Abstract
In order to predict strangulated obstruction requiring partial resection of the small intestine due to severe ischemia, we enrolled 36 patients with strangulated obstruction diagnosed intraoperatively and compared for background factors, vital signs, blood test, and CT value of ascites and/or fluid in the dilated intestine between cases required and not required partial resection of the intestine. We applied a multiple logistic regression analysis using statistically significant three factors and determined a discriminant to predict strangulated obstruction requiring partial resection of the small intestine due to severe ischemia. The three factors were the CT value, CRP value and Alb value. By this discriminant, we had a sensitivity of 78.6% (11/14 cases), a specificity of 86.4% (19/22 cases), and an overall prediction rate of 83.3%. We also compared predictions of bowel resection by this discriminant with actual results in other 17 patients with strangulated obstruction at our hospital and 25 patients at another hospital. As a result, we had sensitivities of 100 and 60%, specificities of 80 and 90%, and prediction rates of 86.7 and 72%, respectively, suggesting this discriminant is useful to predict strangulated obstruction requiring partial resection of the small intestine due to severe ischemia in clinical settings.
