Background: Because it is not easy to make a clinical decision regarding surgical treatment in patients with acute attacks of ulcerative colitis, an objective, simple, criterion is needed to determine the optimum timing for colectomy. The aim of this study was to retrospectively examine to what extent an activity index (AI) can evaluate the clinical course in such acute attacks.
Methods: One hundred and twenty-seven patients with moderate or severe attacks of ulcerative colitis were examined. AI values and the decline in AI values were compared between surgical and nonsurgical groups after 1 week and 2 weeks of medical therapy. To evaluate the clinical course of acute attacks, cutoff AI values were set at every 10 points between values of 180 and 210. The positive predictive value for surgery was examined.
Results: AI values in the surgical group were significantly higher than those in nonsurgical group at pretreatment, and after 1 or 2 weeks of medical therapy. The decline of AI values in the nonsurgical group was significantly higher than that in the surgical group after 1 or 2 weeks of medical therapy. At pretreatment, the prediction of colectomy was less than 50% at any of the cutoff values. After 1 week of therapy, approximately 60% of patients with an AI value greater than any of the cutoff AI values required colectomy. After 2 weeks of therapy, 30 of 43 (70%), 28 of 38 (74%), 24 of 29 (83%), and 17 of 21 (81%) patients with AI values greater than 180, 190, 200, and 210, respectively, required colectomy. Overall accuracy was 86%, 87%, 88%, and 83% for cutoff AI values of 180, 190, 200, and 210, respectively. Because the overall accuracy and positive predictive value for colectomy at AI values of 200 were significantly higher than these parameters at other AI values after 2 weeks of therapy, an AI value of 200 was regarded as the cutoff value most able to predict colectomy.
Conclusions: We concluded that patients with an AI value in excess of 200 after 2 weeks of medical therapy would require surgical treatment.