original reports
A birthday party, home-made ice cream, and
an outbreak of Salmonella enteritidis phage
type 6 infection
H Dodhia, J Kearney, F Warburton
Summary: An outbreak of food poisoning caused by Salmonella enteritidis
Key Words:
phage type (PT) 6 followed a children’s birthday party. Thirty of 37 (attack
dairy products
rate 81%) attendees were ill, with diarrhoea and fever being the most common
disease outbreaks
symptoms. One child required hospital admission for rehydration. No deaths
eggs
occurred. Stool specimens from 24 out of 25 people who submitted them were
ice cream
positive for S. enteritidis PT 6 and one was positive for S. enteritidis PT4.
Epidemiological investigation indicated that home-made ice cream was the vehicle salmonella food poisoning
of infection. Fresh shell eggs used raw in the ice cream were the likeliest source of
infection. No ice cream or eggs were available for microbiological analysis.
S. enteritidis PT6 accounts for a small but increasing proportion of all
salmonellas in humans and its epidemiology is thought to be similar to S.
enteritidis PT4. Thus current recommendations on the use of raw eggs in
cooking continue to apply.
Communicable Disease and Public Health 1998:1, 31-4.
General outbreaks (affecting members of more than
one private residence or residents of an institution) of
infectious intestinal disease associated with eating ice
cream are uncommon in England and Wales. Two out
of 20 general outbreaks associated with the
consumption of milk and dairy products between 1992
and 1996 were reported as associated with ice cream1.
Both were due to S. enteritidis PT4; one as a result of
cross contamination of ice cream stored at the wrong
temperature and the other as the result of using raw
shell eggs in the manufacture of home-made ice cream.
In this paper we report an outbreak of S. enteritidis
PT6 food poisoning associated with eating home-made
ice cream at a birthday party in North London in
October 1996. Seven adults and 30 children, mainly
aged 4 or 5 years, attended the party. Thirty people
developed symptoms of gastroenteritis within 24
hours of the party.
The commonest salmonella reported to the
PHLS Communicable Disease Surveillance Centre is
S. enteritidis PT4 – 13 184 cases of infection were
reported in 1996 and 15 457 (provisional) in 1997 –
which accounts for most of the outbreaks of
salmonellosis. S. enteritidis PT6 was the second
commonest phage type of S. enteritidis, with 989
reports in 1996 and a peak between July and
November. Laboratory reports for 1997 showed an
unusual spring peak (figure)2.
COMMUNICABLE DISEASE AND PUBLIC HEALTH
Methods
Epidemiology
A retrospective cohort study was undertaken. The
outbreak control team defined a primary case as: any
person who attended the birthday party and
developed one or more of diarrhoea, vomiting,
abdominal cramps, and fever in the next seven days.
Parents of all the children and adults who attended
the party were interviewed by telephone using a
standard questionnaire a week after the party. They
were asked about the complete list of food items on
the menu, symptoms, and the date of onset, and if any
FIGURE Laboratory reports of Salmonella enteritidis PT6:
England and Wales, January 1995 to May 1997
350
Number of laboratory reports
Introduction
1997
1996
1995
300
250
200
150
100
50
0
Jan
Feb Mar Apr May Jun
Jul Aug Sep Oct Nov Dec
Month of report
VOL 1 NO 1 MARCH 1998
31
original reports
TABLE 1 The food specific attack rates for items served at the birthday party
Ate food
Did not eat food
Ill
Not ill
Attack
rate (%)
Ill
Not ill
Attack
rate (%)
Relative
risk
95% confidence
interval
p value*
Sausages
25
3
89
5
2
71
1.25
(0.77 to 2.03)
0.255
Potato cake
22
1
96
8
4
67
1.43
(0.95 to 2.16)
0.0375
Sweet corn
16
1
94
14
4
78
1.21
(0.92 to 1.56)
0.3376
Cucumber
20
3
87
9
2
82
1.06
(0.77 to 1.46)
1.00
Ice cream
30
2
94
0
3
0
–
(12.04 to ∞)
0.00152
Cornet
25
1
96
4
4
50
1.92
(0.96 to 3.86)
0.00674
Flake
24
1
96
5
4
56
1.73
(0.96 to 3.12)
0.01177
†
* p value calculated using Fisher’s exact test
† not estimable
other members of their household had developed
similar symptoms. The hostess was asked if she or any
of the other food handlers had been ill in the week
before the party.
Data were entered twice to minimise errors. Singlevariable analysis was carried out in Epi Info version
6.04 3 . Relative risks (RRs) with 95% confidence
intervals were calculated. Multivariable analysis was
carried out using logistic regression in the statistical
package GLIM4. Food items that showed any sign of
an association with illness (p<0.5) from single-variable
analysis were examined to find their independent
effects after adjusting for the effects of the other food
items.
Environmental
The kitchen and the food preparation area were
inspected. The only food eaten at the party available
for microbiological examination was an unopened
carton of a proprietary whip topping. Samples of a
brand of potato cake and vegetarian sausages were
obtained from the shop where the products were
originally bought. These were sent for analysis to the
Central Middlesex Public Health Laboratory and the
PHLS Food Hygiene Laboratory.
Environmental
Ice cream was the only home-made item. It was
prepared the previous day in one batch, using raw
shell eggs, sugar, and a proprietary whip topping, and
stored overnight in a domestic freezer. A chocolate
flake was added to the ice cream on the day of the
party. The freezer was working normally. The hostess
could not remember where she had bought the eggs,
and none was left for microbiological examination. All
the ice cream, in cones with chocolate flakes, was eaten
TABLE 2 Multivariable analysis of food items of interest
from single variable analysis
Microbiology
Stool specimens obtained for microbiological
examination were examined in several hospital
microbiology laboratories. Isolates were sent to the
PHLS Laboratory of Enteric Pathogens (LEP) for
confirmation and phage typing.
Sausages
Potato cake
Results
Epidemiology
Thirty people met the primary case definition (attack
rate 81%). Twenty-seven cases had diarrhoea, 25 fever,
21 vomiting, and nine abdominal cramps, starting
within 24 hours and lasting two to three days. One
child required admission to hospital for rehydration.
Attack rates for the food items eaten are shown in
tablei1. The food specific attack rates were highest
for the potato cake, sweet corn, cornet, and ice cream.
32
Single variable analysis showed that eating potato
cake, ice cream, cornet and chocolate were all
significantly associated with illness. After fitting all
the food items into a multivariable model only the ice
cream remained significantly associated with illness
(table 2). The data were reanalysed treating as cases
only those people who fitted the original definition
and from whom S. enteritidis PT6 had been isolated
(n=24). This did not alter the findings: ice cream was
still the only food significantly associated with illness
(p = 0.025).
VOL 1 NO 1 MARCH 1998
Food
item
Odds
ratio
(95% confidence
p
interval)
value*
No
Yes
1.00
2644
2.62×10 -43 to ∞
0.48
No
Yes
1.00
3.4×10 -4
0 to ∞
0.61
Sweet corn
No
Yes
1.00
5.95×10 -7
0 to ∞
0.30
Ice cream
No
Yes
1.00
5.02×10 8
0 to ∞
0.01
Cornet
No
Yes
1.00
1.03×10 8
0 to ∞
0.39
Flake
No
Yes
1.00
0.08
0 to ∞
1.00
* p value calculated using Chi-squared test for change in deviance
COMMUNICABLE DISEASE AND PUBLIC HEALTH
Microbiology
Twenty-five stool specimens were sent for
microbiological examination. Twenty-four were
positive for S. enteritidis PT6 (including one from the
hostess, who reported no symptoms of illness) and one
was positive for S. enteritidis PT4. All 24 isolates of S.
enteritidis PT6 were indistinguishable by molecular
typing (J Threlfall, personal communication).
No pathogens were found in any of the food items
sent for analysis.
Discussion
The analytical investigation supported the hypothesis
that the vehicle of infection was home-made ice cream
made with raw eggs. The attack rate was high and
the incubation period relatively short, suggesting a
large inoculum. The association of outbreaks of food
poisoning with the consumption of raw shell eggs in
uncooked products is well recognised5. It is likely that
contaminated egg(s) used to make the ice cream
caused this outbreak. Substantial levels of S. enteritidis
(≥104 organisms per millilitre) have been isolated from
the contents of clean unbroken shell eggs 6-10, a dose
capable of causing illness11. Only one in 12 000 eggs is
estimated to contain >104 salmonellas per millilitre of
contents and the risk to any one individual is also
considered low, but there is significant public health risk
if eggs contaminated at low levels are stored for
prolonged periods at room temperature10.
It is possible that the ice cream could have been
cross contaminated during preparation from
contamination on the surface of the raw shell egg, and
that salmonellas could have multiplied afterwards.
Multiplication would have been unlikely during the
short time when the ice cream was prepared, however,
and while it was frozen or chilled. The egg surface
would have needed to have been heavily
contaminated, and this would probably have been
associated with contamination by other faecal and
environmental organisms6.
An egg survey undertaken by the PHLS has shown
that 0.7% of UK eggs from high street retail outlets
are contaminated with S. enteritidis 12. The survey did
not differentiate between contamination of egg
contents and shell. A further survey showed a
prevalence of S. enteritidis at any concentration
(including <104/ml, which is unlikely to cause illness)
of 0.2% (about 0.1% from content and 0.1% from
surface contamination) 10. Most of the salmonellas
isolated in both surveys were S. enteritidis PT4. S.
enteritidis PT6 was not isolated from eggs sampled in
COMMUNICABLE DISEASE AND PUBLIC HEALTH
either of these surveys. The results of the most recent
egg survey, carried out in 1995, are awaited.
S. enteritidis PT6 has been reported as a cause of
outbreaks of food poisoning and is nearly always
associated with raw eggs or chicken (P Wall, personal
communication). It is a pathogen of poultry and its
epidemiology is believed to be similar to that of S.
enteritidis PT4. There are two published reports of
outbreaks caused by S. enteritidis PT6, one a variant
(S. enteritidis PT6b) associated with lemon meringue
prepared using fresh shell eggs and the other
associated with egg sandwiches 13,14.
Many recipe books and magazines include recipes
for home-made ice creams that require little or no
heating and contain fresh shell eggs. It is also possible
that the lack of recent publicity about salmonella and
eggs has led the public to think that the problem no
longer exists. In a recent Office of National Statistics
omnibus survey of over 10 000 randomly selected
respondents, 96% had heard of a salmonella warning
but only 35% said that they tried to avoid raw or soft
cooked eggs15. This outbreak is further evidence of
the importance of the chief medical officers’ warning
to avoid the use of uncooked or partially cooked eggs
and to use pasteurised egg products instead5,16. This
is particularly important for vulnerable groups, such
as children and elderly people. This outbreak
illustrates that the control of salmonella infection is
important at all levels of the food chain, from infection
in poultry and egg production, to the duration and
temperature at which eggs are stored, and the use of
raw eggs in uncooked foods by vulnerable groups of
people.
original reports
at the party. Other food items were sent for
examination – whip topping, potato cake, and
vegetarian sausages.
Other foods were prepared on the day of the party
in the kitchen, mainly by the hosts, with some help
from relations. Commercially bought food items were
prepared and cooked according to the manufacturers’
instructions. None of the food handlers reported
having been ill in the week before the party.
Acknowledgements
Dr M Susman (consultant in communicable disease
control, Barnet Health Authority), Mr R Phillips, Ms
P Sharpe, Mr I Kimmett, Mrs J Wrigley
(Environmental Health London Borough of Barnet),
Dr J Threlfall, Ms L Ward (LEP).
References
1. Djuretic T, Wall P, Nichols G. General outbreaks of
infectious intestinal disease associated with milk and dairy
products in England and Wales: 1992 to 1996. Commun
Dis Rep CDR Review 1997; 7: R41-5.
2. CDSC. Salmonella infections: monthly report. Commun
Dis Rep CDR Wkly 1997; 7: 208.
3. Dean AG, Dean JA, Coulombier D, Burton AH, Brendel
KA, Smith DC, et al. Epi Info, Version 6: a word processing,
data base and statistics program for epidemiology on
microcomputers. Atlanta: CDC, 1994.
4. Francis B, Green M, Payne C. The GLIM system: release 4
manual. Oxford: Clarendon Press, 1993.
5. Advisory Committee on the Microbiological Saftey of Food.
Report on salmonella in eggs. London: HMSO, 1993.
6. Morgan D, Mawer SL, Harman PL. The role of home-made
ice cream as a vehicle of Salmonella enteritidis phage type
4 infection from fresh shell eggs. Epidemiol Infect 1994;
113: 21-9.
7. Humphrey TJ, Whitehead A, Gawler AHL, Henley A, Rowe
B. Numbers of Salmonella enteritidis in the contents of
naturally contaminated hen’s eggs. Epidemiol Infect 1991;
106: 489-96.
VOL 1 NO 1 MARCH 1998
33
34
original reports
8. Mawer SL, Spain GE, Rowe B. Salmonella enteritidis phage
type 4 and hen’s eggs. Lancet 1989 ; i: 280.
9. Humphrey TJ, Cruickshank JG, Rowe B. Salmonella
enteritidis phage type 4 and hen’s eggs. Lancet 1989; i: 281.
10. PHLS Food Surveillance Group. Salmonella contamination of
stored hens’ eggs. PHLS Microbiology Digest 1994: 11; 203-5.
11. Benenson AS. (editor). Control of communicable diseases
manual. 16th edition. Washington DC: American Public
Health Association, 1995.
12. de Louvois J. Salmonella contamination of eggs. Lancet
1993; 342: 366-7.
13. Brugha RF, Howard AJ, Thomas GR, Parry R, Ward LR,
Palmer SR Chaos under canvas: A Salmonella enteritidis
PT 6B outbreak. Epidemiol Infect 1995; 115: 513-17.
14. Holtby I, Tebbutt GM, Grunert E, Lyle HJ, Stenson MP.
Outbreak of Samonella enteritidis phage type 6 infection
associated with food items provided at a buffet meal.
Commun Dis Rep CDR Rev 1997; 7: R87-90.
15. Walker A. Food safety in the home: a survey of public
awareness. London: ONS/HMSO, 1996.
16. Chief Medical Officer. Raw shell eggs. London: Department
of Health, 1988. (EL/98/P136).
VOL 1 NO 1 MARCH 1998
H Dodhia
senior registrar in public health medicine
Bromley Health Authority
J Kearney
specialist registrar in public health medicine
Barnet Health Authority
F Warburton
statistician
PHLS Statistics Unit
Address for correspondence:
Dr Hiten Dodhia
Bromley Health Authority
Global House
10 Station Approach
Hayes
Kent BR2 7EH
tel: 0181 315 8315
fax: 0181 315 8356
email: [email protected]
COMMUNICABLE DISEASE AND PUBLIC HEALTH