External validation of a prediction score for identifying a mild course of Crohn's disease

Crohn's disease has a heterogeneous course. Early identification of patients likely to follow a mild trajectory remains challenging. The prediction score for a mild course of Crohn's disease (PreMiCC) was recently proposed, but external validation is lacking.

This single-center retrospective study included patients diagnosed with Crohn's disease between 2004 and 2024. The PreMiCC score was calculated at diagnosis using the original variables: age, C-reactive protein, perianal disease, endoscopic severity, and complications. Patients were classified as having a score ≤2 or >2. Outcomes included mild disease course, need for biologic therapy, Crohn's disease-related surgery or hospitalization, and time to treatment escalation.

A total of 342 patients (51% male; mean age 35 ± 16 years) were included, with a mean follow-up of 9 ± 5 years. At diagnosis, 55% had a PreMiCC score ≤2. A mild disease course occurred in 31% of patients and was more common in those with a score ≤2 than >2 (42% vs. 17%; P < 0.001). Patients with a score >2 more frequently required biologic therapy (80% vs. 54%; P < 0.001) and hospitalization (15% vs. 7%; P = 0.023), whereas the need for surgery did not differ between groups (20% vs. 17%; P = 0.415). The PreMiCC score independently predicted biologic therapy [odds ratio = 1.65, 95% confidence interval (CI) = 1.33-2.04; P < 0.001]. Time to treatment escalation was shorter in patients with a score >2 (log-rank P < 0.001). The model showed acceptable discrimination (area under the receiver operating characteristic curve = 0.708, 95% CI = 0.649-0.768), with 74% sensitivity, 55% specificity, 42% positive predictive value, and 83% negative predictive value.

This first external validation confirms the predictive value of the PreMiCC score in a real-world cohort, supporting its use as a simple tool for early risk stratification in Crohn's disease.