This article has Open Peer Review reports available.
Lessons learned from a textbook outbreak: EHEC-O157:H7 infections associated with the consumption of raw meat products, June 2012, Limburg, Belgium
- Toon Braeye†1Email author,
- Sarah Denayer†2,
- Klara De Rauw3,
- Anmarie Forier4,
- Jurgen Verluyten5,
- Ludo Fourie5,
- Katelijne Dierick2,
- Nadine Botteldoorn2,
- Sophie Quoilin1,
- Pascale Cosse5,
- Jeannine Noyen5 and
- Denis Pierard3
© Braeye et al.; licensee BioMed Central. 2014
Received: 29 April 2014
Accepted: 9 September 2014
Published: 15 December 2014
On 5 June 2012 several enterohemorrhagic Escherichia coli, EHEC, O157:H7 infections were reported to the public health authorities of Limburg.
We performed a case-control study, a trace back/forward investigation and compared strains isolated from human cases and food samples. A case was defined as anyone with a laboratory-confirmed E. coli O157:H7-infection in North-East Limburg from May 30 2012 till July 15 2012. Family members with bloody diarrhea were also included as cases. E. coli O157 was isolated by culture and the presence of the virulence genes was verified using (q)PCR. Isolates were genotyped and compared by Pulsed Field Gel Electrophoresis (PFGE) and insertion sequence 629-printing (IS629-printing).
The outbreak involved 24 cases, of which 17 were laboratory-confirmed. Five cases developed Hemolytic Uremic Syndrome (HUS) and fifteen were hospitalized. Cases reported a significantly higher consumption of “steak tartare”, a raw meat product (OR 48.12; 95% CI; 5.62- 416.01). Cases were also more likely to buy meat-products at certain butcheries (OR 11.67; 95% CI; 1.41 - 96.49). PFGE and IS629-printing demonstrated that the vtx1a vtx2a eae ehxA positive EHEC O157:H7 strains isolated from three meat products and all seventeen human stool samples were identical. In a slaughterhouse, identified by the trace-back investigation, a carcass infected with a different EHEC strain was found and confiscated.
We present a well described and effectively investigated foodborne outbreak associated with meat products. Our main recommendations are the facilitation and acceleration of the outbreak detection and the development of a communication plan to reaches all persons at risk.
Foodborne diseases, Shiga-toxigenic Escherichia coli, Enterohemorrhagic Escherichia coli, Meat products, Case control studies, Electrophoresis, Gel, Pulsed-Field
KeywordsEnterohaemorrhagic E. coli – EHEC Escherichia coli O157 Food-borne infections Vero cytotoxin-producing E. coli – VTEC Outbreaks Case control studies PFGE IS629-printing
Enterohemorrhagic Escherichia coli, EHEC, (also called Verotoxin-producing E. coli, VTEC, or Shiga-like toxin-producing E. coli, STEC, associated with hemorrhagic colitis) is known to easily spread via food and cause infections with symptoms ranging from mild watery diarrhea to hemorrhagic colitis and hemolytic uremic syndrome, HUS . In 2011, 12 EU countries reported 63 foodborne outbreaks, caused by EHEC strains . The most commonly associated serogroup is O157 [3, 4]. The first published large outbreak with E. coli O157:H7 occurred in 1982 in the United States and involved contaminated hamburgers. The transmission of E. coli O157 occurs primarily by ingestion of inadequate processed contaminated food or water, but also through contact with manure or animals at farms and petting zoos [5, 6]. Person to person transmission has been described in daycare centers, nurseries and certain institutions .
In Belgium, governmental departments, laboratories and the scientific institute of public health, IPH, are collaborating for the prevention and management of outbreaks. Three National Reference Laboratories, NRL, for VTEC are officially recognized; one for human samples (National Reference Center, NRC), one for food samples (National Reference Laboratory for Foodborne Outbreaks ,NRL-FBO) and one for veterinary samples. The IPH hosts the NRL-FBO, organizes the network of the human reference centers and offers epidemiological support during outbreaks . The Federal Agency For the Safety of the Food Chain, FASFC, is concerned with the safety along the food chain; hygiene inspection, sampling and it manages the system for tracing food products back and forward . Notification of foodborne outbreaks to the local public health authority is mandatory in the Flemish region of Belgium; two or more cases of foodborne disease, linked or possibly linked, should be reported by clinicians or laboratories.
E. coli O157 was first described in Belgium in 1987. From 1994 to 2011, between 47 and 103 EHEC cases were registered yearly in Belgium. The most frequently detected serotype was O157, 24.1% of the EHEC positive stools was E. coli O157:H7/H-[10, 11]. The NRC detected an association between E. coli 0157 and HUS. In 2011 twelve EHEC strains were detected in HUS cases, ten of these were serotyped as E. coli O157:H7. In 2012 all O157-isolates tested positive for the virulence factors intimin (eae) and enterohemolysin (ehxA). Intimin is associated with the formation of “attachment-effacement”-lesions and enterohemolysin is a plasmid-encoded toxin.
Previously two Belgian outbreaks with a microbiologically confirmed link between human cases and the origin of the EHEC infection have been published. One outbreak in 2006 with only 2 cases, of which none developed HUS, could be linked to contact with farm animals . A larger outbreak in 2007 with 12 cases, of which five developed HUS, could be linked to ice cream contaminated with E. coli O26 and E. coli O145 .
On June 5th 2012 two cases diagnosed with an EHEC-infection were reported to the department of public health of Limburg. One day earlier another case had already been reported. We describe the further course of this outbreak and the investigation towards the source.
Case finding & exploratory interviews
The outbreak was detected and the cases were identified through the system of mandatory reporting. The public health authority contacted local general practitioners and hospitals to inform them on the outbreak and ask them to report additional cases. The public health authority also contacted all cases telephonically for an interview. Information on personal characteristics (age, gender), family size, symptoms in family members, possible sources, diet history, food preferences and in which shops they recently bought food products was collected.
A case in the case-control study was defined as anyone with a laboratory-confirmed E. coli O157:H7 infection in the North-East Limburg region from May 30 2012 until June 14 2012, the start of the case-control study. Out of the 13 laboratory-confirmed cases known by the start of the case-control study, we randomly selected eleven cases and invited them to an electronic survey. The two cases that were not selected, provided the information necessary to compose a list of suspicious supermarkets for the survey. Two hundred persons living in the direct neighbourhood of the cases were invited to the same electronic survey. A standard survey is available for case-control studies in foodborne outbreaks. As the exploratory interviews, conducted by the start of the case-control study, pointed towards meat products, we shortened this survey by excluding specific questions about diary, vegetable and fruit consumption. The final survey included 13 meat products, three dairy products (including ice-cream) and six other food products (vegetables and sprouts). The survey also included questions on outdoor dining at a restaurant or snack bar and shopping at a supermarket in a list of five supermarkets. Finally the survey gathered information on symptoms and the onset date of these symptoms. The persons who completed the survey and did not report any symptoms were considered controls.
In the univariate analysis we calculated unadjusted odds ratios. Logistic regression was used for multivariate analysis. Variables attaining a significance level of p < 0.2 in univariate analysis were included in multivariate analysis. Overall, a significance level of p < 0.05 was set as statistically significant. Analysis were performed using SAS software (version 9.3, SAS Institute Inc., Cary, NC, USA).
In Belgium, beef meat was traceable back and forward through the use of unique “Sanitel-numbers” and data recorded in a computerized management system. Cattle could be followed through an individual identification number from farm through slaughterhouse, cutting plant, butcheries into retail. Once suspicion had risen on a meat product, samples were taken from the complete production and distribution chain associated with this product. All data concerning sampling is encoded in the computer system ‘Foodnet’ of FASFC. Samples were analyzed at the NRL-FBO IPH. Human samples were further analyzed at the NRC UZBrussel.
For the detection of EHEC in stool, an in-house multiplex PCR was used on cultures of clinical samples. Initial typing of EHEC isolates was done using biochemical tests, in-house PCR and agglutination. Subtyping of the vtx genes was done using a PCR method described by Scheutz et al. .
For the detection in and isolation of EHEC from meat and carcass samples, three detection and/or isolation methods were used in parallel, including the recently published ISO/TS 13136:2012 for the detection of EHEC, ISO 16654:2001 for the detection and isolation of E. coli O157 and the VIDAS® E. coli O157 system (BioMerieux) according to the manufacturer’s instructions. Strains were further confirmed for the presence of verocytotoxin genes (vtx1 and vtx2), the intimin gene (eae) and the enterohemolysin gene (ehxA) using the GeneDiscCycler Instruments (Pall Genesystems, France) according to the manufacturer’s instructions.
E. coli O157:H7 isolated from stool samples of cases was compared to E. coli O157:H7 isolated from food products and a carcass swab using pulsed field gel electrophoresis (PFGE) and insertion sequence 629-printing (IS629-printing). IS629-printing is a rapid genotyping method based on the variable presence of insertion-sequence 629 throughout the E. coli O157:H7 genome described by Ooka et al. . Pulsed field gel electrophoresis was conducted as described in the PulseNet protocol for E. coli O157.
Course of the outbreak
A first case was reported to the department of public health of Limburg on June 4. On June 5 two other cases from a single household were reported. By June 6 seven cases had been reported. Two days later, June 8, the FASFC conducted a first investigation. Samples from meat products were taken in shops frequently mentioned during exploratory interviews with cases. The meat products were traced back to a slaughterhouse and cutting plant, where more samples were taken. The first laboratory results were available on June 14. The case-control study was set-up on June 13 and the invitations to fill in the online survey were distributed on June 14. The results of the case-control study were available on June 18.
Results from the case-control study, Limburg, Belgium, 2012
Lower 95% CI
Upper 95% CI
Results of sampling
Food samples were taken at the butcheries of five different supermarkets, one meat establishment, one cutting plant and one slaughterhouse. On June 14, the NRL-FBO confirmed the presence of vtx1a vtx2a eae ehxA positive E. coli O157 in two samples of raw beef meat and “steak tartare” collected on June 8. Sampling continued until July 4. A total of 73 samples were analyzed by the NRL-FBO. In total, three vtx1a vtx2a eae ehxA positive E. coli O157 and one vtx1a vtx2c eae ehxA positive E. coli O157 could be isolated. Ten other samples were suspected for the presence of the epidemic strain at the screening level but the strain could not be isolated due to the presence of a high number of background flora. Five other samples were suspected for the presence of EHEC but the detected virulence genes were different from the ones of the epidemic strain.
The management of a foodborne outbreak involves several steps. The outbreak needs to be detected as early as possible, followed by a timely identification of the suspected food vehicle and finally resulting in adequate control measures.
1) Early detection
In Belgium, early identification of an outbreak depends on the alertness and subsequent reporting by clinicians and labs. As also illustrated by this outbreak, a higher alertness is justified during warmer periods. Preoccupied customers, leaving their products in the car, and malfunctioning refrigerators will allow for E. coli growth during these periods. E. coli grows from 8°C, with an optimal growth at 37°C. Even at 5°C the pathogen remains hazardous .
The first cases developed symptoms on May 30. A first case was reported on June 4. On June 8 the first investigation towards the source of the outbreak started. We detected the outbreak relatively fast as compared to other outbreaks, e.g. the E. coli O104:H4 outbreak in Germany was only detected two weeks after the outbreak started . Further shortening these intervals will require that clinicians report faster. Public health authorities should focus on an easy and accessible notification system and increase the overall awareness of clinicians about foodborne outbreaks by e.g. frequently reporting about such outbreaks. Outbreak detection by laboratories will be facilitated by the development of new techniques and methods. Molecular information, collected by IS629-printing of every EHEC O157-strain found in clinical samples, is brought into a national database which allows for early detection of outbreaks even in geographically and temporally less obviously associated cases. A database on both national and European level based on PFGE information is in development for EHEC-strains isolated from food.
E. coli is a continuously evolving species. New and emerging EHEC-clones will belong to both serogroup O157 and non-O157 serogroups. With the publication of ISO/TS 13136 in November 2012, a standardized method for the detection in foodstuff of other EHEC-serogroups (O26, O103, O111, O145) is available and used at the NRL-FBO. For any vtx-positive stool sample multiplex PCR is used to detect EHEC. Peripheral labs may not utilize these more advanced and intensive techniques to detect EHEC. This emphasizes the role of reference labs.
2) Timely identification of the suspected food vehicle
On June 14 the first EHEC-positive meat products were identified. Results from the case-control study were available from June 18. This is 14 and 19 days after the first case experienced symptoms and five and ten days after the first investigation started. Public health authorities communicated for the first time about possible sources of this outbreak 15 days after its onset. The median time to communication by governmental sources for European outbreaks is 31 days . In this outbreak, the main delay was not due to the investigation itself, but due to the time till the detection of the outbreak and the start of the investigation. For the timely identification of the suspected food vehicle, we benefited from the textbook character of this outbreak; contaminated bovine-derived products are the most common source of infection, 75% of EHEC outbreaks are linked to their consumption and a previous Belgian survey showed a prevalence of 0.73% for STEC O157 in beef [19, 20].
A case control study allows for a wide search, it can indicate any item from a survey as possible source, and it can be used to direct microbiological investigations. Studies have been inconclusive about the response representativeness of electronic surveys versus mail surveys, but since an electronic survey is both cheap and fast (in this study there were five days between the set-up of the study and the results) and our population probably had high internet usage, we opted for an electronic survey . After the authorities communicated about the outbreak, the first microbiological results were published in the media . This information was communicated while cases and controls could still fill in the survey, this will likely have interfered with the case-control study. A possible solution to avoid such a possible bias is to use a virtual cohort. This method is based on sales records. It calculates relative risks based on the amount of a certain product sold to cases and the amount sold to persons not identified as cases .
A strain identical to the one isolated from human samples was isolated from three meat products. We did not isolate this specific strain from any other sample taken during the trace-back investigation, but we did isolate another EHEC O157 strain from a carcass swab. No environmental sampling was performed in the slaughterhouse, though this could have aided in finding the specific cause (e.g. contaminated during the evisceration process, contaminated equipment and/or uncontrolled environmental conditions) of this outbreak.
3) Control measures and secondary effects
As soon as the presence of vtx1a vtx2a eae ehxA positive E. coli in meat products was confirmed, the FASFC performed a trace-back investigation using the system of Sanitel numbers. All batches of contaminated beef meat still in retail were put under seizure and destroyed. The butcheries and slaughterhouse were temporarily closed and thoroughly cleaned and disinfected. After the disinfection, butcheries at supermarkets and the slaughterhouse restarted their activities at a slower pace and under close monitoring. The personnel of these establishments was reminded about hand hygiene, cross-contamination and the inspection of animals. They also received a new self-checklist.
Posters and leaflets were distributed at the affected supermarkets. National media coverage of the outbreak aided in publicizing the source of the outbreak and lowering the public anxiety. Still three new cases were reported after national news media reported on the outbreak. These cases kept consuming meat products bought during the outbreak period which were stored in the freezer or refrigerator.
Out of our 24 cases, 5 cases (21%) developed HUS. This amount is relatively high compared to other described EHEC outbreaks, though O157:H7 infections are known to have the strongest association with HUS worldwide . It is known that the use of conventional antibiotics exacerbates Shiga toxin-mediated cytotoxicity [25, 26]. The three children who developed HUS did however not receive antibiotics.
We described a foodborne outbreak, involving 24 cases of which 5 developed HUS, in which the source could be traced back to a slaughterhouse. PFGE and IS629-printing proved that beef products, sold at different supermarkets, were the source of this EHEC O157:H7 outbreak. Contaminated food products and a contaminated carcass were removed. There was a nine days delay between the start of the outbreak and the start of the investigation. Detecting an outbreak as fast as possible and reporting the results to all concerned are the greatest challenges towards significantly improving outbreak management in Belgium. A lot of different authorities, institutes and laboratories, each with its own mandate, are involved in outbreak investigation and management. An outbreak in which the source is proven and subsequently halted showed the effectiveness of the management of foodborne outbreaks in Belgium.
We would like to thank the teams of the public health authorities involved in this outbreak, especially Chantal Dehollogne and Sofie Th eunis from the local team of the agency of care and health. We would like the physicians, local laboratories and other health care providers for their collaboration.
- Karmali MA, Petric M, Lim C, Fleming PC, Steele BT: Escherichia coli cytotoxin, haemolytic-uraemic syndrome, and haemorrhagic colitis. Lancet. 1983, 2: 1299-1300.View ArticlePubMedGoogle Scholar
- European Food Safety Authority, European Centre for Disease Prevention and Control: The European Union Summary Report on Trends and Sources of Zoonoses, Zoonotic Agents and Food-borne Outbreaks in 2011. http://www.efsa.europa.eu/en/efsajournal/pub/3129.htm,
- Wells JG, Davis BR, Wachsmuth IK, Riley LW, Remis RS, Sokolow R, Morris GK: Laboratory investigation of hemorrhagic colitis outbreaks associated with a rare Escherichia coli serotype. J Clin Microbiol. 1983, 18: 512-520.PubMedPubMed CentralGoogle Scholar
- European Centre for Disease Prevention and Control (ECDC): Annual Epidemiological Report. Reporting on 2010 Surveillance Data and 2011 Epidemic Intelligence Data. 2013, Stockholm: ECDCGoogle Scholar
- Ferens WA, Hovde CJ: Escherichia coli O157:H7: animal reservoir and sources of human infection. Foodborne Pathog Dis. 2011, 8: 465-487. 10.1089/fpd.2010.0673.View ArticlePubMedPubMed CentralGoogle Scholar
- Heuvelink AE, van Heerwaarden C, Zwartkruis-Nahuis JTM, van Oosterom R, Edink K, van Duynhoven YTHP, de Boer E: Escherichia coli O157 infection associated with a petting zoo. Epidemiol Infect. 2002, 129: 295-302.View ArticlePubMedPubMed CentralGoogle Scholar
- Pennington H: Escherichia coli O157. Lancet. 2010, 376: 1428-1435. 10.1016/S0140-6736(10)60963-4.View ArticlePubMedGoogle Scholar
- Muyldermans G, Gaëtan M, Litzroth A, Amber L, Ducoffre G, Geneviève D, Quoilin S, Sophie Q, Medical-Technical Advisory Board: Establishment and reinforcement of the national reference centers for human microbiology in Belgium. Arch Public Health. 2012, 70: 16-10.1186/0778-7367-70-16.View ArticlePubMedPubMed CentralGoogle Scholar
- Maudoux JP, Saegerman C, Rettigner C, Houins G, Van Huffel X, Berkvens D: Food safety surveillance through a risk based control programme: approach employed by the Belgian Federal Agency for the Safety of the Food Chain. Vet Q. 2006, 28: 140-154. 10.1080/01652176.2006.9695220.View ArticlePubMedGoogle Scholar
- Robaeys G, Surmont I, Lemmens P, Coremans G, Vantrappen G, Vandepitte J: Haemorrhagic colitis and verotoxin-producing Escherichia coli O157 in Belgium. Lancet. 1987, 1: 1495-1496.View ArticlePubMedGoogle Scholar
- Buvens G, De Gheldre Y, Dediste A, de Moreau A-I, Mascart G, Simon A, Allemeersch D, Scheutz F, Lauwers S, Piérard D: Incidence and virulence determinants of verocytotoxin-producing Escherichia coli infections in the Brussels-Capital Region, Belgium, in 2008-2010. J Clin Microbiol. 2012, 50: 1336-1345. 10.1128/JCM.05317-11.View ArticlePubMedPubMed CentralGoogle Scholar
- Van Den Branden D, De Schrijver K: Een cluster van E. coli O157: H7-infecties met een hemolytisch uremisch syndroom na een verblijf op een vakantiehoeve. Vlaams infectieziekte bulletin. 2007, 60: 2-Google Scholar
- De Schrijver K, Buvens G, Possé B, Van den Branden D, Oosterlynck O, De Zutter L, Eilers K, Piérard D, Dierick K, Van Damme-Lombaerts R, Lauwers C, Jacobs R: Outbreak of verocytotoxin-producing E. coli O145 and O26 infections associated with the consumption of ice cream produced at a farm, Belgium. Euro Surveill. 2007, 2008: 13-Google Scholar
- Scheutz F, Teel LD, Beutin L, Piérard D, Buvens G, Karch H, Mellmann A, Caprioli A, Tozzoli R, Morabito S, Strockbine NA, Melton-Celsa AR, Sanchez M, Persson S, O’Brien AD: Multicenter evaluation of a sequence-based protocol for subtyping Shiga toxins and standardizing Stx nomenclature. J Clin Microbiol. 2012, 50: 2951-2963. 10.1128/JCM.00860-12.View ArticlePubMedPubMed CentralGoogle Scholar
- Ooka T, Terajima J, Kusumoto M, Iguchi A, Kurokawa K, Ogura Y, Asadulghani M, Nakayama K, Murase K, Ohnishi M, Iyoda S, Watanabe H, Hayashi T: Development of a multiplex PCR-based rapid typing method for enterohemorrhagic Escherichia coli O157 strains. J Clin Microbiol. 2009, 47: 2888-2894. 10.1128/JCM.00792-09.View ArticlePubMedPubMed CentralGoogle Scholar
- Palumbo SA, Call JE, Schultz FJ, Williams AC: Minimum and maximum temperatures for growth and verotoxin production by hemorrhagic strains of Escherichia coli. Journal of Food Protection. 1995, 58: 352-356.Google Scholar
- Werber D, Krause G, Frank C, Fruth A, Flieger A, Mielke M, Schaade L, Stark K: Outbreaks of virulent diarrheagenic Escherichia coli–are we in control?. BMC Med. 2012, 10: 11-10.1186/1741-7015-10-11.View ArticlePubMedPubMed CentralGoogle Scholar
- Mondor L, Brownstein JS, Chan E, Madoff LC, Pollack MP, Buckeridge DL, Brewer TF: Timeliness of Nongovernmental versus Governmental Global Outbreak Communications. Emerg Infect Dis. 2012, 18: 1184-1187. 10.3201/eid1807.120249.View ArticlePubMedPubMed CentralGoogle Scholar
- Chahed A, Ghafir Y, China B, Dierick K, De Zutter L, Piérard D, Daube G: Survey of the contamination of foodstuffs of animal origin by Shiga toxin producing Escherichia coli serotype O157:H7 in Belgium from 1999 to 2003. Euro Surveill. 2005, 10: 33-36.PubMedGoogle Scholar
- Nguyen Y, Sperandio V: Enterohemorrhagic E. coli (EHEC) pathogenesis. Front Cell Infect Microbiol. 2012, 2: 90-View ArticlePubMedPubMed CentralGoogle Scholar
- Kuusi M, Nuorti JP, Maunula L, Miettinen I, Pesonen H, von Bonsdorff C-H: Internet use and epidemiologic investigation of gastroenteritis outbreak. Emerg Infect Dis. 2004, 10: 447-450. 10.3201/eid1003.020607.View ArticlePubMedPubMed CentralGoogle Scholar
- Woodburn MJ, Raab CA: Household food preparers’ food-safety knowledge and practices following widely publicized outbreaks of foodborne illness. Journal of Food Protection. 1997, 60: 1105-1109.Google Scholar
- McCartney G, Cowden J, Murray S, Ahmed S: The use of a new virtual cohort study design to investigate an outbreak of E. coli O157 linked to a supermarket delicatessen. Epidemiol Infect. 2010, 138: 1439-1442. 10.1017/S0950268810000257.View ArticlePubMedGoogle Scholar
- Tarr PI, Gordon CA, Chandler WL: Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome. Lancet. 2005, 365: 1073-1086.PubMedGoogle Scholar
- Safdar NSA: Risk of hemolytic uremic syndrome after antibiotic treatment of escherichia coli o157:h7 enteritis: a meta-analysis. JAMA. 2002, 288: 996-001. 10.1001/jama.288.8.996.View ArticlePubMedGoogle Scholar
- Wong CS, Mooney JC, Brandt JR, Staples AO, Jelacic S, Boster DR, Watkins SL, Tarr PI: Risk factors for the hemolytic uremic syndrome in children infected with Escherichia coli O157:H7: a multivariable analysis. Clin Infect Dis. 2012, 55: 33-41. 10.1093/cid/cis299.View ArticlePubMedPubMed CentralGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.