Knowing when to go public in an outbreak situation is challenging. But it’s better than silence.The most important conclusions from this Public Health Wales report are:

• Issuing a proactive press release without naming the premises resulted in this decision becoming the media focus rather than the outbreak.
• In future outbreaks proactive media engagement without naming the premises should be avoided.

On the 27th May 2015 the Shared Regulatory Services Communicable Disease Team (Cardiff)  identified two cases of Campylobacter (one in Cardiff and the other in the Vale of Glamorgan) that were linked to the same premises (Premises A) in Cardiff.

This triggered an immediate investigation and an Outbreak Control Team was subsequently convened, declaring a formal outbreak on 4th June 2015.

In total there were 33 cases meeting the case definition of which 11 were microbiologically confirmed as Campylobacter jejuni. No cases were hospitalised. 24 cases ate at Premises A on 17th May. Of the remaining cases, seven ate on 16th May and one on 18th May. The final case ate on 7th June.

Repeated environmental visits were undertaken and issues that could potentially lead to cross contamination were identified. Premises A voluntarily closed on 4th June to address these issues and reopened on 6th June.

Of the 33 cases, 31 participated in a case control study.  These all ate between the 16th and 18th May.  The study revealed that 100% (31) of included cases had eaten from the salad bar compared to 84.9% (45/53) controls (p=0.024). In addition, 30 of 31 cases (96.8%) had eaten pasta salad from the salad bar, compared with 22/50 controls  (44%) (odds ratio 38; 95% CI 5.3–1611). Adjustment for other exposures using logistic regression did not materially change the association with eating pasta salad. A similar but independent association with eating noodles from the salad bar was also identified but few of the cases (6/31) had consumed noodles.

Environmental investigation found areas of non compliance with statutory food hygiene regulations and confirmed that several poor food hygiene practices had been identified that potentially could result in pasta salad cross-contamination within the kitchen area.

It was therefore concluded that eating pasta salad from the salad bar between 16th and 18th May 2015 was significantly associated with acquiring Campylobacter infection in this outbreak, and that for the small number of individuals who ate noodles this may have been independently associated with acquiring Campylobacter infection. The identification of non-meat items (often salad) in Campylobacter outbreaks is a reoccurring theme.

Ensuring good food hygiene is always the sole responsibility of the Food Business Operator. Nevertheless, it is important to note that this outbreak identified several issues which have implications more widely.

These included:

• Issues with the interpretation and implementation by the food business of the Food Standards Agency E. coli O157 Control of Cross Contamination Guidance (revised December 2014).
• Issues relating to the Primary Authority’s response in outbreak situations (relevant to Food Business Operators with multiple outlets).

The specific points of concern are explained in the discussion section of this report. 

Related to these issues, investigations highlighted three matters which may have implications for other high throughput food businesses.  These were:

• Not using physical separation as the primary control measure to prevent cross-contamination.
• An over reliance on two-stage cleaning as a control measure which may fail during busy periods.
• The need to design out (as much as possible) any potential for human error resulting in cross-contamination.

Following this outbreak, improvements with respect to these three matters have been implemented in Premises A and all other similar premises nationally that are under the same ownership.

The outbreak was declared over on 25th August 2015.

Conclusions

1. There were 33 cases of Campylobacter associated with this outbreak. Eleven were microbiologically confirmed.
2. This had the features of a point source outbreak. All but one case ate at Premises A on the weekend 16-18th May. The final confirmed case ate at the premises on 7th June.
3. Epidemiological and environmental investigation identified cross-contamination of the pasta salad as the most likely source of the outbreak for the cases on 16-18th May. No source was identified for the case on 7th June.
4. Environmental investigation found areas of non compliance with statutory Food Hygiene Regulations and confirmed that several poor food hygiene practices had been identified that potentially could result in pasta salad cross-contamination within the kitchen area.
5. The interpretation and application of the December 2014 revised version of the Food Standards Agency E. coli O157 Control of Cross Contamination Guidance by the Food Business Operator of Premises A resulted in the business not using physical separation as the primary control measure to prevent cross-contamination. This and over reliance on two-stage cleaning as a control measure was potentially not effective in preventing cross-contamination. This guidance was then used by the Food Business Operator to defend such arrangements and structural layouts as being in line with the recommendation of this guidance.
6. Implementation of some control measures in this outbreak were delayed by involvement of the Primary Authority.
7. Being unable to interview food handlers involved in this outbreak at an early stage in a structured format away from Premises A hampered outbreak investigation and control.
8. Issuing a proactive press release without naming the premises resulted in this decision becoming the media focus rather than the outbreak.

Recommendations

1. The Food Standards Agency E. coli O157 Control of Cross Contamination Guidance (revised December 2014) should be reviewed in light of the issues identified in this outbreak.
2. The Food Standards Agency should work with the Better Regulation Delivery Office to develop advice for Primary Authorities on providing timely and effective responses to outbreak investigations.
3. Proactive follow-up for example via telephone of all confirmed Campylobacter cases in Wales should be routine practice by all Local Authorities. This supports early detection of outbreaks, the application of control measures to be timely and prompt hygiene advice to be given to cases.
4. Local Authorities should ensure that they retain sufficient Environmental Health staff with Food Safety and Communicable Disease skills to be able to proactively follow up communicable disease cases and investigate suspected outbreaks.
5. Although direct poultry contact or consumption is known to be the most common source for Campylobacter infection in humans, the Food Standards Agency Campylobacter Reduction Strategy should note for consideration that outbreaks in Wales have also been linked to non meat products such as salads. This could of course in some cases represent cross contamination but they may wish to consider looking at the body of evidence from such outbreaks across the United Kingdom to inform the Strategy going forward.
6. The use of ‘Requests for Co-operation’ under health protection legislation should be considered early in outbreak investigations in order to effectively interview food handlers.
7. In future outbreaks proactive media engagement without naming the premises should be avoided.