Among those affected by the summer’s E.Coli outbreak, 122 people were hospitalised. More than half were female (57%) and had a median age of 35 (ranging from 6-85 years).
The STEC O145 t5:206 outbreak strain possesses a virulence gene profile (stx2a, eae+) which is associated with an increased likelihood of Haemolytic Uraemic Syndrome (HUS) - a condition that can lead to kidney failure and death. There were seven cases of HUS confirmed across the UK.
There were two deaths in patients within 28 days of confirmation of infection with the STEC outbreak strain t5:206. Neither had HUS and both were adults who had underlying medical conditions.
Based on the information available, the Food Standards Agency (FSA) believes one of these deaths is likely linked to STEC infection and the associated contaminated product - lettuce.
“Based on our extensive food chain analysis and epidemiological evidence, underpinned by two types of case studies, the likely cause of the outbreak was sandwiches containing Apollo lettuce,” Tina Potter, head of incidents for the FSA told Food Manufacture.
How does E. coli get into our produce in the first place?
E. coli is a bacteria commonly found in the gut of warm-blooded organisms. While most strains are harmless, some can cause serious problems such as Shiga toxin-producing E. coli (STEC).
The majority of the time, STEC outbreaks are linked to raw or undercooked ground meat products, raw milk, and faecal-contaminated vegetables.
Aidan Taylor, microbiology lecturer at Reading University, explained more: “E. coli that ends up on crops starts life in the gut of animals, whether that’s humans or livestock. When animals defecate they pass these bacteria into the environment, where they are now looking for a new host to infect. To do this, they must be ingested by the new host, either by drinking contaminated water or eating contaminated food.
“Modern sewage networks normally deal with this problem by taking human waste directly and safely to treatment plants where the bacteria are killed before they have the chance to interact with, or infect, anyone else. Of course, we have all seen in recent years how regularly this system fails due to over-stretched infrastructure, leading to untreated sewage leaking into water courses.”
On farms the situation is more complex, as animal faeces are often composted to generate manure which is then used as fertiliser on crops.
Aidan Taylor, microbiology lecturer - Reading University
He continued: “The other source of E. coli contamination on ready to eat crops comes from water. Lettuce in particular requires regular irrigation which has the potential to coat the crop in pathogenic bacteria if the water is contaminated. Since these crops are not cooked and, importantly, washing is largely ineffective at removing bacteria like E. coli, the risk of infection to consumers is high.
“Irrigation water can be contaminated from various sources, including close proximity of livestock and human sewage leaks. Mains water is generally safe, but rare contamination events can occur when infrastructure is damaged. Irrigation systems which use local, open sources such as ponds and canals are much more susceptible to contamination by livestock.
“Proper separation of livestock and irrigation water sources on farms is critical, particularly for ready to eat crops. However, contaminated water may also reach crops during heavy rainfall as run off from adjacent fields, or indeed from large scale flooding. Containment is much more challenging in this scenario and is but one factor that requires robust infrastructure to mitigate the effects of climate change on crop production.”
Highly irregular
Speaking with a number of industry representatives, it’s evident that this was a rare and unusual event.
This was a “freak” incident, Jim Winship director of the British Sandwich Association explained, as he emphasised the robust processes already in place among members.
Investigations are still on-going, but when something like this happens it’s always prudent to reflect and look at one’s own processes.
Tom Hollands, innovation and technical director of sandwich manufacturer Raynor Foods, agreed with Winship’s thoughts.
Whilst Raynor Foods was not impacted by the event, the company did treat it as a near-miss and subsequently looked at what lessons could be learned from this.
So…what happened?
As the team took stock, one of the interesting nuggets of information they unearthed was around whole-headed product.
Whilst the micro standards for wholesalers or suppliers dictate that pre-cut fruit and vegetables are tested, the same does not apply to whole-headed product that’s required to be washed or prepared by the customer.
The irrigation water must be tested and if that comes back clean, one can generally assume that the product is also clean. Another safety layer to that is to test the soil, but that’s not a requirement.
It’s also worth noting new strains of bacteria like E. Coli can be problematic as the chemicals used in the wash may not be suitable to address them. Although this was not the case in this scenario, as t5:206 is known to us.
Moreover, with certain pathogens, chlorinating can sometimes fall short if the bacteria are hiding in the xylem (capillaries) of the fresh fruit or veg (this is more likely with Listeria).
Hollands acknowledges that the systems in place are robust, but he also wonders whether there is a gap to be filled in whole-head testing. During our conversation, he ventured that it could be beneficial to extend sampling and screening to cover this and introduce an additional step wherein supply chain partners advise when there are slightly elevated levels of E. coli.
Another possibility Hollands raised was infected seed. In this case, washing the lettuce once it had grown would not have been sufficient to remove the bacteria and this would have come down to a failure much earlier in the supply chain.
“When we were running Rosemary Gardens here, that was one of the key areas that we focused on. We chlorinated all the seeds in a reasonably strong dilution before we started growing them in high care conditions,” said Hollands. “We’d also test the water very frequently as well as the final crop.”
Taylor offered further context on systems failures that could occur earlier in the chain. As referenced earlier, animal poo is often used on farms as fertiliser, but if it’s properly matured manure, you can pretty much assume it will be largely free of pathogens like E. coli.
“These bacteria will survive for several months in manure but are killed off over time by a combination of temperature, sunlight and drying,” Taylor explained.
“For this reason, thorough rotating and composting of manure is essential prior to spreading. Even so, spreading should occur well in advance of seeding crops to ensure any remaining pathogenic bacteria die off on the field: the FSA recommendation is at least 6 months. Other cautious behaviour can help reduce the risk of contamination further, particularly in the case of ready to eat crops like lettuce. These include a minimum 12-month gap between livestock grazing or fresh manure spreading and crop harvesting, as fresh manure and animal dung carries a much higher risk of contamination than matured manure.”
As mentioned above, the actual cause of this specific outbreak is still unknown. Whilst it’s linked to Apollo lettuce, how it happened remains unclear.
What has the investigation found?
Whilst the industry experts raised important questions, speaking more recently with the FSA, the agency explained that, so far, no evidence has emerged that irrigation water was the source of contamination. Neither was the use of manure a factor of concern in this particular incident.
The agency also noted that in some cases, irrigation water had not been used on the fresh produce.
Although the FSA investigated potential contamination of the seeds in an effort to be thorough, it’s also not likely to have been a contamination at this stage either, as there would have been a significantly wider impact on the supply chain, across multiple countries, based on the way seedlings are used in the sector.
One of the other possibilities raised by the industry experts during our initial conversations was the issue of temperature abuse. E. Coli can’t grow at certain temperatures and is killed when product is cooked – so if a sandwich, which is not heated suffered temperature abuse (for example, was left in a car for too long between a store and consuming or a supermarket fridge door was left open) it could have reached an optimal state for E.Coli to grow.
But speaking with the FSA, the agency confirmed that temperature is also not likely to have been a factor of concern.
When asked whether we can expect a definitive answer on the source and reason behind the outbreak, Potter said: “At this stage, we don’t expect further clarification on our hypothesis that the outbreak was caused by Apollo leaves in sandwiches. Based on our extensive food chain analysis and epidemiological evidence to date, we find this conclusion to be reliable.”
Lessons to take away
The first case was reported on 15 May 2024, with cases peaking on 20 May and the presence of STEC identified on 24 May. Although the product recall was not implemented until 14 June, it was considered a speedy response as there needs to be several cases reported before an outbreak is confirmed.
Taynor noted that the UKHSA “was quick to undertake epidemiological tracing, which involves surveying the infected individuals for what foodstuffs they have consumed, when and where”.
He continued: “This is further complicated by the incubation time from ingestion to appearance of symptoms being up to several days. Individuals may not reliably remember every food item they have consumed days in the past. To make the connection between the relatively low number of cases and a single food item was done as rapidly as can be reasonably expected.
“It is important to note that although the recall was implemented, testing at the related food producers had all returned negative for STEC by this point, and the outbreak was already coming to an end irrespective of the recall. The implication is that only a small amount of lettuce was contaminated, most likely from a single source, and so the contaminated sandwiches had already left stores well before the recall. It is not reasonable to expect a recall to have been called fast enough to significantly reduce cases, given infections peaked four days before the presence of STEC had even been identified, let alone the source of contaminated food.”
Winship was less flattering about the recall, describing it as “nonsense”.
Because of the short shelf life of sandwiches, “it’s very unlikely that the products being recalled would have had a problem attached to them”.
“The speed at which those products come in, are made-up, and then consumed is very short,” he continued – a mere few days. “It doesn’t leave a lot of time for assessing every item that comes in. It’s very much reliant on good quality suppliers, doing testing and making sure their systems operate.
“[The sandwich industry] recognises the risks involved in having a very such a short shelf life, [and] they put – and always have – a huge amount of investment into food safety.”
He continued: “They [recalls] work in terms of longer shelf-life products but actually in terms of short shelf-life products, we need to either move very rapidly if something is identified – and that’s not very practical – or we need to have systems that don’t do the damage that a recall does when it’s not necessary.
“We understand that the FSA has a duty to protect consumers and our members would be very keen to make sure that that continues to be the case. But at the same time, we need to recognise that the short shelf life of some chilled foods makes it really difficult to deal with that. There’s no point in in creating an issue if there is not a risk anymore.
“We’re trying to sort out with the FSA how those sorts of incidents get handled in the future.”
For Winship, the biggest concern is the limited number of environmental health officers with knowledge of chilled foods available to audit.
The limited knowledge, particularly in chilled foods, of how the chilled food industry operates can create some conflicts.
Jim Winship, director - British Sandwich Association
“We’re looking at putting additional auditing of ingredient suppliers in place to try and make sure that those their systems are sufficiently robust as our industry is reliant on ingredients coming into them in a safe condition, whether it’s meat or salad.
“As we are assembling products that are largely pre-prepared, we need suppliers to do everything they should be doing to make sure that those ingredients are 100% safe before they even reach the factories.”
Potter also told Food Manufacture that work is underway to identify lessons for the future.
“There are a number of well-established mechanisms for industry and government to work together, including working groups as well as direct liaison at regular intervals with major food businesses. These mechanisms proved successful again in the most recent response,” said Potter.
“As with all emergency responses, there’s always scope for learning…
“Work is ongoing at industry level, to identify options available to enhance monitoring and controls, with a view to reducing further similar serous outbreaks. It’s commonly accepted E. Coli should not be present on fresh produce, it simply should not be there, and systems are in place to prevent contamination. The effort now is focused on exploring options currently available for enhanced monitoring of the non-O157 E. Coli STEC groups.
“At the heart of this will be data sharing, to ensure there’s early information and common understanding. This is particularly important when supply chains are common amongst different businesses, each business’s pieces of the jigsaw are essential for a fuller picture.”
Hollands agreed microbiological data sharing would be a great way forward for improving food safety in the future.
“The FSA is keen for isolates to be shared with them for whole genomic sequencing that allows them to build up a map,” he said, before adding that there may be hesitation from the sector to do this for fear of ‘being caught out’ if something is found.
From Hollands’ perspective though, he doesn’t believe this will be used as some sort of honey trap, but rather a beneficial tool for all in the sector.
For Taylor, improvements will need to made through more regular testing for microbial contamination at the harvesting stage. However, he recognises this would incur additional costs and time for producers, with regulation difficult to implement beyond UK farms.
He concluded: “The exact source of contamination is yet to be firmly identified, and indeed this is extremely difficult to trace post incident.”
