Introduction

One of the most serious outbreaks of foodborne disease ever reported in the European Union was a verotoxigenic E. coli O104 outbreak identified in Germany in May 2011. Cases related to this outbreak were detected in several European countries, including a cluster of cases which were exposed in the Bordeaux region of France. As of 27/7/2011, a total of 782 HUS cases, including 29 deaths, and 3128 non-HUS cases, including 17 deaths, were reported to the European Centre for Disease Prevention and Control (ECDC).¹ A Task Force of the European Food Safety Authority (EFSA) reported on 5/7/2011 that fenugreek seeds imported from Egypt were the most likely source of the outbreaks in Germany and France.² This outbreak is a reminder of the potential severity of disease associated with VTEC infections, and the magnitude of outbreaks that can result from contamination of food produced and distributed on a large scale.

Fortunately, there were no cases related to this outbreak identified in Ireland. However, the reported verotoxigenic E. coli (VTEC) incidence rate in Ireland is generally high relative to other European countries, and has been rising steadily over the last five years. In 2008 and 2009, Ireland reported the highest VTEC incidence rate of any Member State in the European Union.^3,4 Infection has historically been most commonly associated with VTEC serogroup O157, with smaller numbers of non-O157 VTEC reported, although this may reflect diagnostic bias as the techniques for the detection of non-O157 VTEC are more complex than for VTEC O157, and different policies may exist in different laboratories in Ireland for the routine examination of stool specimens.

The dominant transmission routes reported for VTEC infection in Ireland have been person-to-person spread, especially in creches/childminding facilities and among families with young children, and waterborne transmission associated with exposure to water from untreated or poorly treated private water sources.^5-8 Other important transmission routes identified internationally include food (often minced beef products or fresh produce such as lettuce and spinach), and contact with infected animals or contaminated environments.^{3,4 9-11}

Materials and Methods

Infection due to Enterohaemorrhagic E. coli (EHEC) is notifiable (S.I. 707 of 2003) since 2004 by clinicians and laboratory directors. This report focuses on cases that conform to the case definition used for VTEC enhanced surveillance. Click here. Enhanced epidemiological information was supplied as in previous years by HSE personnel, and VTEC confirmation and typing data were provided by the HSE Dublin Mid Leinster Public Health Laboratory at Cherry Orchard Hospital (DML-PHL). Data from all sources are maintained in the Computerised Infectious Disease Reporting (CIDR) system. Outbreaks of VTEC are notifiable since 2004 and data are provided to CIDR by regional public health departments. The data presented here were retrieved from CIDR on April 27^th 2011.

Data from the CSO 2006 census were used to provide denominator data for the calculation of incidence rate.

Results

Incidence

In 2010, there were 199 confirmed and probable cases of VTEC notified, equating to a crude incidence rate (CIR) of 4.7 per 100,000 (Table 1). If only confirmed VTEC cases are considered, the 197 cases (CIR=4.7 [4.0-5.3]) notified this year represent a 17% decrease overall on the number of confirmed cases notified in 2009. Non-O157 VTEC made up 41% of cases in 2010, however, and this overall change in VTEC case numbers was made up of a 30% decrease in the number of VTEC O157 cases and an 11% increase in non-O157 cases reported compared to 2009 (Figure 1).

Table 1. Number and crude incidence rates confirmed and probable VTEC, Ireland 2004-2010

^a Data from the 2006 census were used to calculate rates

^b Confirmed cases include 116 VTEC O157 cases, 67 VTEC O26 cases and 14 VTEC strains of other serogroups. Two probable cases were reported on the basis of being epidemiologically linked to laboratory confirmed cases (VTEC O157 in one instance and VTEC O121 in the second instance).

Figure 1. Annual number of confirmed and probable VTEC cases by serogroup, Ireland 1999-2010

Of 192 cases where information was available on symptoms, 147 (77%) were symptomatic, 75 (51%) of which developed bloody diarrhoea. Nineteen individuals (9.5%) developed HUS compared to 24 (10.0%) last year (21% decrease). And where reported, 42% of notified cases required hospitalisation (72/173).

Seasonal distribution

Typically, VTEC cases are most commonly associated with late summer; overall this year, 48% of cases were reported in quarter 3. However, the seasonal distribution varied by serogroup, with VTEC O26 being more common in quarter 2 while VTEC O157 remained more common in quarter 3. It is possible that the overall seasonal distribution noted previously was biased by the dominance of VTEC O157 in the national dataset, and that these variations in seasonal distribution by serogroup reflect a seasonal difference in sources or transmission routes for different serotypes.

Figure 2. Seasonal distribution of VTEC cases by serogroup, Ireland 2010

Regional distribution

Overall this year the highest VTEC incidence rates were reported in the HSE-MW and HSE-NW, where the rates were over twice the national crude rate (Table 2). As in previous years, the HSE-E reported the lowest overall crude incidence rate (Table 2), just over one quarter of the national rate this year. The crude incidence rate in the HSE-NE is also consistently low relative to other areas.

When the incidence by HSE-area is examined by serogroup, the incidence rate across HSE-areas for VTEC O157 is similar across six of the HSE-areas, with only HSE-E and HSE-NE showing lower incidence rates (Table 2 and Figure 3). The elevated overall incidence rates in the HSE-MW and HSE-NW were strongly influenced by their high reported incidence rates for non-O157 infections. Historically, the HSE-NW (and more recently the HSE-MW) reported relatively high numbers of non-O157 VTEC infections. While it is possible that there is a true geographical difference in risk for different serogroups, it is likely that this regional variation in non-O157 VTEC incidence to some extent reflects regional differences in laboratory diagnostic practice for non-O157 infections.

Review of regional HUS incidence due to confirmed or probable VTEC infection gives a slightly different perspective on the relative importance of VTEC by region (Table 2 and Figure 3). The HSE-S reported the highest VTEC-associated HUS incidence rates in 2010, followed by HSE-M, HSE-MW and HSE-NW. The eastern part of Ireland including HSE-E, HSE-SE and HSE-NE displayed the lowest VTEC-associated HUS incidence rates.

Table 2. Number and crude incidence rate confirmed and probable VTEC by serogroup and HSE area, and number and crude incidence rate VTEC-associated HUS by HSE-area, Ireland 2010

*Rates per 100,000 calculated using CSO census 2006 for denominator data

Figure 3: Crude incidence rate VTEC O157, non-O157, and VTEC-associated HUS by HSE-area, Ireland 2010

VTEC typing

In 2010, all isolates from the 197 confirmed VTEC cases were referred to the DML- PHL, and their serotype and verotoxin profiles are displayed in Table 3. As usual among VTEC O157 in Ireland, isolates containing the genes for verotoxin 2 (vt2) were more common (86%) than strains containing both vt1 and vt2. VTEC O26 strains containing only vt1 made up 55% of all VTEC O26 reported, with 36% of VTEC O26 containing the genes for both vt1 and vt2.

HUS cases in 2010 were associated with VTEC O157 strains containing vt2 or both vt1 and vt2, and with VTEC O26 containing vt2 or both vt1 and vt2, but not with any of the VTEC O26 strains containing vt1 alone.

Table 3. Serotype and verotoxin (VT) profiles for VTEC isolates as determined at the PHL HSE Dublin Mid Leinster, Cherry Orchard Hospital in 2010 by HUS status as recorded on CIDR

^a One HUS and one non-HUS case in 2010 were reported on the basis of being epidemiologically linked to laboratory confirmed cases, and thus no isolates were available for inclusion in this table

In 2010, the DML-PHL introduced a new, more highly discriminatory service whereby all human VTEC isolates were routinely typed by pulsed field gel electrophoresis (PFGE) rather than referred to the Health Protection Agency laboratory in the United Kingdom for phage typing. This confirmed that no large undetected clusters/outbreaks occurred among laboratory confirmed VTEC cases in Ireland in 2010. On a small number of occasions, indistinguishable isolates were identified among pairs of cases reported as sporadic (and sometimes between a family cluster and another case reported as sporadic). On these occasions, the HSE PH departments were informed and review of case histories undertaken. However, no epidemiological evidence of links were uncovered on any of these occasions.

Risk factors

Under enhanced surveillance for VTEC, risk factor information is routinely collected on VTEC notifications (Table 4).

Table 4. Number of cases (and percentage where information received) for which specified risk factor was reported, Ireland 2010

^aThis is a composite variable recoded from two different water supply exposure variables in CIDR

Among VTEC cases in Ireland in 2010, exposure to farm animals or their faeces and exposure to private well water were common among cases; 52.4% and 42.9% reported these exposures respectively. This is consistent with the low incidence of VTEC infection among residents in the largely urban HSE-E population and the higher incidence recorded in more rural parts of the country.

Unlike salmonellosis, foreign travel plays only a minor role in VTEC infection in Ireland, with the majority of infections acquired indigenously. The countries where the small number of travel-associated Irish VTEC cases had travelled to during their potential incubation periods were Bulgaria, Cyprus, Sri Lanka and United Kingdom.

For the 19 cases where food was suspected as the cause of illness, burgers and other minced beef products were listed as suspected for seven cases, sausages were listed for three cases, and other meat products and foods for five cases. Where tested, no foods were found positive for the VTEC strains implicated in the human cases, although one raw sausage product tested was found positive for an unrelated VTEC O8 strain.

Outbreak and environmental investigations

Forty-five VTEC outbreaks were notified in 2010, which included 103 of the 199 VTEC notifications. The majority of outbreaks (96%) were family outbreaks with only two general outbreaks notified. Both general outbreaks involved private households and childcare arrangements/creche facilities: five persons in total were reported ill between these two outbreaks. One general outbreak was reported as being due to person-to-person spread while the transmission route for the second general outbreak was reported as unknown.

Twenty-four outbreaks (53%) were caused by VTEC O157, sixteen (36%) by VTEC O26, four (9%) by other non-O157 and one (2%) was caused by a mixture of VTEC strains. The suspected modes of transmission reported are listed in table 5.

Table 5. VTEC outbreaks in Ireland 2010 by suspected mode of transmission

Person-to-person spread is an important mode of VTEC transmission particularly between young children, and was suspected to have played a role in 30 (67%) VTEC outbreaks in 2010 in which 65 persons were reported ill.

Figure 4. Number of VTEC outbreaks by suspected transmission route and year, Ireland 2004-2010

Note: In this figure, reported transmission routes were grouped for simplicity. Any outbreak where food contributed was reported as foodborne, any outbreak where water contributed was reported as waterborne, any outbreak where animal contact contributed was reported as Animal contact. Person-to-person outbreaks include only those outbreaks reported as being due only to person-to-person transmission.

Unusually, the second most common transmission route reported for VTEC outbreaks in 2010 was animal contact, which was reported to have contributed to four outbreaks (9%).

Unlike previous years, when exposure to private well water was a commonly reported transmission route (Figure 4) , there was only one VTEC outbreak reported as waterborne in 2010 (Figure 4). This family outbreak was reported associated with a private untreated well. No VTEC organisms were identified in the suspected water supply.

Separately, a family outbreak of two VTEC cases was reported to the United Kingdom's outbreak reporting system comprising two UK residents who had returned home following a visit to Ireland. Follow-up investigation by the HSE-E showed the outbreak also to be waterborne.  A private well water sample taken from a premises at which they had stayed was found positive at the DML-PHL for a VTEC strain indistinguishable from one of their infections.

Unlike previous years, no foodborne VTEC outbreaks were reported in 2010 (Figure 4). But for over one quarter (n=12) of VTEC outbreaks, the transmission route was reported as unknown.

In 2010, one sporadic VTEC case was reported in a laboratory worker who had exposure to VTEC during the course of their work.

Discussion

In 2010, the reported incidence of VTEC infection in Ireland decreased by 17% overall, with the VTEC O157 rate decreasing by 30% relative to 2009. The concomitant reduction of 21% in the VTEC-associated HUS numbers suggests that this was a true decrease in VTEC incidence. Interesting in 2010 in Scotland, there was also a decrease in the VTEC O157 incidence rate to 4.1 per 100,000 (11% decrease) and in England and Wales, VTEC O157 case numbers fell by 23%.^12,13

 The CIR for VTEC overall of 4.7 per 100,000 population in Ireland, however, remains high relative to Europe.^3,4 The reported VTEC incidence in 2009 at European community level was 0.7 VTEC cases per 100,000 population, with Denmark (2.90/100,000), Sweden (2.46/100,000) and the United Kingdom (2.19 per 100,000) reporting the next highest rates after Ireland.

While VTEC O157 infections declined in Ireland in 2010, the reported incidence rate for non-O157 infections increased, following the trend of recent years. This is likely to be due to improved surveillance for non-O157 infections rather than a true increase in non-O157 infections. Moreover, it is probable that these infections are still under-diagnosed; a strong geographical variation in the reported incidence of non-O157 suggests that there may be regional variation in laboratory diagnostic policy. Given the recent outbreak in Europe, and its association with a non-O157 VTEC strain, the detection and reporting of non-O157 infections become even more essential. Moreover, as data accumulates on non-O157 infections, it may be possible to investigate if there are epidemiological differences between VTEC serogroups in Ireland. For example, the seasonal distribution of VTEC O26 infections in 2010 appeared to differ from VTEC O157. In future years, it should be possible to verify if this is a genuine feature of VTEC O26 infections or an artifact in 2010.

Person-to-person spread is a major transmission route for VTEC infection.^5,6 This year, person-to-person transmission contributed to 64% of VTEC outbreaks in Ireland. Transmission between young children is particularly common, possibly because of lower levels of hygiene in this group, and again in 2010, two general outbreaks were reported which included transmission in childcare facilities/arrangements. The HPSC recently developed guidelines for creche owners in relation to infectious diseases. The document was issued for consultation during the spring of 2011 and will be finalized shortly.

However, an outbreak reported as being due to ‘person-to-person transmission’ cannot entirely be considered definitive as the index case must have acquired their illness originally from a source.

The second most commonly reported transmission route for VTEC outbreaks in Irelandin 2010 was animal contact (9% of outbreaks). Contact with animal faeces was previously shown to be a strong risk factor for VTEC infection in Scotland;⁷ moreover in 2009 a large outbreak of VTEC O157 infection occurred in England which was associated with an open farm.⁸ That outbreak resulted in 93 confirmed cases of illness, 17 of whom developed HUS, and is a stark reminder of the potential role of direct animal contact and contact with animal faeces in disease transmission and the importance of hand hygiene messages, in particular for children, following contact with farm animals and pets. Guidance for pet farm operators and on the recreational use of farmland, for the prevention of VTEC and other zoonotic infections are available on the HPSC website. Click here.  

Exposure to untreated or inadequately treated private well water has repeatedly been reported as a risk factor for VTEC infection.^{5, 7, 8} Waterborne transmission was reported for only one outbreak in 2010 relative to 7 and 10 outbreaks respectively in 2008 and 2009. In general, the weather in 2010 was drier than normal and this may have contributed to a smaller role for water in VTEC outbreaks in 2010.¹⁵

In 2010, DML-PHL in Cherry Orchard undertook routine molecular typing (i.e. PFGE) of all human VTEC isolates; the service demonstrated that there were no large undetected VTEC outbreaks as a diversity of profiles were obtained for the majority of VTEC isolates tested. Among the small number of indistinguishable isolates identified, no epidemiological links were identified by public health. While it is possible that these indistinguishable pairs of isolates represent more common clones in the environment, it is also possible that they represent diffuse outbreaks (possibly foodborne) and public health departments will continue to investigate any potentially linked cases identified in this manner to outrule this possibility. The severity and scale of the VTEC O104 outbreak in Europe is a reminder of the value of laboratory services in outbreak detection and investigation.

Patricia Garvey and Paul McKeown, HPSC, and Anne Carroll and Eleanor McNamara, HSE-DML PHL

Acknowledgements

The authors wish to acknowledge the co-operation of clinicians, microbiologists, medical scientists, SMOs, SPHMs, surveillance scientists, infection control nurses, PEHOs, and EHOs in providing the information on which this report is based.

References

1. ECDC. 2011. Epidemiological updates on the VTEC O104 outbreak. Click here.
2. EFSA Tracing seeds, in particular fenugreek (Trigonella foenum-graecum) seeds, in relation to the Shiga toxin-producing E. coli (STEC) O104:H4 2011 Outbreaks in Germany and France. 2011.  Click here.
3. EFSA. 2009. The Community Summary Report on Trends and Sources of Zoonoses, Zoonotic Agents, Antimicrobial resistance and Foodborne outbreaks in the European Union in 2008. Click here.    
4.  EFSA 2010. The Community Summary Report on Trends and Sources of Zoonoses, Zoonotic Agents, Antimicrobial resistance and Foodborne outbreaks in the European Union in 2009. Click here.
5. Garvey, P. et al. 2010. Epidemiology of verotoxigenic E. coli in Ireland, 2007. Epi-Insight: 11(9)
6. Locking et al. 2010. Escherichia coli O157 Infection and Secondary Spread, Scotland, 1999–2008 EID 17(3): 524 Click here.
7. O’Sullivan et al. 2008. Increase in VTEC cases in the south of Ireland: link to private wells? Eurosurveillance 13(39) click here.
8. HPSC. 2008. Press release. Householders must properly maintain private water supplies following increase in contamination – HPSC. Click here. 
9. Locking et al. 2001. Risk factors for sporadic cases of Escherichia coli O157 infection: the importance of contact with animal excreta. Epidemiol Infect. 127(2):215-20. Click here.
10. Griffen. 2010. Review of the major outbreak of E. coli O157 in surrey, 2009. Click here. 
11. CDC. Ongoing multistate outbreak of Escherichia coli serotype O157:H7 infections associated with consumption of fresh spinach— United States, September 2006. MMWR 2006; 55(38): 1045-6.
12. HPS. 2011. Gastro-intestinal and foodborne infections: Escherichia coli O157, Salmonella and Campylobacter - laboratory reports, 2010. Click here.
13. HPA. 2011. E. coli O157 Annual Totals. Click here.
14. Met Eireann –Annual Summary of the weather in 2010.  Click here.