Distinguishing features of self-limiting adult small-bowel intussusception identified at CT

PURPOSE: To determine if clinical or computed tomographic (CT) findings can be used to distinguish self-limiting cases of adult small-bowel intussusception from those requiring surgery. MATERIALS AND METHODS: Thirty-seven cases of adult small-bowel intussusception were identified by a retrospective computerized search of 69,040 abdominopelvic CT examinations performed over a 4-year period. Two independent readers recorded CT features. Clinical findings and outcomes were determined by review of all available medical records. Outcome was classified as either surgical or self-nnlimiting. Association between predictive variables and outcome was assessed by the Fisher exact test and logistic regression models. A multivariate, stepwise, logistic regression model was used to determine the best predictors of outcome. RESULTS: Six patients (16%) underwent surgery, and all had lead-point tumors. Thirty-one patients were cared for conservatively (84%) and none required surgery at a mean follow-up of 5.2 months (range, 0-46 months). Multivariate, stepwise, logistic regression analysis showed intussusception length was the only variable that was independently predictive of outcome. All 20 patients with an intussusception length of 3.5 cm or less, as measured by either reader, had cases that were self-limiting. Seventeen patients had an intussusception length greater than 3.5 cm, as measured by either reader. Eleven patients had an intussusception that was self-limiting, and six patients had an intussusception that required surgery. CONCLUSION: Intussusception length is the main factor in distinguishing the majority of small-bowel intussusceptions detected at CT that are self-limiting from the minority that require surgery. An intussusception that is shorter than 3.5 cm is likely to be self-limiting.