An official website of the United States government.
The principal aim of the study was to estimate the number of cases of gastroenteritis,
or intestinal infectious diseases (IID), occurring in the population of England, and find
out how many people with IID consulted their general practitioners (GPs).<P>
We sought to identify as many as possible of the disease-causing organisms, or
pathogens, responsible for IID. We then compared our estimate of the actual
number of cases of IID in the population of England and presenting to their GPs, and
the pathogens responsible for illness, with the routine national surveillance data
from laboratory reports to the Public Health Laboratory Service (PHLS)
Communicable Disease Surveillance Centre (CDSC). We also set out to identify the
factors which might lead to IID, and the costs which might result.
Because it is impossible to separate out with any precision or reliability those cases
of IID which result from food poisoning and those cases resulting from other causes,
our study necessarily addressed all cases of IID and not merely the cases caused by
eating contaminated food. We therefore included in our study cases infected with
pathogens known to be spread predominantly from person to person, and cases with
pathogens usually held responsible for food poisoning, as well as those cases who,
although suffering from IID according to our definition, had no pathogen found in their
stools. Our definition of IID was: any person with loose stools or significant vomiting
lasting less than two weeks, in the absence of a known non-infectious cause and
preceded by a symptom-free period of three weeks. Vomiting was considered
significant if it occurred more than once in a 24-hour period and if it incapacitated the
case or was accompanied by other symptoms such as cramps or fever. <P>
The study attempted to estimate the accuracy of laboratory reporting to the PHLS
and CDSC; it did not attempt to determine the accuracy of national food poisoning
statistics, which depend upon statutory notifications by doctors on the basis of
clinical suspicion.<P>
The specific objectives of the study were:
<UL> <LI> To estimate the number and aetiology of cases of IID in the population,
presenting to GPs, and having stool specimens sent routinely for laboratory
examination.
<LI> To compare these numbers and the aetiologies with those recorded by the
national laboratory reporting surveillance system.
<LI> To estimate the prevalence of asymptomatic infection with agents associated
with IID.
<LI> To document differences between cases of IID (in the population and presenting
to GPs) and similar but well people (controls).
<LI> To estimate the socio-economic burden of IID and its distribution.
RESULTS:
We estimated that 20% of the population of England suffered IID in a year, and
3% of the population presented themselves to their GP.<BR>
This means that nine and a half million cases of IID occur annually, of which one and a
half million present to their GPs. Half a million have stools sent for microbiological
examination. In our study, despite using extensive microbiological testing, no target
organism was found in about two thirds of cases in the community, and nearly half
those presenting to the GP. In normal practice, a much greater proportion of ‘negative
stools’ is reported: over three-quarters of the stools submitted in the enumeration
component of this study, which observed normal practice, were negative. This may
be due to a number of factors: the diarrhoea may be non-infectious, or due to a
pathogen which cannot be identified, or which is no longer present in sufficient
numbers to be detectable, or whose identification is difficult, or not attempted.
Viruses, almost half of which are SRSV (Small Round Structured Viruses), account
for about 16% of cases of IID in the community. Yersinia and Aeromonas are almost
as common, or more common, in controls as cases, and their clinical significance is
therefore unclear. If these are excluded from calculations, viruses are as common as
bacteria in association with IID in the community in cases where a target organism
was identified. In cases presenting to GPs, viruses were detected in over 20%, with
rotavirus accounting for one third of these. Bacteria are, however, much commoner:
excluding Yersinia and Aeromonas, nearly 40% of cases presenting to GPs have a
bacterial pathogen identified.
<P>
We estimated by our direct method that for every 136 cases of IID in the
community, 23 presented to a GP, 6.2 had a stool sent routinely for microbiological
examination; 1.4 had a positive result; and one was reported to the PHLS’ CDSC.<BR>
The ratio varies according to the organism. Approximately three cases of
salmonellosis, a predominantly foodborne disease, occur in the community for
every one reported to PHLS CDSC, whereas as many as 1,500 or more cases of
SRSV infection, which is often spread from person to person, may occur for every
one reported to the PHLS CDSC.
Put in another way: for every 1,000 cases of IID in the community, 160 presented to
their GP, 45 had a stool sent routinely for microbiological examination, 10 had a
positive result, and 7 were reported to PHLS, CDSC.<P>
We estimated by our indirect method the ratio to be 88 cases in the community
to every one reported to PHLS CDSC. <BR>
This ratio is lower than the ratio of 1:136 calculated by our direct method and this
suggests that the indirect method may underestimate the community rate in relation to
cases reported in the national data. This would occur if, as we suspect, national
surveillance tends to over-represent the proportion of cases which are part of
outbreaks. In other words, in estimating the ratio by this method, the national
surveillance system’s limitations in identifying apparently sporadic cases are partially
offset by its greater efficiency in identifying cases which are part of outbreaks.<P>
We found many differences between cases and controls.<BR>
When analysed by each of the enteropathogenic organisms, social factors and
crowding, travel abroad, and bottle-feeding of infants were associated with an
increased risk of IID. We also found that cases of infection with almost all organisms
are consistently less likely than controls to have consumed certain foods (pulses,
salads and rice prepared at home, fruit, pasteurised dairy products and fish) in the
previous ten days. This may have arisen from the study design but we can find no
evidence for this. It may therefore be a true association. We believe further research
is warranted to confirm or refute this observation, as, if it is a true association, it may
have implications for the prevention of IID.
<P>
We found the consumption of very few specific foods to be associated with an
increased risk of suffering from IID. <BR>
There are a number of possible reasons for this, including the fact that most of our
cases suffered from infection with organisms spread predominantly from person to
person. A second explanation is that the time period we asked about — ten days
prior to the onset of illness in cases — was too long to allow us to discriminate
sufficiently between cases and controls, i.e., over that time period so many controls
would also have eaten common foods that there was no difference between them
and cases. If these explanations do not fully explain the lack of positive
associations, a third explanation is that current understanding, based as it is almost
completely on either the investigation of outbreaks or cases sufficiently ill to present
to their GP, is not applicable to sporadic cases. The absence of an association
between IID and, for example, the consumption of chicken in the home, in our study
is indeed true of the mild, sporadic cases which constitute most of the burden of
illness which occurs. This may be because such cases are linked to lightly
contaminated foods, possibly as a result of cross-contamination from more heavily
contaminated products.
<P>
We estimated the average cost of a case of IID, whatever its cause, in England
to have been £79 at 1993–1995 prices.<BR>About 36% of this cost falls to the NHS, 8% is a direct cost to the case and 55% is
the cost to employers in lost production by the case or a carer. The average cost of a
case presenting to a GP is £250; the average cost of a case presenting to a GP with
Salmonella, a predominantly foodborne organism, is £606, and the average cost of a
case presenting to a GP with SRSV, which is often spread from person to person, is
£176. We estimated that IID in England cost at least three-quarters of a billion
pounds a year. Cases presenting to their GP account for over half of this total. We
found that cases presenting to the GP are ill for an average of 8.6 days. A quarter of
these cases had symptoms persisting three weeks after illness.<P>
The final report, "<a href="http://www.foodbase.org.uk/results.php?f_report_id=320" target="_new">A Report of the Study of Infectious Intestinal Disease in England</A>" is available at Foodbase, an open access repository of the <acronym title="Food Standards Agency"> FSA</acronym>.<P>
Find more about this project and other FSA food safety-related projects at the <a href="http://www.food.gov.uk/science/research/" target="_blank">Food Standards Agency Research webpage</a>.