Background 

Over the past several years, the area of pre-harvest food safety research in beef cattle has grown tremendously. Several interventions have been identified that may control pathogenic bacteria in the feedlot prior to slaughter. Current methods for determining the prevalence of E. coli O157:H7 in cattle include fecal grab samples from the rectum of animals, fecal pat samples from the floor of the pen and hide sampling. New information indicates that the type of sample collected may yield different results, a factor that could be important in evaluating pre-harvest interventions. It is important to conduct a side-by-side comparison of sampling methods in order to determine the best method.   

Previous research has reported that the primary site of colonization of E. coli O157:H7 in cattle is adjacent to the “recto-anal-junction” (RAJ), however it is unclear if fecal grab samples include bacteria from this colonization site. Fecal pat samples have also been shown to be inconsistent in determining prevalence rates as the surface of fecal pats is more likely to test positive than the center of the fecal pat. Determining the best method for fecal testing is important as fecal pat samples from a feedlot pen floor areoften used to determine the efficacy of an intervention or prevalence of a pathogen in a particular setting.   

Previously published research sampled seven sites from beef feedlot cattle (Keen and Elder; J Am Vet Med Assoc 2002; 220:756-63). These sites were feces, oral cavity, and five hide locations (lumbar region, ventral aspect or bottom portion of the neck, ventral abdominal midline, dorsal thoracic midline and distal aspect of the left hind leg). Different sites on hides have also demonstrated varying likelihood of E. coli O157:H7 recovery. Hide testing is an important supplement to fecal testing because hides are a primary source of contamination of the carcass.   

The objective of this study was to determine ideal sampling methods to estimate the prevalence of E. coli O157:H7 and Salmonella in feedlot cattle to potentially provide standardization of methods for future epidemiological and intervention studies.

Methodology 

Study 1—Sampling Sites 

Nine samples were collected from 87 animals from the following sites:

• Fecal grabSalmonella was recovered from at least one site from all fifty animals analyzed. The sites of greatest recovery were the oral cavity (94 percent), perineum (88 percent) and hock (94 percent). Feces had the lowest recovery rate (50 percent). samples were taken directly from the rectum of the animals 
• Swabs of the rectoanal junction (RAJ) 
• Oral samples 
• Hide samples from the: 
• Flank region 
• Perineum 
• Ventral aspect of the neck 
• Ventral abdominal midline 
• Dorsal thoracic midline 
• Distal aspect of the left hind leg or hock

Study 2—Fecal Sampling 

Freshly voided fecal pats were identified for appropriate size and consistency for the purposes of this study. Five separate samples from each fecal pat were collected and a homogenized sample from the five was also prepared in the lab and included in the analysis. Variation between the samples from the same fecal pat were examined, as well as the homogenized sample to determine if it provided an estimate of E. coli O157:H7 prevalence as accurate as analyzing all of the samples separately.   

E. coli O157:H7 were isolated using an immunomagnetic separation (IMS) and confirmed using H7 agglutination and API 20 tests were used to confirm. Relative sensitivities for each culture methodology were calculated to determine the most sensitive method.

Findings 

Study 1—Sampling Sites 

E. coli O157 was recovered from 14.9 percent of the animals. Of these positive animals, the greatest estimates of prevalence were obtained from fecal grab samples (46.2 percent), perineum swabs (38.5 percent) and hock swabs (38.5 percent). Estimates of prevalence obtained from swabs of the terminal rectum, dorsal midline, oral cavity, flank, brisket and ventral neck were 15.4, 15.4, 7.7, 15.4, 23.1 and 23.1 percent, respectively.

Salmonella was recovered from at least one site from all fifty animals analyzed. The sites of greatest recovery were the oral cavity (94 percent), perineum (88 percent) and hock (94 percent). Feces had the lowest recovery rate (50 percent).

Study 2—Fecal Sampling 

Five samples per fecal pat were collected from 48 fecal pats. In addition, a composite of the five samples was generated. Of the fecal pats sampled, 22.9 percent were positive at any one of the five sites. Of the positive fecal pats, 36.4, 9.1, 9.1, 18.2 and 27.3 percent of samples were positive at one, two, three, four or five locations, respectively. Of the composite samples, 14.6 percent of samples were positive, and were not significantly different from the estimate from sampling one site. As a result, the researchers concluded that collecting a composite sample did not improve prevalence estimates.

Implications   

The best hide sites for pathogen recovery were similar for Salmonella and E. coli O157:H7, however feces was the best overall sample for detecting of E. coli O157, but the worst for detecting of Salmonella. In the fecal analyses, compositing samples was not more effective in determining E. coli O157:H7 prevalence than taking multiple samples. This study provided important information on effective sampling sites and methods for determining E. coli O157:H7 and Salmonella prevalence in cattle.