Meeting 4 : Experience of E. coli O157 outbreaks on "Open" Farms in the Devolved Administrations
Held on Tuesday, 30 November 2009
Health Protection Agency
7th Floor Holborn Gate
330 High Holborn
London, WC1V 7PP
Notes of the meeting
Members:
Professor George Griffin (Chair)
Dr Meirion Evans
Mr David Eves
Ms Karen Jones
Dr Chris Low
Apologies
Professor David Strachan
Secretariat:
Dr Isobel Rosenstein
1.0 The identification and management of cases and outbreaks of VTEC in Scotland (with special reference to “Open” Farms)
1. John Cowden (Consultant Epidemiologist, Health Protection Scotland) gave an overview of the areas of VTEC activity within the remit of Health Protection Scotland, including identifying cases, clusters and common exposures, forming hypotheses, controlling current incidents and preventing/mitigating further incidents.
2. Dr Cowden noted it was important to consider other serotypes of Verocytotoxin producing E. coli as well as O157.
He also remarked there are some strains of E. coli O157 which do not seem to have VT genes or produce verocytotoxin, but cause similar clinical symptoms in humans.
3. Dr Cowden described the partnership working of Health Protection Scotland (HPS) with its major stakeholders. HPS works with the Health Protection Teams of 14 NHS Boards. The Teams consist of Consultants in Public Health Medicine and Health Protection Nurses. Partnerships with other Bodies are not regulated but there is an agreement to work together (in England it is more structured).
The Scottish E. coli O157/VTEC Reference Laboratory (SERL) is based in Edinburgh. The Reference Laboratory for other GI pathogens (Salmonella, Shigella, C. difficile) is in Glasgow. It was noted that Scotland no longer has a campylobacter reference laboratory.
HPS has an informal relationship with LA Environmental Health Departments.
The HPS also works with the Scottish Agricultural College (SAC) which is equivalent to the VLA in England on relevant public health issues. The SAC is funded by the Scottish Government to do so. The HPS also works with the Scottish Government through the Health Department, Animal Health (UK wide) Rural Department and the Health Promotion Unit of NHS Scotland and national Bodies in the same way as the Health Protection Agency.
4. In Scotland, there are 32 Local Authorities but they do not have a pivotal role in laboratory surveillance. Their EHOs however do collect information on cases once identified. The 40 or so NHS laboratories report identifications to HPA and are invited to send any presumptive E. coli O157 isolates to the Reference Laboratory which acts as a national surveillance centre for the whole of Scotland. They also send faecal specimens from cases in whom they suspect but have failed to identify VTEC.
5. There is informal networking with GPs. NHS24 (equivalent to NHS Direct in England) pick up calls from the public and could provide additional surveillance information.
6. The aspects of the Public Health etc (Scotland) Act 2008 relating to notification become active on 1 January 2010. The clinical notification will be “Clinical Syndrome due to E. coli O157 Infection” with the notifiable organism E. coli O157.
7. As health is a devolved matter, there is no necessity for the law to be identical North and South of the Border. In England (and as far as Dr Cowden knows Wales as well) the relevant clinically notifiable disease is likely to be “Presumed infectious bloody diarrhoea”. Scotland and England and Wales will consequently be out of “kilter”
However as in England, both the clinical syndrome and the organism will be notifiable in Scotland (the latter by the laboratory). HPS has been producing non statutory guidance on the notification process but like Health Protection Agency has no statutory power in this area.
8. Dr Cowden described the HPS enhanced surveillance for VTEC. HPS has the responsibility to ensure data collected by HPS feeds into policy. The enhanced surveillance is the surveillance of both individual cases and outbreaks. Once an outbreak is recognised it additionally becomes the unit of surveillance for “ObSurv”, the outbreak surveillance system which collects a standard minimum dataset on all outbreaks involving more than one household. Having said that, because of their rarity and importance, HPA tends to collect fuller and more detailed information on outbreaks of VTEC than on other less important pathogens.
9. Data on laboratory isolates of E. coli O157 for Scotland were presented and compared with England, N. Ireland and Wales.
Incidence rates for Scotland are consistently higher than for the rest of the UK. It is unclear why. Although Scotland is more rural this cannot be the whole explanation, as both rural and urban rates are higher in Scotland. It could, however, be some aspect of rurality which is commoner in Scotland. There are, for example, more areas provided by private water supplies in rural Scotland than in the rest of rural UK. This speculative explanation would not however account for the higher number of urban cases of E. coli O157 in Scotland compared with England. While better ascertainment in Scotland might be an explanation, there is no good evidence for this, and rates for other bacterial GI pathogens such as Salmonella and Campylobacter are broadly the same in Scotland and England.
10. Other theoretically possible, though untested, reasons for the differences are wider contact tracing in Scotland or better detection. It is noted that diagnostic laboratories in Scotland send negative faecal specimens from suspicious cases to the Reference Laboratory as well as any presumptive isolates. The faecal specimens are particularly important if initial cultures are negative and especially if the patient is presenting with bloody diarrhoea – SERL will do a direct PCR on the faecal sample.
11. The Scottish strategy for management of E. coli O157 changed following the central Scotland outbreak in 1966, with mandatory inspections of butchers premises and licensing. The Scottish E. coli Task Force produced a series of recommendations in 2001. Price, Waterhouse Cooper are currently reviewing the recommendations of the Task Force to ensure they have been implemented effectively.
12. In Scotland, once a local outbreak has been declared, the Outbreak Control Team (OCT) is chaired by the Consultant in Public Health Medicine (employed by the NHS Board). It is usual but not compulsory for HPS to provide a member of the OCT. If a national outbreak is declared, the OCT would be chaired by an HPS consultant. The LA would field a member of the OCT in any local outbreak, with the EHO visiting the farm if appropriate. The NHS Board and the LA have a partnership (similar to the situation when the PHLS was in existence, the CCDC was employed by the Trust). NHS Boards do not work under a common management and include a Director of Public Health in the Management Team who also covers non infectious aspects as part of their remit.
13. Before an OCT is called officially, there may be a ‘watching’ or ‘orange light’ period where the situation is monitored by a ‘Problem Assessment Group’. It is not as formal as an OCT but enables updates to be assessed and collective decisions to be made. Outbreaks may be defined on the basis of microbiological or epidemiological links, i.e. infection with indistinguishable strains or common exposures. If the common exposure is a visit to an “Open” Farm, say, then it may be the source even if the strains differ, as farm animals may be carrying a number of different strains.
14. The HSE Guidance was discussed, noting this was national guidance and therefore used in Scotland and promoted by HPS. Dr Cowden noted the problems in deciding when open farms were implicated, and – when implicated – deciding on criteria for closure and subsequent re opening.
15. Dr Cowden referred to Guidance for the Public Heath Management of Infection with Verotoxigenic Escherichia coli (VTEC) developed for the Health Protection Network Steering Group by Health Protection Scotland.
A Guidance Development Group chaired by Dr Cowden provided oversight to the development of the Guidance produced in 2008. This Guidance is currently due for review. It was noted that the Scottish Guidance had been presented to the Health Protection Agency Gastrointestinal Diseases Programme Board with a view to it being adapted for use in England.
The Guidance would be circulated to the Investigation Committee for their information.
Action: Secretariat
16. Specific reference was made to the Section in the Guidance on the management of suspected sources and vehicles, and livestock and their environment in particular. An analogy was drawn between the problems with enforcement at a butchers premise with closure of an “Open” Farm i.e. at a poorly managed butchers premise such as Barrs, enforcement stopped the butcher selling RTE Food (i.e. cooked sliced meat) but not raw meat (because the raw meat was “allowed” to contain E. coli O157). A farm could not be closed on the basis that E. coli O157 was present because the cattle must be assumed to be carrying it regardless of the other circumstances at the farm.
17. Parental responsibility has to be a consideration in exactly the same way as there is a responsibility to cook meat properly in the home environment.
However there has to be a responsibility by regulatory authorities to ensure that parents are adequately informed. HPA has for some years circulated an HPS/SAC leaflet on reducing livestock related risk to rural communities and organisations such as the scouts.
It was noted that the Health Protection Agency had produced a new leaflet on E. coli O157 and Farms for the public. It was unclear how widely it had been circulated although it was available from the Health Protection Agency website.
Action: Secretariat to circulate to the Committee
18. Reference was made to two research projects, firstly a collaborative study between HPS and Glasgow University Veterinary School demonstrating that cattle and human population density appeared to be related to prevalence of E.coli O157 but that the mechanism is complex. Secondly a national case control study of sporadic E.coli O157 in 2001 which showed that sporadic infection comprised 80% of cases in Scotland and that contact with farm animal faeces (not pets) was the risk factor most strongly associated with sporadic infection.
19. Dr Cowden concluded by reference to his and Mary Locking’s recently published editorial in the BMJ (which had been circulated to the Committee prior to the meeting).
It was noted that GPs have a key role in being alert to cases or outbreaks and in advising patients on personal hygiene to prevent secondary spread. Given that risks on Farms cannot be eliminated only minimised, the message of scrupulous hygiene must be emphasised especially since there are no criteria which have been agreed to prompt action.
2.0 The identification and Management of Cases and Outbreaks of VTEC in N. Ireland and the role of Health Protection
20. Lorraine Doherty (Assistant Director of Public Health, Health Protection for the Public Health Agency, N. Ireland) provided an overview of E.coli O157 in N. Ireland.
21. Previously there were 19 Hospital Trusts in Northern Ireland which were amalgamated into 5 Health and Social Care Trusts on 1st April 2008. Similarly the four Health Boards were amalgamated into 2 organisations on 1st April 2009 – the Health and Social Care Board and the Public Health Agency. The Health and Social Care Board in N. Ireland is equivalent to the English Strategic Health Authority. The Public Health Agency is a single Body providing public health functions in Health Protection, Health Development, Service Commissioning and Screening and R&D. The Health Protection Service is based in the PHA bringing together CCDCs, Surveillance Staff and Health Protection Nurses. The Microbiology and Virology Laboratories however are not included although closely linked.
22. Once the routine laboratory isolates a presumptive E. coli O157, the laboratory telephones the CCDC who then informs the district council based EHO and a questionnaire is gone through with the patient. The questionnaire data is shared with the CCDC and a decision is taken on what public health action is required.
23. Dr Doherty referred to the N. Ireland Strategy for the Surveillance, Prevention and Control of E. coli O157 (which had been circulated to the Committee prior to the meeting). It had been published by the N. Ireland E. coli O157 Taskforce in September 2006 following the recognition of the problem with E. coli O157 and is currently being reviewed.
24. The largest reported outbreak in N. Ireland was in 2008 with 17 cases; 10 children were less than 6 years old. Two cases were hospitalised, 1 with HUS. The primary cases were associated with a visit to an “Open” Farm – contact was with goats. There were a number of cases where human to human secondary transmission occurred.
25. Dr Doherty described the progress of the outbreak and the way it was managed. The first case was a 4 year old child who attended a nursery where animals from an “Open” Farm had been brought in for the children to ‘pet’.
It was decided to close the nursery (although this would not be normal practice) in view of the visit from the “Open” Farm. There was a very good relationship between the Public Health Agency and the Education Department who appreciate the advice from the Agency. There was extensive contact tracing and further cases were identified including pupils from a second nursery who had visited the first nursery for the session with the animals. The Belfast Education and Library Board undertook a lot of public advice and involvement with the parents.
26. Visits to and advice for the “Open” Farm was provided by the Public Health Agency as well as discussions with the Department of Agriculture and Rural Development (DARD), the private Vet (who undertook initial animal testing) and HSE (NI).
27. The Farm voluntarily suspended visits to schools and enhanced hygiene measures. The HSE inspected the Farm and found it very satisfactory (in terms of signage and washing facilities). The District Council also inspected (food hygiene).
It was noted that the inspection protocol and relationships between HSE and LA is different from that in England with HSE inspecting both “Open” and commercial farms.
It was also noted it was necessary to establish a relationship with a private Vet in order to do the animal testing at the Farm.
A goat which had been brought to the nursery was found to be carrying E. coli O157.
28. The outbreak continued with a cousin of an earlier case being affected. She attended another nursery and had been symptomatic at the nursery. This prompted a multi disciplinary risk assessment, the outcome of which was extensive contact tracing and closure of the second nursery (for consistency). Further cases were identified.
29. An adult who had visited the “Open” Farm was then found to have E. coli O157. The son of the adult patient who attended another nursery was found to be symptomatic. There was again extensive contact tracing with further cases identified. This nursery was not closed but thoroughly cleaned over a weekend to minimise disruption.
30. Another child who had attended a party at the same “Open” Farm was found to have E. coli O157 which prompted further contact tracing with further cases identified. There was no evidence of poor practices at the Farm.
31. The next case was a child who had visited the Farm as part of a school party (3 different schools were involved).
Extensive contact tracing revealed further cases.
32. In view of the case associated with a school visit, the Farm closed voluntarily for 2 days following a formal multi professional visit.
A number of actions were advised by HSE NI which involved young animals, rotation, cleaning etc.
The Committee requested further information on these actions.
Action: Dr Doherty to provide further details
Farm workers were tested (they were negative).
It was noted that although the Farm had previously been advised to remove goats from the public area, they had not done so.
Following the final inspection, the goats were permanently removed.
33. All positive isolates of E.coli O157 were sent to the Reference Laboratory at Centre for Infections for typing which demonstrated a single strain (type 31) from all human cases and the goat.
34. Following the outbreak, consideration was given to how well the outbreak investigation and subsequent management had been undertaken.
There was good inter sectoral working, a robust assessment of the risk with timely action taken.
A challenge was the interface with the Veterinary Service and the need for a private Vet to undertake the sampling. Also a challenge was the isolation and identification of the farm animals and the visit and review of the “Open” Farm environment.
35. There were a number of recommendations following the outbreak including the need for arrangements to be in place for Veterinary input and risk assessment, jointly with DARD and HSE NI.
Also for guidance for use by EHOs in E. coli O157 outbreaks and for guidance for the educational sector.
Dr Doherty was unsure whether HSE produced different standards and guidance for N. Ireland in relation to Farms and would follow this up.
Action: Dr Doherty
36. Further work was needed to be able to link clinical cases of HUS with laboratory E. coli O157 data.
There is a need to establish a NI E. coli O157 register, agree arrangements for farm visits by Vets/DARD and for sampling and further clarify risks and liaison arrangements for outbreaks.
3.0 The Identification and Management of Cases and Outbreaks of VTEC in Wales
37. Dr Gwen Lowe (Consultant in Communicable Disease Control, National Public Health Service for Wales) gave an overview of VTEC management in Wales.
38. In Wales, Local Health Protection Teams are broadly equivalent to the Health Protection Agency HPUs.
There are 5 (2 combined for acute issues – so 4 operational teams) across Wales, 1 based in Cardiff with the Communicable Disease Surveillance Centre which provides specialist epidemiological support across Wales. The Microbiology Laboratories are based in NHS premises but are part of NPHS Wales management. The wider NPHS includes Local Public Health Directors, Registrars, Specialists in Communication, Public Health, the Pharmacy Team and Library Service; all able to provide help during outbreak investigations. The Local Public Health Directors and Registrars are trained in health protection, hence there is provision for surge capacity and a consistency of approach across NPHS Wales.
39. It was noted that NPHS Wales is currently undergoing a re organisation and is now part of Public Health Wales NHS Trust.
40. An on call rota is in place, those on call cover the whole of Wales (population 3 million) out of hours. This comprises a Specialist Registrar, a Public Health Consultant, a CCDC and a Regional Epidemiologist. The Specialist Registrar is not on call at night.
If a case of E.coli O157 is confirmed, information about previous cases across Wales can be accessed through a secure network although the amount of detail is variable.
41. Information on cases is shared at the weekly teleconference attended by each of the 5 Health Protection Teams and CDSC, where links may be made on cases initially thought to be sporadic.
42. There are strong links with other key Agencies, particularly Local Health Boards. It was noted that the 22 current Boards are to be amalgamated to 7.
A draft multi Agency generic outbreak plan has just been developed and is expected to be ratified in the next few months by the Welsh Assembly Government.
43. Dr Lowe described the process for handling a single VTEC case.
As soon as a presumptive diagnosis is made by the local laboratory the details are telephoned to the Health Protection Team. Urgent action is required of the LA EHO for follow up. The EHO will contact the case and complete the questionnaire. It was noted that there are two versions of the questionnaire in circulation. Initially, on notification a standard form is used. If an outbreak is declared an enhanced questionnaire may be developed which is specific to the particular circumstances of that outbreak. In general the EHO will usually visit any premises which are suspect (biologically plausible) visited by the case during the incubation period.
A judgement is used on visits where several premises are mentioned in the questionnaire, in that not all may be visited, only the most likely. In normal hours, the EHOs take the lead on undertaking the actions. Out of hours, whilst the EHOs would still usually take the lead, not all LAs have on call rotas and if an appropriate EHO cannot be located the on call team led by the CCDC would take the lead out of hours.
44. Dr Lowe then described the O157 outbreak in 2005 and the subsequent investigation. There were 157 cases, 118 laboratory confirmed (73% school age) 44 schools had primary cases. 31 were admitted to hospital and there was a death (5 years old). 10 were transferred to England for dialysis.
45. The outbreak first presented on 16th September when the Microbiologist telephoned the Local Health Protection Team with 2 cases of presumptive E.coli O157.
The same morning, an Infection Control Nurse telephoned to advise there were 8 cases of bloody diarrhoea seen in the same hospital. All were primary school aged children from different schools.
46. The first OCT meeting was held on the afternoon of 16th September.
Dr Low explained that one definition of an outbreak would be “two or more linked cases”. However it is recognised that in practice this is a fluid definition and judgement also has to be exercised.
Often if there was a suspicion or concern about a potential outbreak or issue an “incident meeting” is held first (similar to Scotland’s “Problem Assessment Group”). This can be convened at very short notice and is more flexible in the very early stages (when it is not obvious if a serious problem exists) than calling a formal OCT meeting. An incident meeting can be converted into an OCT meeting at any stage and is a multiagency meeting to consider the facts and decide on further action.
In this case an incident meeting was called, an outbreak was declared and the incident meeting converted immediately into an OCT meeting. An OCT was therefore established and the first cases interviewed that evening (with the standard form).
47. A second OCT meeting was held the following day (Saturday) and the first adult case presented (school meals supervisor).
A third OCT was held on Sunday morning, new cases presented throughout the weekend.
The only common feature for all the cases was the school environment. At this stage, fruit and water had been ruled out, milk may still have been a possible cause. But school dinners were the most likely source. Because of the need to urgently initiate a case control study, staff from around NPHS Wales were called in.
By Sunday afternoon, the link with school dinners was established – the association was with meat supplied as ready cooked and sliced, not re heated, with gravy on the top.
48. The cooked sliced meat came from a common source.
Follow up microbiological sampling and analytical epidemiological investigations continued to confirm this as the source after control measures were implemented.
49. At this stage, a decision had to be made on the continued provision of school meals. A decision based on the evidence to date was to continue with providing school meals but with the removal of meat. This was made at the fourth OCT meeting held on the Sunday afternoon.
A helpline was set up which picked up cases in other LA areas.
There was also proactive involvement with the media. This was seen as a public health benefit to enable information to be given to the public.
Having several OCT meetings in quick succession and working closely enabled these decisions to be made.
50. Early outbreak control therefore was very successful. The OCT membership was broad and consisted of 33 regular members (core and co opted) and 39 others who attended at least one meeting.
By communicating frequently and information shared with roles and responsibilities for each organisation explicitly agreed, the subsequent public inquiry found that ‘But for the quality of the analysis and control measures the outbreak would have been considerably more severe and prolonged’.
51. The meeting then discussed training of EHOs in Wales.
The initial degree is usually a BSc in Environmental Science followed by on going in house training on epidemiological methods, outbreak management and case control studies.
A Masters in Public Health (at Cardiff University) is available as day release from work.
52. In Scotland, it is the Health Protection Nurses who are most likely to be in contact with cases. In house training is available through courses run by the Royal Environmental Health Institute of Scotland.