Accuracy and cost of diagnostic strategies for patients with suspected Crohn's disease.

OBJECTIVE To evaluate accuracy and cost of non-invasive diagnostic strategies including magnetic resonance imaging, intestinal ultrasonography, ileocolonoscopy and video-capsule endoscopy in suspected Crohn's disease. METHODS A decision-analytic model was used to assess the costs in low (25%), intermediate (50%) or high (75%) pre-test probability of Crohn's disease. Based on the published accuracy of diagnostic modalities and Bayes' rule, we calculated post-test probability of Crohn's disease using different strategies, starting from ileocolonoscopy, ultrasonography or magnetic resonance. Each strategy was considered successful when post-test probability was >95% or <5%. RESULTS With low pre-test probability, only ileocolonoscopy as the first investigation could exclude or confirm Crohn's disease while a normal ultrasonography may exclude Crohn's disease. With high pre-test probability, ileocolonoscopy or ultrasonography as the first test may confirm Crohn's disease, but at least 3 negative tests are required to exclude Crohn's disease. The cost to diagnose one patient was cheapest utilising an ultrasonography-based strategy both in low (ultrasonography €1076; ileocolonoscopy €2005; magnetic resonance €4515) and high pre-test probability of Crohn's disease (ultrasonography €321; ileocolonoscopy €712; magnetic resonance €1412). CONCLUSION The accuracy and cost of these strategies depend on pre-test probability of Crohn's disease and vary according to the first test used. Ileocolonoscopy plus ultrasonography is the most accurate and less expensive initial diagnostic strategy.