Background
Enhanced surveillance systems for STEC across the UK combine detailed clinical and epidemiological data (including symptoms, travel, food, and animal exposure) collected on enhanced surveillance questionnaires (ESQ) with the microbiological characterization of strains using whole genome sequencing (WGS) [Reference Butt1]. Diagnostic laboratories report presumptive cases of STEC based on PCR or culture, directly to health protection teams (and to local authorities in Wales), who undertake public health management including collection of information via the STEC ESQ within 48 h.
Faecal specimens from suspected cases of STEC and/or isolates of STEC are referred to the UKHSA Gastrointestinal Bacteria Reference Unit (GBRU) in London or the Scottish E. coli Reference Laboratory (SERL) in Edinburgh. Faecal specimens testing positive for STEC by PCR are cultured and all isolates of STEC are sequenced. Characterization includes clonal complex and sequence typing, serotyping, stx typing, and SNP typing [Reference Chattaway2].
Descriptive epidemiology
Of the 288 reported cases, confirmed to be linked by WGS within a five SNP cluster, 286 were symptomatic primary cases; four (two in England and two in Scotland) were symptomatic secondary cases. Symptom onset dates of the primary cases ranged from 29 April 2024 to 17 June 2024 (Figure 1). Primary cases had a median age of 29 years (range: 1–89) and were predominantly female (57%) (Figure 2). Cases were geographically dispersed across the UK. For cases where information was available (n = 263), 49% of cases were hospitalized, and 80% of symptomatic cases reported bloody diarrhoea. There were nine cases of haemolytic uraemic syndrome (HUS), and two deaths among these confirmed cases (Table 1).
Figure 1. Confirmed primary cases by date of symptom onset (n = 248).
Note: Onset date is unavailable for 38 cases.
Figure 2. Age/sex pyramid of cases (n = 288).
Table 1. Clinical information including reported symptoms (n = 263)
Note: Data is unavailable for 25 cases.
Hypothesis generation
Data collected through UKHSA’s National Enhanced STEC Surveillance System (NESSS), and complementary enhanced surveillance in Scotland, Wales, and Northern Ireland, were reviewed. A case–case study was conducted using English cases in the outbreak cluster as cases, and other English STEC cases as controls. Cases and controls with the same age profile and with sample dates from April 2024 to May 2024 were included. The analysis indicated pre-packaged sandwiches as a possible hypothesis, OR 4.91 (95%CI 1.51–15.1, P 0.004). A similar study in Wales identified the same hypothesis. In logistic regression with subsequent cases (outbreak cases n = 59, control cases n = 64) the final model included pre-packaged sandwiches (OR 3.88, 95%CI 1.65–9.57, P 0.002), iceberg lettuce (OR 2.99, 95%CI 1.24–7.48, P 0.016), and eating out (OR 2.17, 95%CI 0.91–5.37, P 0.08) as significant exposures for the outbreak cases.
In total, 11 of 15 cases interviewed with a trawling questionnaire reported consuming pre-packaged sandwiches from different national retailers. Additionally, given the incubation period of STEC, the epi curve indicated that the exposure period for the cases must have been very brief and therefore suggestive of a short shelf-life product. Based on the generated hypothesis, we undertook an analytical study using targeted questionnaires with more detail on pre-packaged sandwiches, eating out, and salad consumption. Case data were correct as of 09 June 2024 (n = 43).
Analytical studies
Following the initial case–case studies, outbreak cases were compared to two sources of control. All controls were frequency matched to cases in age bands and reported no travel outside of the UK in the week before data collection. Controls in study 1 were cases of Salmonella residents in the UK with a notification date from 01 April 2024 to 09 June 2024, (n = 63) and were asked about their food histories for the week prior to their onset with Salmonella. Controls in study 2 were recruited by a Market Panel [Reference Mook3] and reported no diarrhoea in the previous week (06 June 2024).
In both analytical studies, variables significantly associated with outbreak case status (odds ratio (OR) > 1 and P < 0.1) in single variable analysis and age and sex as a priori potential confounding variables were included in a multivariable Firth logistic regression model using a forward step approach for model construction with all models including the potential confounders.
In the multivariable models, cases were significantly more likely to have consumed pre-packaged sandwiches containing lettuce (for Study 1 Model 1 OR 7.1, 95%CI 2.3–21.5, P 0.001; and for Study 2 Model 1 OR 4.8, 95%CI 1.9–12.0, P < 0.001) (Table 2).
Table 2. Multivariable analysis of estimated odds ratios of infection with STEC O145 t5.206
* aOR: Adjusted odds ratio.
** CI: confidence interval.
In Study 2 Model 2 (Table 2), cases were significantly more likely to have consumed a prepackaged sandwich with lettuce compared to any other type of sandwich or no sandwiches (OR 7.1, 95%CI 3.0–17.5, P < 0.001).
A separate model to compare lettuce consumed in sandwiches to lettuce consumed when eating out showed the latter was not significant either (Study 2 Model 3) (Table 2).
Food chain and environmental investigation
Food chain investigations identified the sandwich producer that supplied the retailers during May 2024. The sandwich producer had sourced lettuce from farms in England. Further food chain investigations are ongoing.
Control measures
On 13 June 2024, supplier/producer A and B voluntarily recalled various sandwiches, wraps, and salads because of possible contamination with E. coli. Consumers who had bought the products listed were advised not to eat them, and to return them to the store where they were purchased for a full refund. A further recall occurred on 15 June 2024 by supplier/producer C.
Epipulse communications to ECDC indicated of the 13 European countries who replied, none were affected.
Discussion
Since 2015, across the UK notifications of STEC O157:H7 have declined and STEC O26:H11 and STEC O145:H28 have emerged as a significant cause of gastrointestinal infectious disease and HUS [Reference Butt1, Reference Vishram4, Reference Rodwell5]. Since 2020, STEC O145:H28 has consistently been in the top five most common STEC serotypes reported in England and Scotland [Reference Rodwell5, 6].
Outbreaks of STEC infection have previously been associated with pre-packed sandwiches and salad vegetables, mainly lettuce, in the UK and elsewhere [Reference Mikhail7–Reference Kintz10]. Ready-to-eat salad vegetables are vulnerable to contamination with pathogens at the pre-harvest level via flooding, rainwater run-off, or irrigation water containing animal faeces [Reference Dogan11]. Current methods for washing and decontaminating fresh produce cannot guarantee that pathogens, if present, will be removed. STEC can adhere to leaves and become internalized within leafy vegetables [Reference Zhao12]. The application of controls to minimize the risk of faecal contamination during growing, handling, and processing is therefore of fundamental importance in ensuring the safety of fresh produce.
Monitoring PCR results provided an early indication of the outbreak and surveillance data case–case analysis facilitated rapid hypothesis generation. Subsequent analytical studies established a precise definition and exploration of the composite product. Interdisciplinary collaboration and cooperation from the major food retailers led to voluntarily removal of the implicated product from sale thus reducing the risk of an on-going transmission.
Number of confirmed cases decreased since 31 May 2024, but the outbreak has not yet been declared over. Food chain investigations are ongoing, and the location of the animal reservoir and/or mechanisms of crop contamination are currently unclear. Possible routes of contamination include a failure in control measures protecting the crop from agricultural run-off, contamination of water or growing materials used in lettuce production, or contaminated seeds. The implicated lettuce is a UK product, and no cases are known to have occurred outside the UK. Nevertheless, the international community should be aware of this vehicle of infection and monitor for possible ongoing cases linked to this outbreak of STEC O145:H28, as similar routes of contamination may occur in other countries.
Acknowledgements
The article has been written on behalf of the Incident Management Team. The authors thank the following for their contributions to the outbreak investigations: cases who provided detailed information; Health Protection Teams, the Rapid Investigation Team, the Syndromic Surveillance team, other Field Services Teams, the Gastrointestinal Infections, Food Safety and One Health Division and Local Authorities for their assistance in the investigation of cases and conducting interviews; frontline and reference laboratory colleagues, as well as members of the Diagnostics and Genomic Cell for their laboratory investigations including the generation, analysis and interpretation of WGS data; colleagues at the Food Standards Agency (including Wales and Northern Ireland) who led the Food Chain and Sampling Cell, as well as Food Standard Scotland; colleagues at the Animal and Plant Health Agency who supported environmental investigations into the root cause of the outbreak; to the staff in the Clinical Cell for clinical follow up of cases, assessing severity and liaising with National Health Service England colleagues; the Communications Cell for managing all internal and external communications including proactive public health messaging and to the members of the Data, Epidemiology and Analysis Cell and all the other staff from UK Health Security Agency, Public Health Scotland, Public Health Wales and Public Health Agency of Northern Ireland who assisted with this outbreak investigation and supported the completion of the analytical studies.
Author contribution
All authors were directly involved in the gathering of epidemiological and microbiological data to inform the management and investigation of the outbreak. Outbreak management was led by Amy Douglas, Lesley Larkin, Trish Mannes, Kitty Mohan, and Colin Brown. Interpretation of typing and whole genome sequencing data was led by Anaïs Painset, Claire Jenkins, and Lesley Allison. Coordination of Welsh cases, data, and response was managed by Andrew Nelson, and Christopher Williams. Coordination of Scottish cases, data, and response was managed by Genna Leckenby and Sema Nickbakhsh. Coordination of Northern Irish cases, data, and response was managed by Claire Neill. Follow up of English cases with trawling questionnaires was led by Oluwakemi Olufon and Cary Rees, with data collated and summarised by Orlagh Quinn, Chloe Byers, Claire Brown, and Ann Hoban. Additional analysis of the epidemiological data for hypothesis generation was led by Orlagh Quinn, Yanshi, Sooria Balasegaram, and Ann Hoban. Analytical studies were conducted by Clare Sawyer, Grace King, Ann Hoban, Andre Charlett, Sooria Balasegaram, and Tom Inns. Writing of the manuscript was led by Lesley Larkin, Sooria Balasegaram, Claire Jenkins, Orlagh Quinn, Yanshi, Grace King, and Amy Douglas. All authors reviewed the draft manuscript, submitted contributions, and approved the final version.
Competing interest
None declared.
Ethical standard
UKHSA has legal permission, provided by Regulation 3 of the Health Service (Control of Patient Information) Regulations 2002, to process patient confidential information for national surveillance of communicable diseases.
Funding statement
This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Open access
