How should we select suspected Crohn's disease patients for capsule enteroscopy?

Background: In suspected Crohn's disease (CD), non-diagnostic ileocolonoscopies are often followed by small bowel capsule endoscopy (SBCE). Adequate pre-selection of patients for SBCE is a key to optimize allocation of resources. We aimed to establish a rational approach for the CD diagnostic workflow, based on biochemical profile of patients with suspected CD, targeting an optimization of patients' selection for SBCE. Methods: Multicenter cohort study includes consecutive patients with suspected undergoing SBCE after non-diagnostic ileocolonoscopy. Minimum follow-up period after the capsule enteroscopy was six months. The outcome was confirmation of CD diagnosis. Univariate analysis and logistic regression were performed. Results: In included 220 patients, 62.3% of women were with a mean age of 41 years [26-54]. A confirmed diagnosis of CD was established in 98 patients (44.5%). The initial univariate analysis identified variables above the threshold of marginal statistical association toward CD diagnosis ( p  < .15). The regression model identified high CRP levels (OR 1.028 p  = .128) and low serum Iron (OR 0.990 p  = .025) as the independent variables with consistent correlation with CD diagnosis. Those two variables present a suitable discriminative power (AUC = 0.669, p  < .001) for the diagnosis of CD. Conclusion: In suspected CD, low serum iron and elevated CRP had a statistically significant association with CD diagnosis, being helpful to identify patients with higher CD probability before SBCE. However, the lack of a proper validation of the model leads us to currently recommend SBCE to all patients with suspected CD and negative ileocolonoscopy, as no specific biochemical profile can be used to confidently exclude small bowel CD.