Abstract
Introduction: We conducted a study to determine the computed tomography (CT) findings useful in alimentary tract perforation diagnosis and management. Methods: We reviewed abdominal CT from 180 patients treated for gastroduodenal (n=99), jejunoileal (n=25), and colorectal (n=56) perforation from January 2000 to January 2008. We focused on extraluminal free air (FA) and intraperitoneal fluid collection (FC). We classified FA into upper and lower FA based on the level of the third duodenal portion, and superficial FA and deep FA based on whether FA was adjacent to the anterior peritoneal surface. FC was evaluated by extent and location. Results: FA was detected in 97% of gastroduodenal, 56% of jejunoileal, and 78.6% of colorectal perforations. In upper FA, 97% involved gastroduodenal, 52% jejunoileal, and 66.1% colorectal perforations, and in lower FA 18.2%, 24%, and 58.9%; in lower-deep FA 1%, 16%, and 51.8%. Sensitivity of lower-deep FA findings in diagnosing colorectal perforation was 51.8% and specificity 96.0%. Lower-deep FA was noted only in one case of gastroduodenal perforation, although upper FA was seen in all but three cases. FC extent and site showed no correlation with bowel perforation site. In all cases of gastroduodenal perforation treated successfully conservatively, FC width around the liver was 5 mm or less, with or without some FC in the pelvic cavity. FC in these cases did not increase in 24 hours. Conclusions: We found a correlation between FA and alimentary tract perforation sites. In gastroduodenal perforation, a small amount of FC with no increase in 24 hours was reliable for CT findings in indicating conservative therapy. CT findings were useful in handling alimentary tract perforation. Appropriately interpreting CT findings thus leads to prompt diagnosis and appropriate treatment.
