Skip to main content Accessibility help
×
Home
Hostname: page-component-59b7f5684b-8dvf2 Total loading time: 3.512 Render date: 2022-10-03T12:53:42.277Z Has data issue: true Feature Flags: { "shouldUseShareProductTool": true, "shouldUseHypothesis": true, "isUnsiloEnabled": true, "useRatesEcommerce": false, "displayNetworkTab": true, "displayNetworkMapGraph": false, "useSa": true } hasContentIssue true

Epidemiological investigations identified an outbreak of Shiga toxin-producing Escherichia coli serotype O26:H11 associated with pre-packed sandwiches

Published online by Cambridge University Press:  10 August 2021

Saira Butt
Affiliation:
National Infection Service, Public Health England, London, UK
Lesley Allison
Affiliation:
Scottish E. coli O157/STEC Reference Laboratory, Edinburgh, UK
Bhavita Vishram
Affiliation:
National Infection Service, Public Health England, London, UK
David R. Greig
Affiliation:
National Infection Service, Public Health England, London, UK
Heather Aird
Affiliation:
Food, Water & Environmental Microbiology Laboratory, Public Health England, York, UK
Eisin McDonald
Affiliation:
Public Health Scotland, Meridian Court, 5 Cadogan street, GlasgowG2 6QQ, UK
Genna Drennan
Affiliation:
Public Health Scotland, Meridian Court, 5 Cadogan street, GlasgowG2 6QQ, UK
Claire Jenkins*
Affiliation:
National Infection Service, Public Health England, London, UK
Lisa Byrne
Affiliation:
National Infection Service, Public Health England, London, UK
Kirsty Licence
Affiliation:
Public Health Scotland, Meridian Court, 5 Cadogan street, GlasgowG2 6QQ, UK
Alison Smith-Palmer
Affiliation:
Public Health Scotland, Meridian Court, 5 Cadogan street, GlasgowG2 6QQ, UK
*
Author for correspondence: Claire Jenkins, E-mail: claire.jenkins@phe.gov.uk
Rights & Permissions[Opens in a new window]

Abstract

In October 2019, public health surveillance systems in Scotland identified an increase in the number of reported infections of Shiga toxin-producing Escherichia coli (STEC) O26:H11 involving bloody diarrhoea. Ultimately, across the United Kingdom (UK) 32 cases of STEC O26:H11 stx1a were identified, with the median age of 27 years and 64% were male; six cases were hospitalised. Among food exposures there was an association with consuming pre-packed sandwiches purchased at outlets belonging to a national food chain franchise (food outlet A) [odds ratio (OR) = 183.89, P < 0.001]. The common ingredient identified as a component of the majority of the sandwiches sold at food outlet A was a mixed salad of Apollo and Iceberg lettuce and spinach leaves. Microbiological testing of food and environmental samples were negative for STEC O26:H11, although STEC O36:H19 was isolated from a mixed salad sample taken from premises owned by food outlet A. Contamination of fresh produce is often due to a transient event and detection of the aetiological agent in food that has a short-shelf life is challenging. Robust, statistically significant epidemiological analysis should be sufficient evidence to direct timely and targeted on-farm investigations. A shift in focus from testing the microbiological quality of the produce to investigating the processes and practices through the supply chain and sampling the farm environment is recommended.

Type
Original Paper
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2021. Published by Cambridge University Press

Key Findings

  • Report on the investigation of the first UK-wide foodborne outbreak of Shiga toxin-producing Escherichia coli (STEC) O26:H11

  • Epidemiological investigations identified an association with consuming food purchased at outlets belonging to a national food chain franchise and the consumption of pre-packed sandwiches

  • Highlights the challenges of detecting STEC in produce that has a short-shelf life

  • Recommendation 1: Robust, statistically significant epidemiological analysis be regarded as sufficient evidence to direct timely and targeted on-farm investigations.

  • Recommendation 2: Shift in focus from testing the microbiological quality of the produce to investigating the processes and practices through the supply chain and sampling the farm environment.

Introduction

Shiga toxin-producing Escherichia coli (STEC) belong to a pathogenic group of zoonotic E. coli that cause gastrointestinal disease in humans due to their ability to produce Shiga toxin (Stx) [Reference Croxen1]. There are two types of Stx, Stx1 and Stx2 and 10 subtypes Stx1a, Stx1c, Stx1d and Stx2a-Stx2 g [2]. Shiga toxin targets a specific subset of cells that express the receptor GB3, including podocytes, microvascular endothelial cells in the kidney, platelets, germinal centre B lymphocytes, erythrocytes and neurons [Reference Melton-Celsa3Reference Joseph5]. Certain strains of STEC have the potential to cause haemolytic uraemic syndrome (HUS), a life-threatening condition characterised by renal failure, sometimes with cardiac and/or neurological complications [2Reference Joseph5]. The majority of STEC isolated from patients with severe symptoms also have a gene designated eae that encodes the protein intimin and is commonly used as a molecular marker for the locus of enterocyte effacement (LEE) pathogenicity island. Intimin is one of the large number of proteins encoded in the LEE and is required for the formation of attaching and effacing lesions on host intestinal cells [Reference Furniss and Clements6, Reference Stevens and Frankel7]. STEC infection damages the lining of the gut reducing its capacity to reabsorb fluids resulting in diarrhoea, which may contain blood.

The two most commonly detected STEC serotypes in the UK are STEC O157:H7 and STEC O26:H11 ([Reference Gentle14], https://www.gov.uk/government/publications/escherichia-coli-e-coli-o157-annual-totals/shiga-toxin-producing-escherichia-coli-stec-data-2018), https://www.foodstandards.gov.scot/publications-and-research/publications/whole-genome-sequence-typing-and-analysis-of-non-o157-stec). STEC O157:H7 and STEC O26:H11 are known to colonise the gut of ruminants, such as cattle and sheep and other animals including birds, may act as transient vectors [Reference Thomas8Reference Navarro-Gonzalez10]. Transmission to humans occurs following the consumption of contaminated food or water and direct contact with animals or their environment. In household and institutional settings, secondary, person-to-person transmission of STEC has been described [Reference Byrne11].

The outbreaks of STEC-HUS in the UK in the 1980s were caused by STEC O157:H7 and in response, laboratory methods focused on the use of cefixime-tellurite sorbitol MacConkey (CT-SMAC) agar [Reference Adams12]. CT-SMAC is selective for the growth of this specific serotype, thus compromising surveillance of the other STEC, including STEC O26:H11. In contrast, commercial and in-house polymerase chain reaction (PCR) assays targeting the presence of stx have the potential to detect STEC of all serotypes [Reference Byrne13]. In England and Wales, surveillance for STEC O26:H11 has been continually improving since 2013, in parallel with the increasing number of frontline hospital laboratories implementing PCR ([Reference Byrne13, Reference Gentle14] https://www.gov.uk/government/publications/escherichia-coli-e-coli-o157-annual-totals/shiga-toxin-producing-escherichia-coli-stec-data-2018). However, although all patients who present to primary healthcare and submit a faecal specimen are tested regardless of symptom severity, only the regions of the country where the local hospital laboratory has implemented PCR have the potential to detect STEC O26:H11. In Scotland, the burden of STEC O26:H11 for patients with severe clinical outcomes is well-established as specimens from cases with bloody diarrhoea and HUS have been forwarded to the Scottish E. coli O157/STEC Reference Laboratory (SERL) for STEC testing since 2002 (https://www.foodstandards.gov.scot/publications-and-research/publications/whole-genome-sequence-typing-and-analysis-of-non-o157-stec).

In October 2019, routine microbiological surveillance at SERL identified an unexpected increase in the number of faecal specimens that tested positive by PCR for stx1 but were negative for stx2 and rfbE O157 (a gene encoding a lipopolysaccharide synthesis enzyme unique to E. coli O157), and a multi-agency incident management team (IMT) was convened. This report describes the outbreak investigation, highlights the microbiological and epidemiological challenges encountered and makes recommendations for future practice.

Methods

Microbiology investigations

In the UK, faecal specimens from hospitalised or community cases with symptoms of gastrointestinal disease are cultured in local hospital microbiology laboratories for identification of Salmonella, Campylobacter, Shigella spp. and STEC O157:H7. In England, a sub-set of laboratories, approximately 20%, use commercial PCR assays for the detection of gastrointestinal pathogens, including STEC other than serotype O157:H7 (non-O157 STEC) ([Reference Gentle14], Public Health England (PHE) in-house data). Faecal specimens from patients where there is a clinical suspicion of HUS and/or those testing positive for STEC by PCR and culture-negative for STEC O157:H7 on cefixime-tellurite sorbitol MacConkey agar are submitted to the Gastrointestinal Bacteria Reference Unit at PHE for confirmation by PCR and culture [Reference Jenkins15, Reference Jenkins16]. In Scotland, all faecal specimens from patients with severe gastrointestinal symptoms, specifically bloody diarrhoea and HUS, are submitted to the SERL for PCR and culture. During the outbreak, faecal specimens testing positive for stx1 but negative for rfbE O157 at both SERL and PHE were tested using a PCR for the detection of rfbE O26 (http://old.iss.it/binary/vtec/cont/EU_RL_VTEC_Method_02_Rev_0.pdf), while whole genome sequencing (WGS) results were pending. All strains of STEC isolated from faecal specimens were sequenced, and serotype, stx subtype profile and single nucleotide polymorphism (SNP) type were derived from the genome, as described previously [Reference Chattaway17Reference Dallman20].

Data availability statement

FASTQ reads from all sequences in this study can be found at the PHE Pathogens BioProject at the National Center for Biotechnology Information (accession number: PRJNA315192).

Case definitions

Confirmed: A case of STEC O26:H11 stx1/eae with a sequence that falls within five SNPs of the outbreak SNP type.

Previous analysis of the relatedness of isolates of STEC has shown that isolates from cases epidemiologically linked to the same outbreak fall within the same five SNP single linkage cluster [Reference Dallman21, Reference Jenkins, Dallman and Grant22].

Epidemiological investigations

Prospective and retrospective case ascertainment was undertaken by reviewing stx profiles from faecal specimens where STEC had been isolated but that were pending WGS, and by reviewing all WGS data held in the PHE and SERL databases.

Public Health Agencies in England, Scotland and Wales operate national enhanced surveillance systems for STEC [Reference Byrne11]. Every laboratory confirmed case of STEC O157:H7 is asked to provide a detailed account of their food history, contact with animals and environmental exposures for the 7 days prior to onset of illness using a standardised enhanced surveillance questionnaire (ESQ). In England and Wales, public health follow-up of cases of non-O157 STEC focuses on those that are infected with STEC that are positive by PCR for stx2, because of the association between stx2a and severe clinical outcomes. Cases that are infected with STEC that are positive for stx1 only strains are not routinely administered an ESQ (https://www.gov.uk/government/publications/shiga-toxin-producing-escherichia-coli-public-health-management). In Scotland, ESQs are administered to all SERL confirmed cases of non-O157 STEC regardless of the stx profile. Following an initial review of the ESQ data, outbreak cases were re-interviewed using a trawling questionnaire to collect a more detailed food history.

The UK posted the outbreak on the European Communicable Disease Centre's Epidemic Intelligence Information System sharing the WGS accession numbers to ascertain whether related cases had been seen elsewhere.

Case−case analysis of exposures

Data were extracted from the national enhanced surveillance database maintained by Health Protection Scotland (HPS) for the confirmed outbreak cases residing in Scotland, and a bespoke dataset was created by combining standardised data collected from ESQs and the trawling questionnaires. The null hypothesis for testing was that there was no association between any exposure to food from any specific food outlet and being infected with the outbreak strain.

Controls were selected from among sporadic STEC cases reported in 2018 in the HPS national database. Controls were excluded if travel outside of the UK was reported in the 14 days before onset, if an ESQ was not completed, or if they were younger or older than the age range of the cases associated with the outbreak (between 15 and 64 years). Food exposures were re-coded as binary responses. Binary variables were created for each food outlet, coded as 1 if premise was mentioned as the source of any food items. Odds ratios (ORs) were calculated for each food outlet exposure comparing the odds of outbreak cases reporting the exposure with the odds of controls cases reporting the same exposure. P values were adjusted for multiple testing.

Multivariable logistic regressions were conducted for each exposure, adjusting for age and sex. This analysis was also repeated using all the 2018 STEC cases regardless of age and using logistic regression to adjust for age and sex.

Case−control analysis of exposures

A further analysis was undertaken in order to include exposures reported by cases in England, and to account for overmatching in the case−case analysis by using an existing database of healthy controls. Data were extracted from the national enhanced surveillance system for the confirmed outbreak cases residing in England, and exposure data were provided by HPS for the confirmed cases residing in Scotland. The null hypothesis for testing was: There is no association between any food exposures and being infected with the outbreak strain.

Controls were selected from the national control database maintained by PHE Field Services North West, which is frequency-matched by age with all confirmed STEC cases during the same time period. Controls were excluded if travel outside of the UK was reported or if an ESQ was not completed, if gastrointestinal illness in the month prior was reported, or if they were not resident in England. Food exposures were re-coded as binary responses. ORs were calculated for each food exposure reported by at least 25% of cases, comparing the odds of outbreak cases reporting the exposure with the odds of controls cases reporting the same exposure. The Fisher's exact test was conducted to produce P values, to account for the small numbers of cases (<5 exposed). A multivariable logistic regression was then created to include any food item where the results of the univariate analysis showed P < 0.2 and OR >1.

Food chain investigations

Food, water and environmental samples were collected by local environmental health officers (EHOs) from two food processing sites belonging to a national food chain franchise (food outlet A), and one site belonging to the food business operator (FBO) A, and transported in accordance with the Food Standards Agency Food Law Code of Practice (https://www.food.gov.uk/enforcement/codes-of-practice/food-law-code-of-practice-2015) to Food, Water and Environmental Microbiology Laboratories in Edinburgh and York in cold boxes at a temperature of between 0 and 8 °C and tested within 24 h of collection. Testing of the food samples followed PHE Standard Method F17 based on BS EN ISO 16654:2001 http://img.21food.cn/img/biaozhun/20100729/181/11294219.pdf, as previously described. STEC isolated from food samples were submitted to SERL for confirmation and typing. Growing and/or processing procedures for suppliers and wholesale distributors identified in the supply chain investigation were reviewed.

Results

Analysis of microbiological data

During the week commencing 14th October 2019, SERL detected a higher than average number (n = 11) of faecal specimens that tested positive for stx1 compared to any other week in 2019 (median = 1.2 samples/week, range 0–6). STEC O26:H11 was subsequently cultured from 10/11 of these specimens. During the same week, PHE detected 25 faecal specimens that tested positive by PCR for stx1 which was higher than numbers in previous weeks in 2019 (median = 10 samples/week, range 3–22). STEC O26:H11 was subsequently cultured from 11/25 of these specimens. All isolates were sequenced. Sequencing confirmed the serotype of 21 isolates (SERL = 10; PHE = 11) as STEC O26:H11, stx1a/eae, and all 21 isolates fell within a five SNP single linkage cluster (Fig. 1). Ultimately, between 1st October and 30th November, 32 isolates of STEC O26:H11 fell within the same five SNP single linkage cluster were identified.

Fig. 1. Phylogenetic tree showing the relationship between isolates linked to the outbreak from cases resident in England (n = 16), Scotland (n = 14) and Wales (n = 2) that fell within the same five SNP single linkage cluster highlighted in the blue box. The samples outside the blue box were included to provide context and represent all the isolates in the Public Health England (PHE) archive (2015–2019) that fell within a 25 SNP single linkage cluster of the outbreak cluster. Short read archive accession numbers (SRRs) are provided for the sequences of the isolates sequenced at PHE.

Analysis of epidemiological data

The 32 confirmed cases were distributed across the UK, with 14 cases in Scotland, 16 in England and two in Wales (Fig. 2). Onset dates ranged from 2nd October 2019 to 18th November 2019 (Fig. 3). Twenty-one cases were male (64%) and ages ranged from 3 to 77 years of age, with a median of 27 years (Fig. 4). Clinical outcome data were available for 26 cases, of which 22 (85%) cases developed bloody diarrhoea and six (23%) were hospitalised (Table 1). No cases of HUS or deaths were reported. None of the cases reported travel abroad in the incubation period.

Fig. 2. Geographical distribution of the confirmed cases (n = 32).

Fig. 3. Epidemic curve of confirmed cases by onset date, or sample date where onset date was not available (n = 32). Both cases positioned in mid-November were based on sample dates; one case reported being symptomatic since October and no information was available for the second case.

Fig. 4. Age−sex distribution of the confirmed cases (n = 32).

Table 1. Distribution of clinical symptoms amongst confirmed cases with a completed questionnaire

Exposure data were available for 28 of the 32 cases, of which 26/28 (92.9%) cases reported eating out at various food outlets within 7 days of onset of symptoms. Of these, 17/26 (65.4%) cases reported purchasing food from food outlet A, 3/26 (11.5%) cases reported purchasing sandwiches from food outlet B, and 3/26 (11.5%) cases reported eating out at food outlet C. The remaining three cases ate at three different food outlets.

Case−case analysis of exposures

At the time of the case−case analysis, 14 cases had been confirmed in Scotland, and exposure information was available for 13 cases. In addition to the cases linked to the outbreak, a total of 221 records were extracted from the HPS database, of which 59 reported a history of overseas travel in the 14 days prior to onset and were excluded from the analysis leaving 162 cases not linked to this or any outbreak.

In univariable analysis, among food outlet exposures there was evidence of an association between being infected with the outbreak strain and consuming food purchased at food outlet A (OR = 183.89, P < 0.001), any bakery (OR = 17.6, P = 0.019) and food outlet B (OR = 13.20, P = 0.006) (Table 2). In multivariable logistic regressions for each food item adjusted for age and sex, association with the same exposures were maintained, with the magnitude of association increasing for food outlet A (OR = 197.05, P < 0.001), decreasing for any bakery (OR = 15.73, P = 0.030), and remaining stable for food outlet B (OR = 13.58, P = 0.007) (Table 3).

Table 2. In univariable analysis of the association between being infected with the outbreak strain and consuming food purchased at different food outlets

Table 3. Generalised linear model estimates based upon logistic regression adjusted for age and sex and significance level adjusted for multiple testing

Case−control analysis of exposures

Of 27 confirmed cases at the time of analysis (21 November 2019), questionnaires were available for 24 cases (89%), at varying levels of completion. Three hundred ninety-three controls were included in the analysis, who completed questionnaires from February to October 2019. Fifty-four per cent of controls were female, and the age range was less than 1 year to 71 years with the median age being 26 years. In univariable analysis, among food outlet exposures, there was evidence of an association between pre-packaged sandwiches and a higher OR of being a case (OR = 11.01, P < 0.001). The analysis also showed evidence for eating outside of the home (OR = 5.01, P = 0.020), handling raw beef (OR = 2.84, P = 0.019) and handling or consuming processed meat (OR = 2.48, P = 0.034) (Table 4). In the final multivariable model for food items, adjusted for adulthood and sex, nine food items were retained, of which three were associated with increased risk of the outbreak infection. As all cases were adults, controls under 15 years of age were dropped from the analysis. The most notable exposures were pre-packaged sandwiches, processed meats and raw beef (Table 5).

Table 4. Univariate case−control analysis of food exposures

Table 5. Results of multivariable logistic regression for food exposures (n = 225)

Food chain investigations

The ingredients list of foods consumed by the cases at food outlet A, food outlet B and food outlet C, showed the most common ingredient was a salad mix of Iceberg lettuce, Apollo lettuce and Spinach. Five samples of mixed leaf salad were taken from two sandwich preparation sites for food outlet A (Scotland = 3; North of England = 2), and one sample tested positive for a non-O157 STEC isolate. However, microbiological typing showed that the isolate belonged to STEC O36:H19 stx2e and was unrelated to the outbreak strain. Although this does not constitute direct evidence that this salad was the source of the outbreak as it is a different strain, it does highlight a potential pathway of human exposure to STEC from the mixed leaf salad.

The salad mix was supplied by FBO A. Test results provided by FBO A for samples of this product from 1st August to 28th October 2019, showed that samples had <10 cfu/g for E. coli. EHOs from the local authority for FBO A visited the food processing and distribution site, and took seven samples of the mixed leaf product supplied to food outlet A. One sample with a production date of 30/09/2019 and a use-by date of 04/10/2019 had a result for E. coli of 30 cfu/g but was negative for STEC. The local authority for FBO A took nine environmental swabs including swabs of drains, multi-head packers, lines and post wash product samples on the 4th November 2019. All samples tested negative for E. coli. FBO A reported that they routinely test environmental swabs on a weekly basis to confirm that cleaning has been effective. All swabs taken between August and October 2019 tested negative for E. coli. The EHOs visited the site and were satisfied with the processes that FBO A had in place.

An investigation of the growers supplying FBO A was carried out for the period between 6th of September and 22nd of October. All businesses confirmed that they did not have any retained product or soil samples from this specific time frame and that the results of microbiological sampling during the time frame under investigation had been satisfactory. During the time frame of interest, FBO A was supplied by 11 different growers, 10 farms were located in the UK and one farm was located in France. The IMT reviewed documentation, quality standards and audit reports for all growers supplying FBO A and found them to be satisfactory. Due to the large number of farms involved in the supply chain, the time that had elapsed since the production of the implicated batch of salad leaves, and the absence of an organism linking the mixed leaf salad directly to the outbreak, the IMT concluded it would be disproportionate to investigate further at the on-farm level.

Discussion

In the UK, outbreaks caused by STEC O26:H11 occur less frequently than outbreaks of STEC O157:H7 (https://www.gov.uk/government/publications/escherichia-coli-e-coli-o157-annual-totals). It is uncertain whether this is a true reflection of differences in the burden of disease, animal reservoirs and/or transmission routes, or due to the limitations of the surveillance system for detecting cases of STEC O26:H11. Coverage of the population in Scotland is comprehensive, however, only patients infected with STEC O26:H11 presenting with bloody diarrhoea are captured, and cases with milder symptoms may be missed. Whereas in England, surveillance of STEC O26:H11 is restricted to areas where local hospital laboratories have implemented PCR. At the time of the incident, approximately 20% of local hospital microbiology diagnostic laboratories distributed across England had implemented the commercial GI PCR assay. All hospitals in Wales have implemented the PCR approach.

In England, the STEC Operational Guidance focuses on public health follow up, with respect to administering the ESQ and requiring microbiological clearance of patients in risk groups, of those cases with STEC harbouring stx2 (https://www.gov.uk/government/publications/shiga-toxin-producing-escherichia-coli-public-health-management). Consequently, ESQs on a subset of patients linked to this outbreak and resident in England were not initially available thus hindering the epidemiological analysis in the early phase of the investigation. Evidence in the literature for the association between STEC with stx2a and the potential to cause HUS is well-established [2]. However, other studies have highlighted the role of stx1a as a marker for severe clinical outcomes, specifically as a cause of bloody diarrhoea and increasing the risk of hospitalisation [2, Reference Byrne23]. During this outbreak, the number of patients in England reporting bloody diarrhoea (85%) was higher when compared to those expected for STEC O157:H7, although hospitalisation rates (23%) were lower (bloody diarrhoea 61% and hospitalisation 34% from Byrne et al. 2015 [Reference Byrne11]). In Scotland, hospitalisation rates for all cases of STEC in 2019 were 38% (STEC O157:H7 45%; non-O157 STEC 28%) and bloody diarrhoea was reported by 79% (STEC O157:H7 71%; non-O157 STEC 89%) https://hpspubsrepo.blob.core.windows.net/hps-website/nss/3109/documents/2_stec-in-scotland-2019-full-report.pdf. The high rates of bloody diarrhoea in non-O157 cases in Scotland may reflect the selection bias inherent in the procedure for referral of cases to SERL.

Globally, foodborne outbreaks of STEC O26:H11 have been previously associated with beef products [Reference Werber24, Reference Ethelberg25], dairy products [Reference Jones26, Reference Germinario27], and vegetables and salad items [Reference Peron28, Reference Terajima29]. Outbreaks of STEC O26:H11 caused by person-to-person contact, specifically in nursery school settings, have also been described [Reference McMaster30Reference Scavia34]. These reports of outbreak associated with person-to-person spread provide evidence of the transmissibility of STEC O26:H11, and the need for public health follow-up of cases and exclusion of children aged five and under from school or childcare facilities, until microbiologically clear of the aetiological agent. In England, Scotland and Wales, it is also a requirement that food handlers and clinical/social care staff working with susceptible patients are microbiologically clear of the aetiological agent before returning to work.

Although the outbreak strain was not recovered from the implicated food source, the epidemiological analysis provided robust evidence that one of the components of the mixed leaf salad was the contaminated vehicle. Outbreaks caused by contaminated leafy greens and other salad vegetables are challenging to investigate because these food items are often minor ingredients of a meal that the case may fail to recall [Reference Byrne35]. With respect to the outbreak described here, the implicated mixed leaf salad was a component of sandwich fillings, or garnish. The link between the cases was established because the majority of infected individuals reported buying takeaway food products from the same national food chain franchise (food outlet A), and other food outlets supplied by FBO A. Contamination of produce during cultivation or processing is often transient and caused by a one-off event, such as flooding, or a temporary failure in the production process caused by poor manufacturing processes [Reference Gobin36Reference Kintz38]. Another problematic aspect of investigating outbreaks of STEC is that contaminated food capable of causing illness may contain the pathogen at low levels, and/or be heterogeneously distributed in the food matrix, and so detection of the pathogen is challenging, even when large analytical units of food are tested, and a robust sampling plan is in place [Reference Jenkins39, Reference Zhang40]. The fact that non-O157 STEC, including STEC O26:H11, are not part of the testing algorithm of food samples in the UK, and are difficult to differentiate from other non-pathogenic E. coli that may be present, further confounds detection.

During this outbreak investigation a strain of STEC was isolated from a sample of the mixed salad, although it was not identified as the outbreak strain. The confirmed presence of STEC in a batch of ready-to-eat (RTE) food falling is considered a serious risk to public health. Although not all strains of STEC have been found to cause bloody diarrhoea and/or STEC-HUS, a precautionary approach is appropriate given the uncertainty in the evidence and the potential for severe clinical outcomes. The FSA's current view is that the confirmed presence of STEC in RTE food is an unacceptable risk to public health and that it is appropriate to take action to remove contaminated food from the market (https://acmsf.food.gov.uk/sites/default/files/acm_1191_stec.pdf). The competent authority should be notified through incident reporting procedures and action should be taken to withdraw affected batches from the market in accordance with Article 19 of Regulation (EC) 178/2002. Information on the onward supply of the product is required to determine whether a product recall from end users/ consumers would also be appropriate. Following this incident, the product was no longer available for recall because of the time delay between the sampling and obtaining the result. However, follow up investigations were initiated by the FBO to investigate the source of STEC contamination, and the HACCP-based food safety processes were reviewed.

During the outbreak described in this study, on-farm sampling was deemed inappropriate due the lack of microbiological evidence that the mixed salad was the vehicle, the time since the contaminated product would have been on farm, the absence of evidence of ongoing contamination, and the resource commitment required. Moreover, during the time frame under investigation, the components of the mixed leaf salad had been sourced from many different farms, and that since that time there had been a seasonal shift in growers providing the supplier. However, it should be noted that STEC can survive in soil for many months [Reference Ogden41], so even if the fields have been harvested and it has been some time since the outbreak, it may be still possible to detect the outbreak strain, and given the limitations of testing produce for STEC described above, prioritising on-farm investigations may be the most appropriate course of action. On-farm investigations during previous outbreaks linked to contaminated produce identified ruminants grazing in close proximity to the crops, and poor practice, such as the use of river or pond water for irrigation [Reference Mikhail37Reference Jenkins39, Reference Söderström42]. In North America, in order to gain a better understanding of the risks associated with fresh produce there has been an increase in the emphasis placed on root cause investigations, such as sampling the farm environment, specifically soil, water sources and animal faeces, on-farm environmental audits and assessments of the processes and practices throughout the supply chain (https://www.foodsafetynews.com/2021/01/fda-tests-show-cattle-lot-implicated-in-leafy-greens-e-coli-outbreak/#more-200609, https://www.fda.gov/food/foodborne-pathogens/2020-leafy-greens-stec-action-plan).

Outbreaks of severe gastrointestinal disease caused by salad vegetables and other types of raw produce contaminated with STEC continue to be a major public health concern. Salad vegetables are sold as a ready-to-eat food, as defined in Article 14 of Regulation (EC) 2073/2005 (https://www.fsai.ie/food_businesses/micro_criteria/reg_2073_05.html). Contaminated ready-to-eat foods are regarded as a high risk compared to food that requires cooking prior to consumption. Although the production process involves a stringent washing process that removes 90–99% of external bacteria, low numbers of residual bacteria may be sufficient to cause a significant risk of infection. Outbreaks linked to contaminated produce, such as the one described here, are often transient events. Microbiological testing of food that has a short-shelf life, and where the pathogen may be present in low numbers and have a heterogenous distribution in the matrix, means that interpretation of results can be challenging as negative results do not necessarily mean the food is not the vehicle of infection. Even when PCR detects the presence of stx on the food matrix, confirmation of the presence of STEC can be challenging, as non-O157 STEC are not phenotypically distinguishable from other E. coli. Robust, statistically significant epidemiological analysis be regarded as sufficient evidence to direct timely and targeted on-farm investigations. Such an approach is essential to identify the root cause of outbreaks linked to salad and raw vegetables, and to establish an evidence-base for improving guidance and policy.

Financial support

CJ and DRG are affiliated to the National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Gastrointestinal Infections at University of Liverpool in partnership with Public Health England (PHE), in collaboration with University of Warwick, and are based at PHE. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, the Department of Health and Social Care or Public Health England.

Conflict of interest

There are no conflicts of interest

References

Croxen, MA et al. (2013) Recent advances in understanding enteric pathogenic Escherichia coli. Clinical Microbiology Reviews 26, 822880.CrossRefGoogle ScholarPubMed
EFSA BIOHAZ Panel (2020) Pathogenicity assessment of Shiga toxin-producing Escherichia coli (STEC) and the public health risk posed by contamination of food with STEC. European Food Safety Authority Journal 18, 5967.Google Scholar
Melton-Celsa, AR (2014) Shiga toxin (Stx) classification, structure, and function. Microbiology Spectrum 2, EHEC-0024-2013.CrossRefGoogle Scholar
Tarr, PI et al. (2005) Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome. Lancet (London, England) 365, 10731086.Google ScholarPubMed
Joseph, A et al. (2020) Shiga toxin-associated hemolytic uremic syndrome: a narrative review. Toxins 12, 67.CrossRefGoogle ScholarPubMed
Furniss, RCD and Clements, A (2017) Regulation of the locus of enterocyte effacement in attaching and effacing pathogens. Journal of Bacteriology 200, e00336–17.Google ScholarPubMed
Stevens, MP and Frankel, GM (2014) The locus of enterocyte effacement and associated virulence factors of enterohemorrhagic Escherichia coli. Microbiology Spectrum 2, EHEC-0007-2013. doi: 10.1128/microbiolspec.EHEC-0007-2013.CrossRefGoogle ScholarPubMed
Thomas, KM et al. (2013) Transfer of verocytotoxigenic Escherichia coli O157, O26, O111, O103 and O145 from fleece to carcass during sheep slaughter in an Irish export abattoir. Food Microbiology 34, 3845.CrossRefGoogle Scholar
McCabe, E et al. (2019) An investigation of shedding and super-shedding of Shiga toxigenic Escherichia coli O157 and E. coli O26 in cattle presented for slaughter in the Republic of Ireland. Zoonoses & Public Health 66, 8391.CrossRefGoogle Scholar
Navarro-Gonzalez, N et al. (2020) Carriage and subtypes of foodborne pathogens identified in wild birds residing near agricultural lands in California: a repeated cross-sectional study. Applied Environmental Microbiology 86, e01678–19.CrossRefGoogle ScholarPubMed
Byrne, L et al. (2015) The epidemiology, microbiology and clinical impact of Shiga toxin-producing Escherichia coli in England, 2009–2012. Epidemiology & Infection 143, 34753487.CrossRefGoogle Scholar
Adams, NL et al. (2016) Shiga toxin-producing Escherichia coli O157, England and Wales, 1983–2012. Emerging Infectious Diseases 22, 590597.CrossRefGoogle Scholar
Byrne, L et al. (2014) Epidemiology and microbiology of Shiga toxin-producing Escherichia coli other than serogroup O157 in England, 2009–2013. Journal of Medical Microbiology 63, 11811188.CrossRefGoogle Scholar
Gentle, A et al. (2020) Antimicrobial resistance in Shiga toxin-producing Escherichia coli other than serotype O157:H7 in England, 2014–2016. Journal of Medical Microbiology 69, 379386.CrossRefGoogle Scholar
Jenkins, C et al. (2012) Assessment of a real-time PCR for the detection and characterization of verocytotoxigenic Escherichia coli. Journal of Medical Microbiology 61, 10821085.CrossRefGoogle ScholarPubMed
Jenkins, C et al. (2020) Shiga toxin-producing Escherichia coli haemolytic uraemic syndrome (STEC-HUS): diagnosis, surveillance and public-health management in England. Journal of Medical Microbiology 69, 10341036.CrossRefGoogle ScholarPubMed
Chattaway, MA et al. (2016) Whole genome sequencing for public health surveillance of Shiga toxin-producing Escherichia coli other than serogroup O157. Frontiers in Microbiology 7, 258.CrossRefGoogle ScholarPubMed
Holmes, A et al. (2018) Validation of whole-genome sequencing for identification and characterization of Shiga toxin-producing Escherichia coli to produce standardized data to enable data sharing. Clinical Microbiology 56, e01388–17.CrossRefGoogle ScholarPubMed
Ashton, PM et al. (2015) Insight into Shiga toxin genes encoded by Escherichia coli O157 from whole genome sequencing. PeerJ 3, e739.CrossRefGoogle ScholarPubMed
Dallman, TJ et al. (2018) SnapperDB: a database solution for routine sequencing analysis of bacterial isolates. Bioinformatics (Oxford, England) 34, 30283029. PMID: 29659710.CrossRefGoogle ScholarPubMed
Dallman, TJ et al. (2015) Whole-genome sequencing for national surveillance of Shiga toxin-producing Escherichia coli O157. Clinical Infectious Disease 61, 305312.CrossRefGoogle ScholarPubMed
Jenkins, C, Dallman, TJ and Grant, KA (2019) Impact of whole genome sequencing on the investigation of food-borne outbreaks of Shiga toxin-producing Escherichia coli serogroup O157:H7, England, 2013 to 2017. Euro Surveillance 24, 1800346.Google Scholar
Byrne, L et al. (2020) Association between Shiga toxin-producing Escherichia coli O157:H7 stx gene subtype and disease severity, England, 2009–2019. Emerging Infectious Diseases 26, 23942400.CrossRefGoogle Scholar
Werber, D et al. (2002) A multistate outbreak of Shiga toxin-producing Escherichia coli O26:H11 infections in Germany, detected by molecular subtyping surveillance. Journal of Infectious Diseases 186, 419422.CrossRefGoogle ScholarPubMed
Ethelberg, S et al. (2009) Outbreak of non-O157 Shiga toxin-producing Escherichia coli infection from consumption of beef sausage. Clinical Infectious Diseases 48, e78e81.CrossRefGoogle ScholarPubMed
Jones, G et al. (2019) Outbreak of Shiga toxin-producing Escherichia coli (STEC) O26 paediatric haemolytic uraemic syndrome (HUS) cases associated with the consumption of soft raw cow's milk cheeses, France, March to May 2019. EuroSurveillance 24, 1900305. doi: 10.2807/1560-7917.ES.2019.24.22.1900305CrossRefGoogle Scholar
Germinario, C et al. (2016) Community-wide outbreak of haemolytic uraemic syndrome associated with Shiga toxin 2-producing Escherichia coli O26:H11 in southern Italy, summer 2013. EuroSurveillance 21, 30343. doi: 10.2807/1560-7917.ES.2016.21.38.30343.CrossRefGoogle Scholar
Peron, E et al. (2016) Early findings in outbreak of haemolytic uraemic syndrome among young children caused by Shiga toxin-producing Escherichia coli, Romania, January to February 2016. EuroSurveillance 21, 30170.CrossRefGoogle ScholarPubMed
Terajima, J et al. (2014) Shiga toxin (verotoxin)-producing Escherichia coli in Japan. Microbiology Spectrum 2. doi: 10.1128/microbiolspec.EHEC-0011-2013. PMID: 26104366.CrossRefGoogle ScholarPubMed
McMaster, C et al. (2001) Verocytotoxin-producing Escherichia coli serotype O26:H11 outbreak in an Irish crèche. European Journal of Clinical Microbiology & Infectious Diseases 20, 430432.Google Scholar
Dallman, TJ et al. (2015) The utility and public health implications of PCR and whole genome sequencing for the detection and investigation of an outbreak of Shiga toxin-producing Escherichia coli serogroup O26:H11. Epidemiology & Infection 143, 16721680.CrossRefGoogle ScholarPubMed
Hiruta, N et al. (2001) An outbreak of diarrhoea due to multiple antimicrobial-resistant Shiga toxin-producing Escherichia coli O26:H11 in a nursery. Epidemiology & Infection 127, 221227.CrossRefGoogle Scholar
Moran-Gilad, J et al. (2017) Real-time genomic investigation underlying the public health response to a Shiga toxin-producing Escherichia coli O26:H11 outbreak in a nursery. Epidemiology & Infection 145, 29983006.CrossRefGoogle Scholar
Scavia, G et al. (2018) A case of haemolytic uraemic syndrome (HUS) revealed an outbreak of Shiga toxin-2-producing Escherichia coli O26:H11 infection in a nursery, with long-lasting shedders and person-to-person transmission, Italy 2015. Journal of Medical Microbiology 67, 775782.CrossRefGoogle Scholar
Byrne, L et al. (2016) Epidemiological and microbiological investigation of an outbreak of severe disease from Shiga toxin-producing Escherichia coli O157 infection associated with consumption of a slaw garnish. Journal Food Protection 79, 11611168.CrossRefGoogle ScholarPubMed
Gobin, M et al. (2018) National outbreak of Shiga toxin-producing Escherichia coli O157:H7 linked to mixed salad leaves, United Kingdom, 2016. EuroSurveillance 23, 17-00197. doi: 10.2807/1560-7917.ES.2018.23.18.17-00197CrossRefGoogle ScholarPubMed
Mikhail, AFW et al. (2018) An outbreak of Shiga toxin-producing Escherichia coli O157:H7 associated with contaminated salad leaves: epidemiological, genomic and food trace back investigations. Epidemiology & Infection 146, 187196.CrossRefGoogle ScholarPubMed
Kintz, E et al. (2019) Outbreaks of Shiga toxin-producing Escherichia coli linked to sprouted seeds, salad, and leafy greens: a systematic review. Journal of Food Protection 82, 19501958.CrossRefGoogle ScholarPubMed
Jenkins, C et al. (2015) Public health investigation of two outbreaks of Shiga toxin-producing Escherichia coli O157 associated with consumption of watercress. Applied Environmental Microbiology 8, 39463952.CrossRefGoogle Scholar
Zhang, G et al. (2018) Survey of foodborne pathogens, aerobic plate counts, total coliform counts, and Escherichia coli counts in leafy greens, sprouts, and melons marketed in the United States. Journal of Food Protection 81, 400411.CrossRefGoogle ScholarPubMed
Ogden, ID et al. (2002) Long-term survival of Escherichia coli O157 on pasture following an outbreak associated with sheep at a scout camp. Letters in Applied Microbiology 34, 100104.CrossRefGoogle Scholar
Söderström, A et al. (2008) A large Escherichia coli O157 outbreak in Sweden associated with locally produced lettuce. Foodborne Pathogens and Disease 5, 339349.CrossRefGoogle ScholarPubMed
Figure 0

Fig. 1. Phylogenetic tree showing the relationship between isolates linked to the outbreak from cases resident in England (n = 16), Scotland (n = 14) and Wales (n = 2) that fell within the same five SNP single linkage cluster highlighted in the blue box. The samples outside the blue box were included to provide context and represent all the isolates in the Public Health England (PHE) archive (2015–2019) that fell within a 25 SNP single linkage cluster of the outbreak cluster. Short read archive accession numbers (SRRs) are provided for the sequences of the isolates sequenced at PHE.

Figure 1

Fig. 2. Geographical distribution of the confirmed cases (n = 32).

Figure 2

Fig. 3. Epidemic curve of confirmed cases by onset date, or sample date where onset date was not available (n = 32). Both cases positioned in mid-November were based on sample dates; one case reported being symptomatic since October and no information was available for the second case.

Figure 3

Fig. 4. Age−sex distribution of the confirmed cases (n = 32).

Figure 4

Table 1. Distribution of clinical symptoms amongst confirmed cases with a completed questionnaire

Figure 5

Table 2. In univariable analysis of the association between being infected with the outbreak strain and consuming food purchased at different food outlets

Figure 6

Table 3. Generalised linear model estimates based upon logistic regression adjusted for age and sex and significance level adjusted for multiple testing

Figure 7

Table 4. Univariate case−control analysis of food exposures

Figure 8

Table 5. Results of multivariable logistic regression for food exposures (n = 225)

You have Access Open access
1
Cited by

Save article to Kindle

To save this article to your Kindle, first ensure coreplatform@cambridge.org is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle.

Note you can select to save to either the @free.kindle.com or @kindle.com variations. ‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

Epidemiological investigations identified an outbreak of Shiga toxin-producing Escherichia coli serotype O26:H11 associated with pre-packed sandwiches
Available formats
×

Save article to Dropbox

To save this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Dropbox account. Find out more about saving content to Dropbox.

Epidemiological investigations identified an outbreak of Shiga toxin-producing Escherichia coli serotype O26:H11 associated with pre-packed sandwiches
Available formats
×

Save article to Google Drive

To save this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Google Drive account. Find out more about saving content to Google Drive.

Epidemiological investigations identified an outbreak of Shiga toxin-producing Escherichia coli serotype O26:H11 associated with pre-packed sandwiches
Available formats
×
×

Reply to: Submit a response

Please enter your response.

Your details

Please enter a valid email address.

Conflicting interests

Do you have any conflicting interests? *