Introduction
Listeria monocytogenes is an uncommon cause
of illness in the general population. [1] The annual incidence in European
Union countries is 2-10 cases per million population. [2] In some groups
(the immunosuppressed, neonates, pregnant women and their unborn children),
it can be an important cause of life threatening bacteraemia and meningoencephalitis.
[1,3].
Outbreaks of listeriosis have most often been related to a food source.
[4-11] Because of this risk, pregnant women in the UK are advised to avoid
certain foods, such as camembert, brie, chèvre, blue cheeses and
pâté. [12] We describe an outbreak in pregnant women that
appeared to be linked to consumption of prepacked sandwiches.
Epidemiology investigation and control measures
A cluster of five cases of listeriosis in pregnant women and/or neonates
was identified over a two month period in the autumn of 2003. These
occurred in Swindon (situated in southern England) and the nearby town
of Gloucester (approximately 45 km away). One of the patients gave
birth in Blackpool, a town in the northwest England, but usually lived
in Swindon. None of the cases was fatal. Details are given in the table
below.

Epidemiological information of cases of Listeriosis
When the two index cases were reported, interviews were carried
out by Environmental Health Officers (EHOs) using a standard
food and travel history questionnaire. The isolates of Listeria
monocytogenes were sent for typing to the Health Protection Agency’s
Food Safety Microbiology laboratory (HPA FSML) in Colindale,
London, and were found to have indistinguishable profiles (serotype,
phagetype, and genotype).
Two further cases were then detected in the Swindon area, and
so investigations to find a common source continued. A second
questionnaire was used, asking in more detail about the types
of food eaten within the three months before onset of illness.
These revealed that, apart from shopping at major supermarket
chains, the only other similarity was that three of the patients
had eaten prepacked sandwiches from a single retail outlet within
the Great Western Hospital, Swindon which they had attended for
antenatal appointments, and a fourth patient had probably eaten
them on previous antenatal appointments. This fourth case thought
she had eaten them but could not be 100% certain due to the long
time period asked in the questionnaire and difficulty remembering.
The EHOs visited the outlet and found sandwiches sold during that period
had come from two national suppliers and one local supplier. Daily temperature
records for all the refrigerators and between pack of sandwiches measurements
had been kept, and the refrigeration records were unremarkable. However,
the outlet’s contract with the local supplier had just been terminated
and these sandwiches were no longer available for purchase in the hospital.
An outbreak meeting was held and the following actions were
taken: active surveillance was initiated by alerting local Consultants
in Communicable Disease Control (CsCDC) and microbiology departments,
the outbreak was reported in the national communicable disease
epidemiological bulletin (CDR Weekly), [13] and the HPA FSML
at Colindale was contacted to find out whether any isolates with
a similar profile had recently been identified. Case 5 was notified
by the local microbiologist and, at the same time, information
was supplied by FSML that this was a similar isolate (by typing).
Healthcare workers working with pregnant women and neonates in
the Swindon area were alerted to the outbreak and the local population
was informed via the media (newspaper, radio and television coverage).
The EHOs visited the premises of the local sandwich supplier,
and samples of food and environmental swabs were taken for microbiological
testing for Listeria. A sample from a brie and cranberry sandwich
grew Listeria monocytogenes, as did environmental samples
from the premises (chopping boards, sink plug holes and cleaning
sponge). On further serotyping and molecular testing, these were
shown to be indistinguishable from blood culture isolates from
all the patients at the HPA FSML. They were all typed as serotype
1 / 2, phage type Y, Amplified Fragment Length Polymorphism (AFLP)
type III and were indistinguishable by pulsed field gel electrophoresis
(PFGE ) using Asc1, a rare profile in the UK.
This sandwich supplier voluntarily closed down in order to clean
the premises thoroughly. The EHOs also visited the supplier that
provided meat and cheese for this sandwich maker. Samples were
taken but none yielded listeria.
The hospital retail outlet was given advice about the future
purchase of sandwiches. (see discussion).
Discussion
Listeriosis is not a notifiable disease in the UK, and so it can be difficult
to recognise outbreaks early. This outbreak was detected because most
of the patients (four out of five) presented to the Great Western Hospital
in Swindon or had a link with it. A recent survey of European countries
showed that surveillance systems are in operation in 16 of the 17 countries
surveyed and that in 10 of these countries the infection is statutorily
notifiable. [2] If cases of listeriosis were made notifiable in the UK,
all known cases would be reported, which would help to detect outbreaks
where cases are scattered throughout the UK.
The incubation period for listeriosis can be long (between 3-70
days) [14] and the food questionnaires used in our outbreak investigation
had to cover a period of several weeks . The patients may therefore
have had difficulty remembering exactly what they had eaten during
this period. A link was, however, established for three of the
cases (and possibly a fourth) - these patients all remembered
eating prepacked sandwiches bought from a retail outlet in the
hospital. No link was found for the fifth case. Two previous
outbreak reports have found an association with sandwiches supplied
by external contractors within hospitals. [15,16] In Cardiff
[15], two cases of listeria septicaemia occurred in immunosuppressed
patients who were day cases in the hospital on the same day,
and the only food link found was that both had eaten commercially
prepared sandwiches supplied by the hospital. These sandwiches
were sampled and grew L.monocytogenes with serogroup,
AFLP type and phage type all indistinguishable from the patients’ isolates.
Similarly, four cases of listeriosis occurred in and around the
city of Newcastle in a two month period [16]. This outbreak was
traced back to a caterer who provided sandwiches for the hospital
shop. In our outbreak and the outbreaks in Cardiff and Newcastle,
[15,16] patients who were at risk (that is, immunocompromised
or pregnant) visited the hospital and obtained food that was
contaminated with Listeria . We consider that providers of food
to places with higher than average concentrations of people with
lowered immunity, such as hospital retail outlets, should be
made aware of the need for the highest possible standards of
food hygiene. In January 2006 new food hygiene legislation came
into force in the UK enacting EC Regulations. The new guidelines
[17] recommend that food businesses manufacturing ready-to-eat
foods, which could pose a risk to public health through the presence
or growth of L.monocytogenes, should monitor processing
areas and equipment for the presence of this organism as part
of their sampling plans. In our outbreak and in others, [10,11]
the environment was shown to be contaminated and may have led
to product contamination. The guidelines also recommend that
if the food is to be stored before consumption (that is, if it
has a shelf life) then L.monocytogenes should not exceed 100
cfu/g during this period. If this level cannot be guaranteed,
then it should be absent from 25 g when it leaves the food business
operator. We hope that these new guidelines will prevent outbreaks
such as the one described here.
In conclusion, we report an outbreak of listeriosis that occurred
in pregnant women and was associated with the consumption of
prepacked sandwiches (ready-to-eat food) from a hospital outlet.
However, recent changes in the UK food legislation, if enforced,
should diminish the risk and help prevent further cases/outbreaks
occurring in similar circumstances.
Acknowledgements
Thanks to Jim McLauchlin and Kathy Grant of the HPA FSML for providing
typing and information, and to Christina Rattigan (Maternity Unit, GWH)
and Linda Wearn (Swindon PCT) for assistance with sending out information
to healthcare workers.
|