REPORT ON THE OUTBREAK OF
E COLI O157 IN CHILDREN ATTENDING
MACDUFF PRIMARY SCHOOL, ABERDEENSHIRE
JUNE 1999
Compiled by:
Dr Arun Mukerjee
Consultant in Public Health Medicine (CD&EH)
Grampian Health Board
Jayne Leith & Fiona Browning
Communicable Disease Nurses
Grampian Health Board
Colin Houston
Specialist Officer (Food)
Aberdeenshire Council
JULY 2000
TABLE OF CONTENTS
Chapter / Subject / Page No1 / Summary/Conclusions & Recommendations / 3
2 / Introduction / 4
3 / E coli O157 / 5
4 / Epidemiology / 6
5 / Environmental Investigations and Results / 8
6 / Microbiological Investigations and results / 9
7 / Veterinary sampling and results / 10
8 / Control measures / 11
9 / Communication / 12
References / 13
Appendices / Appendix 1– Outbreak Control Team Membership
Appendix 2 – Sample Enteric Form
Appendix 3 – Aberdeenshire Council’s Enteric Investigation Form
Appendix 4 – Letter to Parents 1 June 1999
Appendix 5 – Letter to Parents 2 June 1999
Appendix 6 – Grampian Health Board’s E Coli O157 patient information leaflet / 14
15
24
26
27
28
Chapter 1
SUMMARY
Four confirmed cases of E coli O157 were notified to Grampian Health Board’s (GHB) Communicable Disease Team and three further cases of enteric illness among children attending one class (3B) of Macduff Primary School, Banff. An outbreak was declared by the Consultant in Public Health Medicine on 2nd June 1999 and an Outbreak Control Team was convened (See Appendix 1 for Membership of Outbreak Control Team).
The potential risk factors identified by the Communicable Disease Team included a visit to Duff House, Banff for a class outing and picnic on 19th May. They had also consumed home-made goats’ cheese in class on 26th May. The Communicable Disease Team arranged for the 30 class children to be screened. 3 adults who had consumed the cheese were also tested. Environmental samples from kitchen utensils and refrigerators where the cheese had been kept were submitted for microbiological examination but no organisms where isolated. The cheese had all been consumed therefore none could be submitted for testing. A sample of raw goats’ milk was tested and grew coliforms suggesting environmental contamination. E coli O157 was not isolated from the goats milk. Faecal samples from various animals on the smallholding where the cheese was produced, including the goat, were positive for E coli O157. Both human and animal samples were phage type 21/28 and indistinguishable from each other by pulsed-field gel electrophoresis. Between 27th May and 4th June, 24 children in the class and 3 adults who ate the cheese were confirmed positive for E coli O157 (96%). The high attack rate among the cheese tasters, the shape of the epidemic curve and illness in an adult who did not visit Duff House but tasted the cheese strongly suggests that the cheese was responsible for the illness.
The outbreak was controlled by taking joint action with the Aberdeenshire Council Environmental Health & Consumer Protection Service, Aberdeenshire Council Education Service, veterinary section of Scottish Agriculture College and local General Practitioners. Epidemiological information was collected and recorded on the appropriate questionnaires (Appendices 2&3). GHB’s Communicable Disease Team regularly updated families of positive cases, health professionals and the school. Daily press updates after the outbreak control meetings were held. Secondary cases were prevented by strictly adhering to exclusion criteria for positive cases, advising about potential transmission routes of E coli O157 and emphasising the importance of hand hygiene. The outbreak was successfully controlled within a fortnight.
CONCLUSIONS AND RECOMMENDATIONS.
1. Consumption of contaminated goat’s cheese was presumed to be the most likely cause of the outbreak of E coli O157 although in the absence of microbiological evidence this could not be confirmed.
2. The risks associated with preparation, storage and service of foods entering educational establishments should be assessed.
3. More research should be focussed on immunity and susceptibility to E coli O157 infection.
Chapter 2
INTRODUCTION
Grampian Health Board’s Communicable Disease Team received notification from the Medical Microbiology Department of Grampian University Hospitals Trust on Tuesday, 1st June 1999 that a child from the Macduff area of Aberdeenshire had been admitted to Royal Aberdeen Children’s Hospital (RACH) with confirmed Escherichia coli (E coli) O157. The hospital had also admitted another child from the same area with gastro-enteritis on 31st May and a further three on 1st June. These five children reported that their symptoms started between Friday, 28th May and Sunday, 31st May. Therefore by June 1st the total number of hospitalised children from the Macduff area was 5.
The parents of the children were interviewed in hospital on 1st June at which time Grampian Health Board’s enteric surveillance forms were completed (Appendix 2). Various common risk factors were identified including a class outing and picnic, to Duff House on 19th May and a class tasting of home produced goats’ cheese on 26th May. Duff House, Banff, is a local historic house with grounds that are open to the public. The visit to Duff House was excluded as a causative factor of the gastro-enteritis because the first child did not become ill until twelve days after the visit. This is outside the usual incubation period of E coli O157. It was further noted that the families concerned tended to socialise together, and although a communal barbecue had been held on Sunday, 30th May, this was discounted as a causative factor as some pupils’ symptoms pre-dated the barbecue. During the course of the interviews the parents intimated that other children were also absent from class 3B at Macduff Primary School.
On 1st June Macduff Primary School was advised that Grampian Health Board’s Communicable Disease Team were treating this incident as a possible outbreak. The Scottish Office, Scottish Centre for Infection and Environmental Health (SCIEH), Glasgow and Grampian Health Board’s Press Team were briefed accordingly. By 2nd June four of the five children hospitalised were confirmed positive for E coli O157. At this time there were three further suspected cases in RACH. An outbreak was officially declared on 2nd June 1999 and the first meeting of the Outbreak Control Team (OCT) was convened at 2pm that afternoon at Summerfield House.
CHAPTER 3
E COLI O157
Escherichia coli (E coli) bacteria are found in the digestive tracts of most humans and animals. There are hundreds of different strains of E coli and usually these bacteria cause no harm (Pennington 1997).
The strains of E coli that are capable of producing toxins are described as Vero cytotoxin producing E coli (VTEC). One VTEC strain that is able to cause serious illness in humans is E coli O157:H7 which can be further sub-typed to assist with epidemiological investigation. The bacteria are capable of invading the gut and can produce a toxin that may affect the kidneys leading to haemolytic uraemic syndrome (HUS).
Over 15% of healthy cattle and 2% of sheep may carry E coli O157 in their gut. Animals are usually unaffected by E coli O157. However a small number may become ill. Animal faeces contaminated with E coli O157 may in turn contaminate water and soil. It is, at present, unknown how long E coli O157 can survive in particular environments but research indicates that it can survive in soil for several weeks (Maule 1999).
E coli O157 can be transmitted to humans in a variety of ways including eating under cooked meat that is already infected, consuming unpasteurised milk or milk products or contaminated water. Transmission can also occur through inadequate hygiene after contact with contaminated soil. Person to person transmission can also occur by faecal-oral route. Inadequate hand hygiene following contact with any of these vehicles can lead to the ingestion of the bacteria. E coli O157 needs very few bacteria to cause infection in a susceptible individual.
The incubation period for E coli O157 varies between 1 and 14 days, usually 3-4 days. The symptoms caused by E coli O157 can vary considerably depending on how the individual reacts to the organism. The individual may be asymptomatic but is more likely to have stomach cramps and diarrhoea. This may resolve but occasionally the disease progresses to the next stage where the diarrhoea becomes bloody. A small number of individuals may develop more serious complications including kidney failure (HUS). The disease tends to be more serious in young children and the elderly.
Diagnosis is made by culture of stool specimens and typing of the bacteria or by demonstration of rising antibody titre in blood serum. Since the majority of Scottish E coli O157 isolates share the same phage type (21/28), other typing techniques must be performed to distinguish potential outbreak samples from the background of unrelated sporadic cases occurring at the same time. Pulsed- field gel electrophoresis is a DNA fingerprinting technique that subdivides isolates within a particular phage type, thus characterising them more accurately.
Chapter 4
EPIDEMIOLOGY
The Incident
The Head Teacher of Macduff Primary School confirmed that on the 24th May one pupil brought a cheese into the school for class 3B to taste. The cheese was made from goat’s milk. It was produced on Sunday, 23rd May by the owner of the goat, stored in a refrigerator overnight and taken to school on Monday, 24th May. The cheese was kept in the school refrigerator overnight and taken home by a member of staff on Tuesday, 25th May to be kept in her domestic refrigerator. The cheese was then brought back to school on Wednesday, 26th May, for consumption by class 3B. A small amount of cheese was put onto a biscuit for each pupil. It was ascertained that not all the pupils had eaten the cheese (Table 2)
School investigation
The Head Teacher confirmed that class 3B comprised of 31 children, one of whom was on holiday throughout the incident (Table 1). She reported that 12 children (39%) were absent from Class 3B on 1st June out of a total of 31. She confirmed that this was unusually high and that no other class was similarly affected. Class lists were faxed identifying children who were absent and who had eaten the goats cheese (Table 2). It was reported that 3 children were on holiday and 2 children had tonsillitis. This left 7 children off with “tummy upsets”. 30 children were present on 26th May however only 25 tasted the cheese. Upon further inquiry it became obvious that the children started to fall ill from 27th May.
Table 1 Composition of Class 3B (All born between March 91 - February 92)
Male / Female / TotalChildren in class / 13 / 18 / 31
Table 2 Relationship between cheese tasting and infection in children of class 3B ( N=30)
Cheese tasters Non cheese tasters
Total who ate cheese / Cases positive / Attack rate / Total who refused cheese / Cases positive / Attack rate / Relative risk25 / 24 / 96% / 5 / 0 / 0% / ¥
Table 3 Relationship between the children tasting the cheese and illness (N =25)
Cheese tasters Non cheese tasters
Symptomatic / Total / Attack rate / Symptomatic / Total / Attack rate / Relative risk19 / 25 / 76% / 0 / 5 / 0% / ¥
Table 4 Relationship between cheese tasting and illness and infection in adults (N=3)
Total / Symptomatic / Positive / Attack rateAdults who ate cheese / 3 / 3 / 3 / 100%
By 4th June a total of 19 children had become ill (Table 3). In addition 3 adults, who also consumed the cheese, developed illness (Table 4). All symptomatic cases were confirmed as E coli O157 positive. One child developed haemolytic uraemic syndrome (HUS) four days after becoming ill which is within the median range of the reported interval for the onset of HUS (EuroSurveillance 2000). None among the cheesemakers household or contacts, including the pupil in class 3B, complained of illness and all were microbiologically negative.
Incubation period.
Figure 1 Date of onset of symptoms
Symptoms developed between 1-9 days after tasting the cheese in those who became ill. This is within the extended incubation period of E coli O157. An incubation period of up to 14 days has been reported in certain incidents (Marsh et al 1992, Kohli et al 1994). All but one child of class 3B visited Duff House 12 days before the first reported illness which is also within the extended incubation period of E coli O157. However, the shape of the epidemic curve (Fig I), the high attack rate among the cheese tasters and illness in one adult who did not visit Duff House but tasted the cheese strongly suggests that the cheese was responsible for the illness. This could not be confirmed by microbiological tests because a sample of cheese was not available although the goat’s faeces were positive for E coli O157.
Chapter 5
ENVIRONMENTAL INVESTIGATIONS AND RESULTS
The incident was reported to the Specialist Officer (Food) for Aberdeenshire Council on 1st June 1999.
Information was gathered on which pupils consumed packed lunches or school meals. The school kitchen was inspected and the food handlers questioned. The investigating officers were satisfied with the conditions in the school kitchen and with the answers provided by the staff.
There was no evidence that the goat had been ill and the owner stated that the teats had been cleaned with anti-bacterial wipes prior to milking. The owner stated that the goat's milk was boiled for one minute before adding the rennet to make the cheese. The milk obtained from the goat to make this cheese was all used in its production. The rennet was only used to make this specific batch of cheese. During a later interview with the producer of the cheese it was confirmed that the milk was not boiled but warmed up.
Samples of milk from the goat and the rennet used in the cheese were submitted for examination and reported as negative for E coli O157. The sample of milk did not comply with the Dairy Products (Hygiene) Regulations 1995. The milk contained a total viable count /gm at 30 degrees of 1.3 X 107 and a coliform count /gm at 30 degrees of 1.6 X 102.