Two sampling methods (rectoanal swabs and rectal fecal grabs) were compared for their recovery of Escherichia coli O157:H7 from feedlot cattle. Samples were collected from 144 steers four times during the finishing period by swabbing the rectoanal mucosa with cotton swabs and immediately obtaining feces from the rectum of each individual steer. The number of steers with detectable E. coli O157:H7 increased from 2 of 144 (1.4%) cattle on arrival at the feedlot to 10 of 144 (6.9%) after 1 month, 76 of 143 (52.8%) after 7 months, and 30 of 143 (20.8%) at the last sampling time before slaughter. Wilcoxon signed-rank tests indicated that the two sampling methods gave different results for sampling times 3 and 4 (P < 0.05) but not for sampling time 2 (P = 0.16). Agreement between the two sampling methods was poor (kappa < 0.2) for three of the four sampling times and moderate (kappa = 0.6) for one sampling time, an indication that in this study rectoanal swabs usually were less sensitive than rectal fecal grabs for detection of E. coli O157:H7 in cattle. Overall, the herd of origin was not significantly associated with E. coli O157:H7 results, but the weight of the steers was. Further investigation is needed to determine the effects of potential confounding factors (e.g., size and type of swab, consistency of feces, site sampled, and swabbing technique) that might influence the sensitivity of swabs in recovering E. coli O157:H7 from the rectoanal mucosa of cattle.
Rank Tests from Partially Ordered Data Using Importance and MCMC Sampling Methods