PURPOSE: To assess the accuracy of radiologic interpretation, in the absenc
e of clinical information, in the differentiation of benign and malignant s
igmoid strictures at barium enema examination.
MATERIALS AND METHODS: On two occasions, four independent observers retrosp
ectively assessed examination findings in 78 patients with documented sigmo
id strictures (43 benign, 35 malignant). Each stricture was graded by using
a five-point scale (definitely malignant to definitely benign).
RESULTS: No significant difference existed between the areas under the rece
iver operating characteristic curves for the two assessments with any obser
ver. Consensus findings indicated agreement among at least three of the fou
r observers in 68 (87%) and 66 (85%) cases at the first and second assessme
nts, respectively. One benign stricture was called malignant at both assess
ments. When consensus existed, the positive predictive value for malignant
strictures was 96% at both assessments (sensitivity, 63% and 66%). Nine mal
ignant strictures were called benign, three at both assessments. When conse
nsus existed, the positive predictive value for benign strictures was 84% a
nd 88% at the first and second assessments, respectively (sensitivity, 88%
and 86%, respectively).
CONCLUSION: The differentiation between a benign and a malignant sigmoid st
ricture can be made in most cases at barium enema examination. When a stric
ture appears malignant, the diagnosis is usually correct, but caution is ad
vised when a stricture appears benign.