Computed tomography in the assessment of suspected large bowel obstruction

BACKGROUND: The aim of this study was to assess the efficacy of computed tomography (CT) scanning in the diagnosis of acute large bowel obstruction. METHODS: Forty-four patients (22 men; 22 women, ages 39-94 years, mean 71 years) with clinical features and abdominal radiographic findings suggesting acute large bowel obstruction (LBO) or pseudo-obstruction were examined with CT. Supine scans were obtained with i.v. contrast medium (unless contraindicated), but (in the majority) without oral contrast. Additional prone and/or decubitus scans were obtained in 33 patients when clarification of a possible transition point on the supine scan was required. CT diagnosis of LBO was made by finding a transition point +/- mass. Final diagnosis was confirmed by surgery, further imaging and/or clinical course. RESULTS: Twenty-two patients had proven mechanical acute LBO of whom 18 had an obstructing carcinoma; 22 patients had no mechanical obstruction. Sensitivity, specificity, Positive Predictive Value, Negative Predictive Value of CT for diagnosis of mechanical LBO were each 91%. Positive and negative likelihood ratios were 10.1 and 0.1, respectively. There were two false-negative CT scans, although one of these was reported as showing segmental mural thickening. A mass was identified on 14 of 17 patients with true-positive CT, subsequently found to have carcinoma. CONCLUSION: Computed tomography with additional selective prone and/or decubitus scanning is highly effective in the diagnosis of mechanical LBO. It is suggested that it replace contrast enema as the initial imaging method.