institutions but negative for nucleic acid testing conducted by the Xi’an CDC. The baseline characteristics of these outbreaks are presented in Table 3.

DIFFERENT EPIDEMIOLOGICAL CHARACTERISTICS OF NOROVIRUS OUTBREAKS:

The pooled estimates of the incidence of the 5 indicators, including the vomiting rate, the diarrhea rate, the attack rate, the asymptomatic prevalence, and the norovirus detection rate, are presented in Figure 1.

VOMITING AND DIARRHEA RATE DURING NOROVIRUS OUTBREAKS:

Vomiting occurred in 90.9% of cases (95% confidence interval [CI]: 84.6–95.9), and it varied significantly by age groups: 94.9% (95% CI: 89.8–98.6) among kindergarten children, 84.5% (95% CI: 66.7–97.0) among primary school students, and 34.8% (95% CI: 16.5–55.5) among junior high school students (χ²=7.921, p=0.019). Specifically, junior high school students showed significantly lower vomiting rates compared to kindergarten children (Z=−2.442, p=0.015). Diarrhea occurred in 23.3% of cases (95% CI: 15.8–31.5), and the highest incidence was noted among kindergarten children (25.8%, 95% CI: 17.8–34.7), followed by junior high school students (14.4%, 95% CI: 6.8–23.9) and primary school students (10.5%, 95% CI: 0.9–25.7) (χ²=2.382, p=0.304). The vomiting and diarrhea rates in AGE cases are key indicators for identifying potential norovirus outbreaks, and thus help in making an initial decision on whether other potential pathogens should be tested.

ATTACK RATE DURING NOROVIRUS OUTBREAKS:

The overall attack rate was 29.1% (95% CI: 22.9–35.6), with significant variation observed across different age groups: 34.8% (95% CI: 28.9–41.0) among kindergarten children, 14.5% (95% CI: 7.3–23.5) among primary school students, and 2.2% (95% CI: 1.6–2.9) among junior high school students (χ²=11.405, p=0.003). Pairwise comparisons revealed statistically significant differences between groups (all p values <0.017). Based on the attack rate of norovirus outbreaks, the scale of an epidemic (as indicated by the number of cases) can be assessed, thereby facilitating the rational allocation of medical resources.

ASYMPTOMATIC PREVALENCE DURING NOROVIRUS OUTBREAKS:

The overall asymptomatic prevalence was 17.4% (95% CI: 9.7–26.3), and the highest was noted among primary school students (32.8%,95% CI: 4.3–69.0), followed by kindergarten children (14.4%,95% CI: 7.4–22.6) (Z=−0.753, p=0.451). No rectal swab samples were collected from asymptomatic junior high school students because they all declined to participate in sampling. The asymptomatic prevalence of norovirus in outbreaks provides valuable data that can inform policy decisions regarding class suspensions.

NOROVIRUS DETECTION RATE DURING OUTBREAKS:

The overall norovirus detection rate in outbreak cases was 65.6% (95% CI: 57.6–73.2), and the highest detection rate was noted in kindergarten children (68.7%,95% CI: 60.0–76.9), followed by primary school students (58.5%,95% CI: 37.7–78.1) and junior high school students (47.6%, 95% CI: 25.4–70.3) (χ²=3.881, p=0.144). According to the norovirus detection rate, and provided that national guidelines requirements are met (with a minimum of 2 confirmed positive cases), the number of collected samples may be appropriately reduced.

Discussion

Schools and kindergartens are particularly vulnerable to norovirus outbreaks, emphasizing the urgent need for more effective prevention and control measures [7].Therefore, we conducted an analysis of 5 key indicators – vomiting rate, diarrhea rate, attack rate, asymptomatic prevalence, and norovirus detection rate – across 44 norovirus outbreaks to address these challenges. The overall attack rate was 29.1% (95% CI: 22.9–35.6) and the rate was highest among kindergarten children (34.8%, 95% CI: 28.9–41.0). Vomiting was reported in 90.9% of the cases (95% CI: 84.6–95.9), most frequently among kindergarten children (94.9%, 95% CI: 89.8–98.6). Diarrhea occurred in only 23.3% of the cases (95% CI: 15.8–31.5), the highest among kindergarten children (25.8%, 95% CI: 17.8–34.7). Significant differences across age groups were observed for all 3 indicators mentioned above. Norovirus was detected in 65.6% of the cases (95% CI: 57.6–73.2) using PCR-based methods. Additionally, the asymptomatic prevalence during the outbreaks was 17.4% (95% CI: 9.7–26.3). No significant differences in either of these metrics were observed across age groups.

In the early stage of an AGE outbreak, it is critically important to preliminarily determine whether norovirus is the causative pathogen, especially in economically underdeveloped regions, as this can optimize the allocation and conservation of public health resources. The vomiting rate observed in this study was 90.9%, significantly higher than the 50% threshold defined by the Kaplan criteria [9]. Additionally, the overall ratio of vomiting to diarrhea cases was 90.9% versus 23.3%, with 94.9% versus 25.8% for kindergarten children and 84.5% versus 10.5% for primary school students. The findings are notably different from 2 international studies referenced in the national guidelines, which reported vomiting-to-diarrhea ratios of 81.0% versus 52.0% in the study by Götz et al [24], and 95.0% versus 74.0% in the study by Rockx et al [25]. These discrepancies suggest that when vomiting rates are lower and diarrhea rates are higher, the possibility of pathogens other than norovirus, such as Sapovirus, Staphylococcus aureus, and Salmonella [26,27], causing the AGE outbreak should be considered. Prompt collection of appropriate samples and testing for alternative pathogens is critical in such scenarios.

Understanding the attack rate of norovirus outbreaks is essential for evaluating the disease burden, determining the scale of an outbreak, and providing valuable references for the effective allocation of medical resources. The attack rate in this study was 29.1% (95% CI: 22.9–35.6), with significant differences among age groups: 34.8% for kindergarten children, 14.5% for primary school students, and 2.2% for junior high school students. This gradient aligns with findings from a meta-analysis conducted in China, which reported attack rates of 9.13% for kindergarten children, 6.07% for primary school students, and 4.19% for junior high school students [10]. However, the higher attack rates observed in our study may reflect differences in the criteria used to define exposure of individuals in educational settings. These findings highlight the importance of tailoring interventions based on group dynamics and exposure levels to effectively control norovirus outbreaks.

Asymptomatic infected individuals play a critical role in norovirus transmission, as they can trigger new outbreaks or prolong ongoing ones [13–15]. Studies have reported that the duration and amount of norovirus shedding in asymptomatic infected individuals were similar to or slightly lower than those in symptomatic patients [18,28–30]. This study found an overall asymptomatic prevalence of 17.4% (95% CI: 9.7–26.3), which is close to the global and regional estimates. For example, Rui Qi et al reported an asymptomatic prevalence of 18.0% globally [21], while Wang Jun et al reported 16.6% in Asia and 17.6% in China [17,22]. There was no statistically significant difference in the asymptomatic prevalence between kindergarten children (14.4%; 95% CI: 7.4–22.6) and primary school students (32.8%; 95% CI: 4.3–69.0). This finding is consistent with the study by Wang et al [17]. Both studies included a very limited number (≤9) of norovirus outbreaks among school-age children. Future research should include more studies on school-age children to obtain more precise estimates of asymptomatic prevalence. These results are instrumental in guiding decisions about class closures during outbreaks, ensuring effective containment without unnecessarily disrupting educational activities.

Assessing the detection rate of norovirus in outbreaks is therefore essential for optimizing resource allocation and improving response efficiency. The norovirus detection rate in this study was 65.6% (95% CI: 57.6–73.2), consistent with the detection rate reported in Anhui Province, China (60.10%) [31]. The national guidelines recommend collecting 20 samples for each outbreak to ensure the identification of at least 2 norovirus-positive specimens, as 2 positive cases are sufficient for outbreak confirmation [8]. However, this study suggests that collecting 5 rectal swab samples from representative cases is generally adequate for confirming a norovirus outbreak. Reducing the sample size from 20 to 5 could significantly decrease testing costs. Currently, the cost of nucleic acid testing for norovirus is approximately 150–200 Chinese Yuan per sample. By reducing sample sizes, financial burdens associated with norovirus detection could be alleviated, particularly in regions like Xi’an, which report hundreds of norovirus outbreaks annually. Scaling this approach nationwide could lead to significant cost savings and more efficient allocation of healthcare resources. If our recommendation is adopted, detection sensitivity should be carefully evaluated. It is recommended to collect stool or anal swab samples during the acute phase (within 2–3 days of symptom onset) and to use real-time RT-PCR for laboratory testing, following national guidelines [8]. All procedures for sample collection, storage, transportation, and testing must strictly comply with national guidelines [8].

Despite its contributions, this study has some limitations. First, the random-effects single-rate meta-analysis employed in this study, similar to those used in previous studies [17,20–23], is limited in its ability to control for confounding bias due to the lack of a control group and demonstrates considerable heterogeneity along with potential publication bias. Although the robustness and reliability of our findings have been assessed in Supplementary File 1, the conclusions derived from this analysis should be interpreted with caution when generalized beyond the current context. Second, norovirus outbreaks were reported less frequently in primary and junior high schools compared to kindergartens, potentially affecting the accuracy of subgroup-specific results. Future studies should systematically collect data from articles on norovirus outbreaks among school-age children across China, thereby achieving a more accurate outcome. Third, this study was conducted during the COVID-19 pandemic, a period marked by stringent non-pharmaceutical interventions such as social distancing, improved ventilation, and enhanced disinfection [32]. These measures may have reduced the asymptomatic prevalence of the findings. Replicating this study under normal social conditions would yield valuable comparative insights. Fourth, some asymptomatic selected individuals refused to be sampled, necessitating the discontinuation of sample collection as other exposed individuals had already left the school following simple random sampling. This may have affected the reliability of the results. Fifth, although all norovirus outbreaks with complete data in Xi’an were included, the sample size was relatively small (only 44 outbreaks). Future studies should include more outbreaks from different regions for more accurate estimates. Sixth, this study did not account for illnesses or asymptomatic infections among family members of these cases, as none were reported to have experienced AGE during the outbreaks.

Conclusions

Our findings highlight the impact of norovirus on student populations in schools and kindergartens and offer the following reference recommendations for the prevention and control of norovirus outbreaks in these settings. First, when the vomiting rate is below 90.9% and the diarrhea rate exceeds 23.3%, it is advisable to consider concurrent testing for alternative common enteric pathogens, such as Sapovirus, Staphylococcus aureus and Salmonella. Second, based on the number of exposed individuals and attack rates for different populations, the number of outbreak cases can be estimated to inform the allocation of medical resources. Third, given the high prevalence of asymptomatic infections, the similar viral shedding observed in asymptomatic infected individuals and symptomatic cases, and the extremely low infective dose (as few as 18–2800 viral particles can cause disease), it is necessary to close all classes with cases and monitor exposed individuals in classes without unaffected cases at home, rather than isolating only symptomatic cases. Fourth, for each acute gastroenteritis outbreak, 5 representative rectal swab samples could be collected to determine whether norovirus is the causative agent.

Data availability statement

The original data supporting our findings were obtained from the Xi’an CDC in China, though access to these data is restricted. The data were used under a specific license for the purpose of this study and thus are not publicly accessible. However, they may be made available upon reasonable request directed to the corresponding authors, subject to approval from Xi’an CDC. However, all extracted data related to the study were included in the article or uploaded as supplementary information. The data supporting the conclusions of this article are available listed in Table 3 and Data set 1.