Download:
-
Numerous acute gastroenteritis (AGE) outbreaks have been attributed to Campylobacter infection in recent years in China (1-5). The main species responsible for these outbreaks is Campylobacter jejuni (3-4) except one outbreak caused by Campylobacter coli (C. coli) in Shunyi reported in 2018 (2). Human sapovirus (HuSaV) is a pathogen that causes sporadic cases and outbreaks of acute gastroenteritis worldwide. An acute gastroenteritis outbreak involving 996 patients occurred in Beijing in July 2021. In this study, we present a comprehensive analysis of the epidemic and laboratory findings related to this outbreak.
-
On July 15, 2021, a sentinel hospital in Beijing identified 13 patients with AGE in the same school. Epidemiological investigation revealed a total of 996 patients in this outbreak, including 958 students (221 male) and 38 staff members. The attack rate among students and staff members were 67.7% (958/1,413) and 21.7% (38/175), respectively. The most common clinical symptoms, ranked from highest to lowest, were abdominal pain (75.6%, 753/996), diarrhea (73.8%, 735/996), nausea (53.3%, 531/996), vomiting (23.0%, 229/996), and fever ≥37.3 ℃ (13.7%, 136/996).
A total of 828 valid survey questionnaires were accomplished (341 AGE cases and 487 controls). From the investigation, one direct drinking water (DDW) system supplied by a secondary water supply system (WSS-S) was identified in the school building. This DDW system offered two types of water: unboiled direct drinking water (UDDW) and boiled. The water source for WSS-S is a groundwater source well (WSW) located within the commercial campus and one neighbor hotel in the same campus also has its own secondary water supply system (WSS-H), which also utilizes the WSW as its water source. Both WSS-H and WSS-S have separate facilities including a water storage tank, water pump, and pipeline. The WSS-H conducted disinfection using chlorine dioxide and the water storage tank and pump were located inside of the hotel. The water quality from WSS-H was supervised effectively. In contrast, the water storage tank and water pump of WSS-S are not located inside the school and lack effective management. Additionally, the sewage well is uncovered and lacks protective facilities during rainfall (Figure 1). Based on the results of the case-control studies, exposure to UDDW was found to increase the risk of illness [risk ratio=3.895, 95% confidence interval (CI): 2.90–5.22; Table 1].
Figure 1.Schematic diagram of the commercial campus and water supply system.
Exposure to uncooked direct drinking water Case number Control number RR value 95% Cl Yes 210 142 3.895 2.906−5.220 No 131 345 Total 341 487 − − Note: “−” means not applicable.
Abbreviation: RR=risk ratio; CI=confidence interval.Table 1. Case-control results for potentially suspicious risk factors.
The first identified patient on July 10, 2021 was designated as D0. Subsequent days were labeled as D1 to D17. Two peaks of patient onset were observed at D4–D5 and D11–D12. On D1–D2, after a rainstorm occurred, some students reported that the water in the school appeared turbid and had an unpleasant odor following the rainstorm. After the DDW was replaced with commercially bottled water, the number of patients decreased significantly. All patients were quarantined after the second peak, and the number of daily patients continued to decline until no new patients were reported after D17 (Figure 2). No cases were identified in the neighbor hotel during this period.
Figure 2.Time distribution of patient onset.
Note: The onset time for patients corresponding to the samples collected on day 6 (D6) was between day 3 and day 6 (D3D6), while the onset time for cases corresponding to the samples collected on day 12 to day 14 (D12D14) was between day 3 and day 12 (D3D12). The number of new patients which onset at D4, D5, D6 and D11 all exceed the maximum value of 50 on the Y-axis (green bar).Stool or anal swabs were collected from a total of 83 patients and 5 canteen staff. In addition, 18 water samples, each 2L, 12 food samples, and 17 environmental smear specimens from the canteen were collected and stored in sterile containers. The laboratory investigation was conducted by local CDC.
Real-time polymerase chain reaction (PCR) was conducted using commercial kits to detect common enteric pathogens. Among the patient samples, the positive rates for C. coli and HuSaV GI, as determined by real-time PCR, were 59.04% (49/83) and 39.83% (39/83), respectively. The mixed positive rate for C. coli and HuSaV GI was 32.53% (27/83). Two out of 17 environmental smear samples, specifically those collected from school toilet pans, tested positive for C. coli. No other enteric pathogens were detected in any of the collected samples. The distribution of pathogen detection results in patient samples collected on various dates is provided in Table 2.
Sample type Sample collection date Time interval for the first onset of clinical symptoms of patients Sample name Sample number Real-time PCR positive rate for C. coli C. coli culture positive rate HuSaV GI positive number C. coli and HuSaV GI positive number Number of unqualified samples for aerobic plate counting Number of unqualified samples for coliform detection Patients D6 D3−D6 Anal swab 19 11 1 0 0 − − D12 D3−D11 Anal swab 15 10 4 7 6 − − D13 D4−D12 Stool 24 11 5 17 8 − − D14 D4−D12 Stool 25 17 6 15 13 − − Canteen workers D6 − Anal swab 5 0 0 0 0 − − Retained food D6 − Food 12 0 0 0 0 − − Environment smear D6 − Smear 17 2 0 0 0 − − Water D6 − WSW 1 − − − − 0 0 − WSS-H 1 − − − − 0 0 − WSS-S 5 − − − − 5 5 D12−D14 − WSS-H 3 − − − − 0 0 − WSS-S 8 − − − − 8 8 Note: “−” means not applicable.
Abbreviation: HuSaV=human sapovirus; WSW=water source well; WSS-H=secondary water supply system for hotel; WSS-S=secondary water supply system for school.Table 2. Detection results in samples collected on different dates.
Campylobacter isolation and antibiotic resistance were performed using an isolation kit and agar dilution method as previously described (2). Seventeen C. coli isolates were obtained from 17 patients. The antibiotic resistance profile of all 17 strains showed co-resistance to nalidixic acid, ciprofloxacin, streptomycin, and tetracycline.
Aerobic plate counting and coliform detection were conducted on a total of 18 water samples. Five samples collected on D6 and 8 samples collected on D12–D14 from the WSS-S did not meet the criteria, with aerobic plate counts exceeding 100 CFU/mL and coliforms being detected. Water samples from WSS-H and WSW satisfied the criteria for both aerobic plate counting and coliform detection (Table 2).
The draft genomes of 17 C. coli isolates from patients were sequenced on the Illumina HiSeq 2500 platform at the Tianjin Genomics Institute. High-quality reads were assembled using the SPAdes genome assembler software (version v3.11.0) (
https://github.com/ablab/spades ). The average nucleotide identity (ANI) and in silico DNA-DNA hybridization (isDDH) values of these 17 genomes and one reference (CP047137), were analyzed using pyani v0.2 software (https://github.com/widdowquinn/pyani ) employing the ANIm and GGDC modules (http://ggdc.dsmz.de/ggdc.ph ) with Formula 2, respectively. Both ANI and isDDH values among the genomes exceeded 96% and 70%, respectively, which meet the taxonomic criteria for classification within the same species (Figure 3A–B and Table 3).
Figure 3.Heatmap of average nucleotide identity (A) and in silico DNA-DNA hybridization (B) values of 17 outbreak C. coli isolate genomes and reference genome CP047137.
CP
046317BJTZ
2021032BJTZ
2021036BJTZ
2021040BJTZ
2021026BJTZ
2021027BJTZ
2021035BJTZ
2021030BJTZ
2021023BJTZ
2021025BJTZ
2021031BJTZ
2021029BJTZ
2021024BJTZ
2021033BJTZ
2021041BJTZ
2021034BJTZ
2021042BJTZ
2021028CP046317 88.50 88.50 88.40 88.40 88.50 88.40 88.50 88.50 88.50 88.60 88.60 88.40 88.50 88.50 88.50 88.50 88.50 BJTZ2021032 98.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021036 98.98 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021040 98.98 00.00 00.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021026 98.99 99.99 99.99 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021027 98.99 100.00 100.00 100.00 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021035 98.98 99.99 99.99 99.99 100.00 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021030 98.99 100.00 100.00 100.00 99.99 100.00 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021023 98.98 100.00 100.00 100.00 99.99 100.00 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021025 98.99 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021031 99.00 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021029 98.99 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021024 98.99 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 BJTZ2021033 98.99 100.00 100.00 100.00 99.99 100.00 99.99 100.00 100.00 99.99 99.99 99.99 99.99 100.00 100.00 100.00 100.00 BJTZ2021041 98.98 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 99.99 100.00 100.00 100.00 BJTZ2021034 98.98 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 99.99 100.00 100.00 100.00 BJTZ2021042 98.99 99.99 99.99 99.99 100.00 99.99 100.00 99.99 99.99 100.00 100.00 100.00 100.00 99.99 100.00 100.00 100.00 BJTZ2021028 98.98 100.00 100.00 100.00 99.99 100.00 99.99 100.00 100.00 99.99 99.99 99.99 99.99 100.00 99.99 99.99 99.99 Note: Numbers in bold represent average nucleotide identity values, and numbers in plain text represent the DNA-DNA hybridization values. Table 3. Average nucleotide identity and in silico DNA-DNA hybridization results for C. coli (%).
Multilocus sequence typing (MLST) for each isolate was determined using the Genome Comparator tool on the PubMLST website (
http://pubmlst.org/campylobacter ). All isolates were identified as sequence type (ST) 1068 and clonal complex (CC) 828. An ad hoc whole-genome MLST (wgMLST) analysis was performed using fast-GeP, with strain RM5611 from the PubMLST database as the reference. This analysis identified 1,697 gene loci in the reference strain and 1,595 core gene loci across all strains, assessing the sequence characteristics of each gene locus. A phylogenetic tree was constructed using SplitsTree 4 and the neighbor-net method (Figure 4) based on wgMLST analysis. Subsequently, core genome SNPs (cgSNPs) were identified by mapping sequencing reads to the draft genome of isolate RM5611 using Snippy version 4.3.5 (https://github.com/tseemann/snippy ). A total of 6,203 cgSNPs were identified, using RM5611 as the reference genome. The cgSNP analysis (Figure 5A) shows high clonality among the 17 strains, with cgSNP differences between them presented in Figure 5B.
Figure 4.Neighbor-net phylogeny for alleles of wgMLST loci of 18 C. coli isolates.
Note: All 18 isolates are classified as ST-1068. The red points represent the 17 isolates associated with the outbreak, while the green points represent sporadic isolates from the pubMLST database (reference). The outbreak cluster is indicated by blue circles.
Figure 5.Phylogenetic tree and difference matrix of 18 isolates based on SNPs. (A) A phylogenetic tree was constructed using SNP data from 18 C. coli isolates. (B) In addition, a difference matrix was generated based on the SNP data for the 18 C. coli isolates.
Note: SNPs were identified by mapping sequencing reads to the draft genome of isolate RM5611 using Snippy version 4.3.5. A total of 6,203 cgSNPs were included, with strain RM5611 serving as the reference genome. -
Case-control studies showed that exposure to UDDW increased the risk of disease. All 17 C. coli isolated from the patients were highly clonal. Water samples from the WSW and treated WSS-H yielded satisfactory results for aerobic plate counting and coliform detection. Exposure to WSS-S was the primary risk factor for the outbreak. C. coli and HuSaV GI were finally identified as the major pathogens in this outbreak, and transmission primarily occurred via UDDW from the contaminated secondary water supply system. This is the first reported outbreak in China where a mixed infection of C. coli and HuSaV caused AGE.
Outbreaks caused by C. coli have been linked to contaminated surface water exposure (6), male sexual activity (7), and transmission through food (2). This outbreak occurred following a rainstorm, where the “rinsing water” from WSS-S appeared turbid and had an odor. Water samples collected from WSS-S on D6 and D12−D14 showed unqualified results for aerobic plate counting and coliform detection, which indicated the contamination. The nearby garbage station and sewage well were likely sources of contamination. Regrettably, during the initial stages of this outbreak, Campylobacter was not given attention. Even though 19 anal swab samples were collected on Day 6, Campylobacter detection was not performed until Day 11, potentially explaining why 11 out of the 19 samples tested positive for Campylobacter through real-time PCR, but only one sample was positive through bacterial isolation.
The C. coli strains isolated in this outbreak were ST1068. There have been no previous reports of ST1068 in China. However, sporadic reports of ST1068 have been observed in Japan, the USA, the UK, Italy, and Canada (8-11). In Japan, ST1068 is the second most dominant type of C. coli isolated in the clinic but no reports of outbreaks caused by ST1068 worldwide. ST1068 is mainly found in livestock, particularly cows and pigs, and it is closely associated with diarrhea in patients. Future research should pay attention to ST1068.
Previous research has shown that HuSaV outbreaks typically last from 10 hours to 6 days with an attack rate of 16%–52% (12). These characteristics align with the second peak observed in this outbreak, as well as the high positive rate of HuSaV GI in samples collected on D12−D14. Notably, no positive results were found in 19 samples collected at D6, while 39 positive results were obtained from 64 samples collected at D12−D14. This suggests that HuSaV GI infection is likely to occur in the later stages of the outbreak, which contributed to the second peak (D11−D13).
During the second peak of this outbreak, the supply of DDW sourced from WSS-S was discontinued, but the use of “rinsing water” sourced from WSS-S continued. Waterborne infections caused by HuSaV and C. coli may occur through exposure to “rinsing water,” and HuSaV may also be transmitted from human to human. This could explain the second peak observed in the epidemic. On D11, all patients were transferred to quarantine, effectively blocking both waterborne and human-to-human transmission. The decrease in the number of new patients following this measure indicates its effectiveness.
The results of aerobic plate counting and coliform detection indicated that the water samples from the WSS-S were poor of quality. But neither C. coli nor HuSaV were identified in the water samples. This was the limitation in this study. Waterborne outbreaks often occur on a large scale and involve complex processes of pathogens examination and tracing. Therefore, effective methods for pathogen identification were very critical.
-
Staff of the participating hospitals and colleagues from Tongzhou CDC.
HTML
| Citation: |
|