Abstract
The performance of a commercially available enzyme immunoassay (EIA) for determining the presence of Shiga toxin I and II in human diarrheal stool samples was evaluated for use as a presumptive test for the presence of Escherichia coli O157:H7 in nondiarrheal bovine fecal samples collected from 10 Kansas cow–calf ranches. The prevalence of E. coli O157:H7 in 2,297 samples, as determined by selective bacterial culture, was 1.6%. The sample prevalence of non-E. coli O157:H7 Shiga toxin–producing bacteria, as detected by the Shiga toxin EIA, was 5.8%. Only 2 of 136 samples that tested positive with the Shiga toxin EIA were positive for E. coli O157:H7 by culture. Compared with bacterial culture, the sensitivity of the Shiga toxin EIA was 5.5% and the specificity was 94.1%. Agreement between the 2 tests, as measured by the kappa statistic, was poor (κ = −;0.002). Although the Shiga toxin EIA was not a good presumptive test for the determination of E. coli O157:H7 in bovine fecal samples because of its low sensitivity (5.5%), it might be a useful test for the detection of Shiga toxin producing non-E. coli O157:H7 organisms in bovine feces.