Summary
Medical records of 245 horses that had been evaluated by use of abdominal radiography between January 1990 and December 1992 were reviewed. One hundred forty-one horses subsequently had a postmortem examination or surgical exploration performed for definitive lesion diagnosis. The signalment, diagnosis, site, and number of enteroliths were obtained from the medical records. Radiographs were evaluated individually by 3 reviewers for the presence of enteroliths, preferred diagnostic view, evidence of large colon tympany, and film quality. Of the 141 cases reviewed, 66.7% (94/141) had confirmed enterolithiasis. Enteroliths were identified in the right dorsal colon of 59 horses, in the transverse colon of 28, in the small colon of 24, and in the ventral colon of 1 (enteroliths were detected in multiple sites in 12 horses). For the 3 reviewers, mean sensitivity was 76.9% and specificity was 94.4%. Mean positive-predictive value was 96.4%, and negative-predictive value was 67.5%. Cases involving only large colon enteroliths were correctly diagnosed 83.2% of the time, compared with 41.6% of the time for cases involving small colon enteroliths. Enteroliths were evident in 54.8% of the horses with radiographic signs of large colon distention. Fifteen horses had enteroliths in the small colon, 4 in the transverse colon, and 4 in the dorsal colon.
Radiographic quality was evaluated and scored as adequate (0), underexposed (−1), overexposed (1), or incomplete. Mean score was −0.5, and there were 21 (14.9%) incomplete studies. For the 75 cases correctly diagnosed via abdominal radiography, there were 14 (18.7%) incomplete studies and a mean score of −0.43. Of the 19 false-negative evaluations, there were 3 (15.8%) incomplete studies and a mean score of −0.69. The most common error leading to the missed diagnoses seemed to be inadequate penetration of the abdomen.
