Background
Foodborne pathogens are the leading cause of illness and death in developing countries costing billions of dollars in medical care and social costs [
1].
E. coli O157: H7, an enterohemorrhagic
E. coli (EHEC), is one of the most common causes of foodborne infections in humans. It infects all age groups and the pathogen is noted for its severe consequences following infection, low infective dose and acid resistance [
2]. Depending on the immune status and the general health of the infected individual, and the dose and virulence of the bacteria, infection with
E. coli O157: H7 can result in mild diarrhea, severe bloody diarrhea, hemorrhagic colitis, or hemolytic uremic syndrome (HUS) leading to kidney failure [
2,
3].
Cattle are the primary reservoirs of
E. coli O157:H7 and ground beef and beef products are identified as major sources of foodborne transmission [
2,
4]. Carcass contamination occurs through skin-to-carcass or fecal-to-carcass transfer of the pathogen during slaughter process at processing plants [
5‐
7]; and this is the major risk factor for human infection. Furthermore, cross-contamination can occur during further processing of carcasses in the processing plants, during distribution and storage of beef at retail markets. Various pre-harvest interventions (vaccination, direct-fed microbial and bacteriophage treatment) to reduce pathogen shedding [
8,
9] and post-harvest intervention technologies such as skin and carcass washing, and the use of antimicrobials [
10] have been developed with varying success.
Despite greater burden caused by foodborne infections in developing countries than developed countries, there is acute scarcity of information on their occurrences [
1]. In Ethiopia, only very few studies can be found regarding
E. coli O157:H7 in animals, animal products or people [
11‐
14]. Therefore producing more information on this particularly important foodborne pathogen is crucial to create awareness in the public and formulate preventive measures along food production, processing, and distribution continuum. The objective of this study was to investigate the occurrence of
E. coli O157:H7 associated with beef cattle at processing plants, retail shops, and in people sampled at health centers in Ethiopia.
Discussion
The aim of the present study was to investigate the occurrence of
E. coli O157:H7 associated with beef cattle production, processing and distribution, and diarrheal diseases in the public seeking health care services. The prevalence of
E. coli O157:H7 both in the fecal and carcass swab samples collected both at the processing plants and retail shops was low compared to previous studies conducted in Ethiopia [
11‐
14]. The fact that we observed low prevalence of
E. coli O157:H7 in the fecal samples (2%) and intestinal mucosa (0.8%) suggests low infection in the beef cattle population studied. This finding can be extrapolated to a larger cattle population of Ethiopia albeit with caution. Although it is difficult to trace back the farms of origin of the cattle slaughtered, it is reasonable to assume that the two processing plants represent wide catchment areas where animals can be brought for slaughter from across the country. Our finding can be confounded by the fact that cattle could be brought both from big commercial feedlots and small scale backyard production systems. Our study suggests the need for large scale on-farm studies to determine the prevalence of
E. coli O157:H7 under different cattle production systems in Ethiopia. The prevalence of
E. coli O157:H7 is associated with herd size [
18] and it is more common under concentrated animal feeding operations compared to small herds raised typically on pasture [
19]. Since the current study was limited only to two beef cattle processing plants, future studies with broader scope representing major beef cattle processing plants in the country are warranted.
The recto-anal junction (RAJ) of cattle is the principal site of colonization for
E. coli O157: H7 [
20] and it was argued that detection of
E. coli O157: H7 in the intestinal mucosa proximal to RAJ indicates persistent infection by colonization rather than pass through the GIT as detected in the feces [
21,
22]. Therefore the low prevalence of
E. coli O157: H7 in the intestinal mucosal swabs observed in the present study indicates low infection in the beef cattle included in this study. Numerically, we observed higher prevalence of
E. coli O157: H7 in the fecal samples than in the intestinal mucosal swabs which would suggest rather a passing through of the pathogen than colonization. This was contrary to observation made by Fox et al., [
23] in which the prevalence of
E. coli O157:H7 in the rectal mucosa was twice as much as that observed in the colon content or feces.
The skin of cattle is a significant source for
E. coli O157: H7 contamination of beef, with the potential for the pathogen transfer onto the carcass during slaughtering and dressing processes [
19]. Skin contamination occurs from direct or indirect fecal contamination in beef cattle production and lairage environments; and plays a significant role for downstream carcass contamination. Cross contamination of skins with feces can also occur when a group of cattle is transported or held together in close quarters thus increasing the prevalence of
E. coli O157: H7 on skins. Level of skin contamination is positively associated with the fecal prevalence of in-coming cattle to the processing plants [
19]. Compared to studies reported in other countries, the prevalence (0.91%) of
E. coli O157: H7 on the skin swabs found in this study is comparatively low. This can be attributed to differences in the factors which can potentially affect skin contamination, including fecal shedding, abattoir management system, farming systems, lairage related conditions, duration of farm/market-to-abattoir transport and hygienic conditions along unloading-to-stunning areas. In this study we note that low prevalence of
E. coli O157: H7 on skin swabs is expected since its observed prevalence (2% in the fecal samples and 0.8% in the intestinal mucosal swabs) in the cattle population was low. The swabbing site could have an effect on the prevalence of
E. coli O157: H7 in the skin swab samples. Contrary to other studies [
24] which obtained skin swab samples from the shoulder of an animal, we swabbed the ventral surface of the animal over the sternum (brisket) extending over the neck area. This was based on the assumption that as cattle rest in sternal recumbence, this site would be in contact with fecal matter on the ground thus maximizing skin contamination. Also, since our swabbing site included the bleeding site this would facilitate entry of the pathogen to carcass surface during slaughtering process. Even though our sampling method (using cotton tipped swabs instead of sponges) and selection of swabbing site (ventral part of the animal) could have contributed to the low apparent prevalence of
E. coli O157: H7 in the skins samples, this effect can be considered minimal compared to the impact of low overall prevalence of
E. coli O157: H7 observed in the cattle population studied.
We observed low level of carcass contamination by
E. coli O157: H7 at the processing plants (0.54%) or retail shops (0.43%). This can be attributed generally to the low prevalence of
E. coli O157 observed in the cattle population (fecal, intestinal mucosal swabs and skin swab samples) as well as on the skins. Even though the prevalence of
E. coli O157: H7 was low, numerically we observed a gradual decline of the prevalence from the fecal (2%), skin (0.9%) and carcass samples (0.5%) at the processing plants demonstrating that current sanitary dressing procedures are effective against
E. coli O157: H7. It can also indicate good sanitary procedures observed at the processing plants and the retail shops. Our results are lower compared to previous studies conducted in Ethiopia. For instance
E. coli O157: H7 was reported in 8% of beef samples collected at abattoirs and retail shops [
12], in 2.7% of beef carcass swab samples collected from a slaughter house [
25], and in 2.1% of beef carcass and cutting board swab samples collected from retailer shops [
26]. Carcass contamination by feces can occur directly from the intestinal content during evisceration, or indirectly from skins during dressing operation or from the abattoir environment such as from contact with conveyer belts. Lack of detection of
E. coli O157:H7 in the environmental samples could be attributed to the low overall prevalence of this pathogen in the present study and the small number of environmental samples (
n = 62).
E. coli O157: H7 was not detected in the stool samples from people with diarrhea seeking health services. Studies reporting
E. coli O157: H7 in humans are limited in Ethiopia despite the common occurrence of diarrhea problems especially in children. One study [
11] conducted in children (
n = 422) under five years of age with acute diarrhea reported 14% prevalence of
E. coli O157: H7. Since our sample size (70 cases) was not sufficient, compared to the aforementioned study (422 cases), to draw conclusive evidence we recommend a more population based study. The human clinical
E. coli O157: H7 isolates (
n = 59) in the above study [
11] exhibited resistance, at varying degrees, to all of the antimicrobials tested with the highest resistance (90%) to ampicillin. Among the
E. coli O157: H7 isolates from beef cattle, skins and carcass samples, even though antimicrobial resistance was rare, all the isolates were resistant to ampicillin. Similarly, Taye et al. [
25] reported 100% resistance to Ampicillin and Amoxicillin. The high prevalence of resistance to the beta-lactam class of antimicrobials both in the clinical and animal origin
E. coli O157: H7 isolates requires further investigation. Even though it is impossible to draw epidemiologic association between the observed high resistance to the beta-lactam classes of antimicrobials and the frequency (quantity) of use of these antimicrobials in this study, we speculate that beta-lactams could be the most commonly used antimicrobials both in humans and cattle in Ethiopia. This hypothesis needs to be further elucidated since determining the amount of antibiotic use both in humans and animals is critical to assess the contribution of antibiotic use to the level of AMR observed in a given country.
Acknowledgement
We thank the cooperation of all the personnel at the processing plants, retail shops, and health centers.