Supplementary Material for: High accuracy of a simplified, practical algorithm in differentiating Crohn's Disease from Intestinal Tuberculosis

Background: The differentiation between intestinal tuberculosis from Crohn's disease remains a challenge, particularly in areas where tuberculosis is highly prevalent. Previous studies have identified features that favor one diagnosis over the other. To determine the accuracy of a standardized protocol in the initial diagnosis of Crohn’s disease versus intestinal tuberculosis. Methods: All patients with suspected intestinal tuberculosis or Crohn’s disease were prospectively recruited. A standardized protocol was applied and the diagnosis was made accordingly. The protocol consists of; history and examination, ileocolonoscopy with biopsies, and tuberculosis workup. The diagnosis of probable intestinal tuberculosis was made based on at least one positive finding. All other patients were diagnosed as probable Crohn’s disease. Patients were treated either with anti-tubercular therapy or steroids. Reassessment was then carried out clinically, biochemically and endoscopically. In patients with suboptimal response, the treatment was either switched or escalated depending on the reassessment. Results: 164 patients were recruited with final diagnosis of 30 (18.3%) intestinal tuberculosis and 134 (81.7%) Crohn’s disease. 1 (3.3%) out of 30 patients with intestinal tuberculosis was initially treated as Crohn’s disease. 16 (11.9%) out of 134 patients with Crohn’s disease were initially treated as intestinal tuberculosis. The initial overall accuracy for the protocol was 147/164 (89.6%). All patients received the correct diagnosis by 12 weeks after reassessment. Conclusion: In our population, most patients had CD rather than ITB. The standardized protocol had a high accuracy in differentiating CD from ITB.