Multidetector helical CT in the evaluation of acute small bowel obstruction: Comparison of non-enhanced (no oral, rectal or IV contrast) and IV enhanced CT

Abstract

Purpose To compare accuracy of non-enhanced CT (NECT) (no oral or IV contrast) and enhanced CT (ECT) (IV enhanced only) to diagnose small bowel obstruction and evaluate reviewer's experience impact. Materials and methods Ninety-nine adult patients underwent 105 NECT and ECT (6 patients had 2 examinations) on a four-detector CT. An abdominal radiologist, an abdominal imaging fellow, a second-year radiology resident retrospectively reviewed NECTs and ECTs separately and independently blinded to outcome. Discrepancy of diameter of proximal and distal small bowel ± a transition was considered indication of mechanical bowel obstruction. Reference standard was surgery in 26 and chart review in 79. Results Mechanical obstruction was present in 56% (59/105). The average sensitivity, specificity, negative and positive predictive and accuracy values for NECT were 88.1% (CI: 80–96%), 77% (CI: 65–89%), 83.0% (CI: 72–95%), 83% (CI: 74–92%), and 83% (CI: 76–90%) with no significant difference between three reviewers. The corresponding numbers for ECT were 87.6% (CI: 79–96%), 75% (CI: 63–88%), 82.6.0% (CI: 71–94%), 82.1% (CI: 73–92%), and 82% (CI: 75–90%) ( p > 0.5). Area under curve (AUC) of ROC curves of three reviewers did not show significant statistical difference ( p > 0.5). Conclusions NECT and ECT have comparable accuracy to diagnose mechanical small bowel obstruction and can be interpreted by reviewers with different levels of expertise.