CROHN’S DISEASE OR INTESTINAL TUBERCULOSIS? A PRACTICAL DIAGNOSTIC APPROACH FOR PAKISTAN: A RAPID SYSTEMATIC REVIEW AND META-ANALYSIS OF DISCRIMINATING FEATURES

Abstract

Background. In Pakistan, a country with one of the world’s highest tuberculosis (TB) burdens, Crohn’s disease (CD) and intestinal TB (ITB) can look almost identical, yet treating one as the other is harmful. We asked which clinical and endoscopic features best separate the two, and what practical diagnostic sequence follows for Pakistani practice. Methods. We pooled nine features reported in nine comparative cohorts from India, China and South Korea (study-level data tabulated in a 2019 review), using logit odds ratios with DerSimonian–Laird random effects, Hartung–Knapp intervals, leave-one-out analysis and an Egger-type test. We searched separately for Pakistani evidence and summarised it narratively. Results. Longitudinal ulcers (OR 15.6; 95% CI 9.2–26.3) and cobblestoning (OR 8.2; 4.9–13.8) pointed strongly to Crohn’s disease, as did, more weakly, aphthous ulcers, rectal and sigmoid involvement, diarrhoea and rectal bleeding. Transverse ulcers (OR 0.10; 0.05–0.18) and a patulous ileocaecal valve (OR 0.15; 0.07 0.31) pointed to TB. No single feature was diagnostic: even the best, longitudinal ulcers, was absent in about half of Crohn’s patients. No Pakistani comparative cohort was found. The one Pakistani Crohn’s cohort (22 patients, Rawalpindi) reported long delays to diagnosis, with about 40% having taken anti-TB treatment first. Conclusions. Endoscopic pattern is the most useful first discriminator, but only in combination with history, biopsy, imaging and, when doubt remains, a time-limited TB treatment trial with a fixed review point. We offer a stepwise approach for Pakistani clinicians, clearly labelled as an evidence-informed proposal that has not been validated locally. A multicentre Pakistani comparative cohort is the most important research priority.