Outbreak of Non-O157 Shiga Toxin-Producing Escherichia coli Infection from Consumption of Beef Sausage
- Steen Ethelberg1,2,
- Birgitte Smith1,
- Mia Torpdahl1,
- Morten Lisby3,
- Jeppe Boel4,
- Tenna Jensen5,
- Eva Møller Nielsen1, and
- Kåre Mølbak2
- 1Departments of Bacteriology, Mycology, and Parasitology, Copenhagen, Denmark
- 2Epidemiology, Statens Serum Institut, Copenhagen, Denmark
- 3Regional Veterinary and Food Control Authority East, Copenhagen, Denmark
- 4National Food Institute, Danish Technical University, Copenhagen, Denmark
- 5Danish Veterinary and Food Administration, Copenhagen, Denmark
- Reprints or correspondence: Dr. Steen Ethelberg, Dept. of Epidemiology, Statens Serum Institut, Artillerivej 5, DK-2300 Copenhagen, Denmark (set{at}ssi.dk).
Abstract
We describe an outbreak of Shiga toxin-producing Escherichia coli O26:H11 infection in 20 patients (median age, 2 years). The source of the infection was an organic fermented beef sausage. The source was discovered by using credit card information to obtain and compare customer transaction records from the computer systems of supermarkets.
Shiga toxin-producing Escherichia coli (STEC) is an enteric pathogen that causes diarrhea, often bloody. In 2%–10% of registered cases, predominantly occurring in children, the infection leads to hemolytic uremic syndrome (HUS), a serious condition characterized by renal failure [1, 2]. The STECs that belong to serogroup O26 are the second most frequently isolated (after O157) in the United States, Asia, and Europe and have in recent years emerged as one of the most frequent causes of HUS (second to O157) [3, 4]. In Denmark during 1997–2003, O157 constituted 26% of human isolates, and O26 constituted 16% [5]. In a number of countries, including Germany [6, 7], the United Kingdom [8], Ireland [9], Austria [10], Italy [11], and Japan [12–15], infections caused by O26, several involving patients with HUS, have been described, but in most instances, the source of these infections remains unknown.
An outbreak of STEC O26:H11 infection was recognized at the Danish Statens Serum Institut on 9 March 2007 by routine PFGE typing of STEC isolates. An investigation was launched by the Danish Statens Serum Institut in cooperation with the Danish Food Safety Authority and the Danish Food Institute.
Methods. In Denmark, STEC infections are notifiable clinically and by laboratory. Strains are sent to the Danish Statens Serum Institut for further characterization, which includes confirmation by the Vero cell assay [16], O:H serotyping [17], and stx gene subtyping [18]. In addition, weekly PFGE profiling is performed as an outbreak detection tool [19]. All of the primary clinical microbiology laboratories where the infections were successfully diagnosed used the same method, relying on PCR analysis for the presence of STEC genes of suspected colonies grown from stool samples [20].
Cases were defined as laboratory-confirmed infection with STEC O26:H11 with the outbreak PFGE pattern detected from 1 February through 1 May 2007. Patients were interviewed using a standard trawling questionnaire; this was followed by a consumer record analysis. Parents were asked to access their home banking systems and to locate all supermarket purchases made during the 3 weeks before the onset of disease. This task was possible if the purchases had been paid for with a credit or debit card. Parents were then asked to e-mail a home bank screen-shot that contained a list of type and location of supermarket, the date when the purchases were made, and the exact amount withdrawn from the accounts. The 2 supermarket chains that were reported by the parents to be used most often agreed to assist in the investigation and searched their central computers for the precise amount paid in combination with the date and the location of the shop. Therefore, lists (including bar code numbers) of all individual items bought on each occasion that the parents shopped in the supermarkets could be produced. The purchases were entered into a database and compared, and food items bought by >1 family were identified and evaluated.
A case-control study was conducted during 28–30 March 2007 with use of telephone interviews. Data on control subjects, matched by birthday, sex, and municipality, were obtained from the Danish population register. The food exposures of concern included various types of meat and chicken, cold cuts, sausages, milk, cheeses, fruits, and vegetables. Data were analyzed by conditional logistic regression with a commercially available software program (SAS; SAS Institute). International alerts were made through PulseNet Europe on 12 March 2007, Enter-Net (the European Union dedicated network for Salmonella infection and Verocytotoxin-producing E. coli infection [21]) on 16 March 2007, and Eurosurveillance on 31 May 2007 [22].
Result. The outbreak comprised 20 laboratory-confirmed cases, all but 2 of which occurred in young children. The median age of the patients was 2 years (range, 0–51 years), and 3 cases occurred in the same family. The outbreak was confined to Denmark (figure 1), except for 1 case that occurred in the Faroe Islands. Overall, the reported symptoms were mild, and there were no cases of HUS and only 1 case involving bloody diarrhea. Several patients had stool samples that tested positive for a second diarrheal pathogen (2 Campylobacter species, 2 Yersinia enterocolitica, 1 norovirus, and 2 eae -positive [stx -negative] E. coli ). Therefore, the date of onset of symptoms was uncertain in some cases; figure 2 shows the week of reporting of cases. The infection strain was serotype O26:H11 and carried the eae and stx >1 genes but not the stx >2 gene. The strain was resistant to sulfonamide and streptomycin and had varying resistance to ampicillin.
Figure 1
Distribution of 19 cases in an outbreak of Escherichia coli O26 infection in Denmark in 2007, by municipality
Figure 2
Number of reported cases of Escherichia coli O26:H11 infection in an outbreak in 2007, by week of reporting. Primary cases are shown in dark shading, and presumed secondary cases are shown in light shading. There were 20 total cases.
Hypothesis-generating interviews of parents of patients did not lead to a clear hypothesis. A week into the investigation, it was therefore decided to try the consumer record method. Seven families agreed to participate. Five families were found to have bought the same type and brand of product, an organic fermented beef sausage, from supermarket A. No other specific products likely to be the source of an STEC infection had been bought by >2 families. Two cases in the sixth family were subsequently associated with the sausage through shopping records from the kindergarten that the 2 children in the family (both of them case patients) attended. Supermarket A reported that the sausage in question was not part of the standard product line but that a single batch had been marketed in all shops beginning on 29 January 2007.
After the sausage had been found on the shopping lists of the third family, a case-control study was initiated. At this point, the sausage was no longer being sold in supermarket A; therefore, questions concerning the general preference of buying sausages of a similar type (e.g., beef sausage and organic sausage) were included. The analysis comprised 14 case patients and 40 matched control subjects; selected results are shown in table 1. Consumption of beef sausage was associated with a matched OR of 21. Several other food exposures, most notably grapes and peppers, were also associated with disease; however, in a multivariable matched analysis of consumption of beef sausage, grapes, and peppers, only beef sausage was associated with disease (matched OR, 2.8; 95% CI, 1.4–170). When parents were directly questioned about whether they remembered buying the specific sausage under suspicion, 5 case parents and 1 control parent answered yes. On 2 April 2007, the sausage was recalled from the market, and the production facility was inspected. The contaminated batch consisted of 19,080 sausages, ∼16,000 of which had been delivered to supermarket A. At recall, all shops reported that the sausage was sold out. However, 4 sausages, 2 of which were in unopened packages and 2 of which had been partly eaten, were recovered from private homes. These sausages were all found to be positive for the infection strain and had a PFGE profile that matched that of the infection strain. In addition, remains of the batch of imported frozen beef used to make the sausages were sampled and subsequently tested positive for the infection strain.
Table 1
Selected results of case-control analysis of an outbreak of Shiga toxin-producing Escherichia coli O26 infection in Denmark in 2007.
Discussion. The combined epidemiological and microbiological evidence revealed that a batch of a particular beef sausage was the source of this outbreak. First, the analysis of supermarket purchases strongly indicated that the sausage was the source. Second, the case-control analysis implicated the sausage. Third, the infection strain was isolated from the sausages and beef used for their production.
In this outbreak, we were able to use information kept in the computer systems of 2 major supermarket chains and obtain detailed lists of what the parents of the case patients had bought before their child became ill. This information proved to be a strong tool in the investigation, and most likely, the source of the outbreak would not have been found without the use of this method. Interviews were often conducted several weeks after exposure, and many parents were not sure of the exact date of onset of symptoms. It was therefore generally difficult for the parents to describe in detail what their children had eaten before becoming ill. None of the 5 parents who, according to supermarket shopping records, had bought the sausage stated this information in the first round of hypothesis-generating interviews. To our knowledge, this method has not been used before but has since been used successfully in the investigation of an outbreak of STEC infection in Iceland [23].
To our knowledge, this is only the second recognized general outbreak of STEC infection in Denmark. The first occurred in 2005 and was also linked to an organic product. It was caused by O157:H− and was associated with pasteurized organic milk from a particular dairy [24]. Non-heat-treated sausages, however, appear to pose a particular food safety hazard. In recent years, infections caused by lightly preserved sausages that contained meat from ruminants have also been described in neighboring Scandinavian countries; in Sweden in 2002, an outbreak of O157:H7 infection was traced to a beef sausage [25], and in Norway in 2006, an outbreak of O103 STEC infection that included 11 cases of HUS was caused by a lamb-meat sausage [26].
In conclusion, the present outbreak demonstrates that STEC O26, a non-O157 STEC, may be present in beef products and can cause foodborne infections. It is therefore potentially problematic if human diagnostic efforts are only directed toward O157 and if quality control programs in the production of beef products are directed toward the detection of O157 only and are not sensitive to the detection of non-O157 serogroups.
Acknowledgments
We thank Charlotte Kjelsø, Steen Villumsen, and Mette Selchau, for help with case-control interviews, Flemming Scheutz and Katharina Olsen, for serotyping and virulence assessments, and public health officers and microbiologists in Århus and Frederiksborg counties, for case notifications.
Potential conflicts of interest. All authors: no conflicts.
- Received July 11, 2008.
- Accepted December 8, 2008.
- © 2009 by the Infectious Diseases Society of America
References
This Article
- AbstractFree
- » Full Text (HTML)Free
- Full Text (PDF)Free
Classifications
-
- Electronic Articles
Services
Citing Articles
Google Scholar
PubMed
Related Content
Share
- Email this article
-
CiteULike
Connotea
Delicious
Facebook
Google+
Mendeley
Twitter
Published on behalf of
IDSA Members: For your free access to this journal, log in via the IDSA members area.
The Journal
Impact Factor: 8.186
5-Yr impact factor: 7.898
For The Media
For Authors
- Services for authors
- Self-archiving Policy
- IDSA Conflict of Interest policy
- Instructions for Authors
- ICMJE Form
- Submit Now!
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy
Corporate Services
Alerting Services
Related IDSA Publications
