Risk Stratification of Patients with Crohn's Disease: A Retrospective Analysis of Clinical Decision Making and Its Impact on Long-Term Outcome.

BACKGROUND AND AIMS: Complications such as need for bowel resections and hospitalization due to Crohn's disease (CD) occur when disease activity persists due to ineffective therapy. Certain "high-risk" features require an early introduction of anti-tumor necrosis factor-α therapy to prevent such complications. We aim to evaluate the prevalence of "high-risk" features among a cohort of patients with CD and examine the association between discordance of early therapy with baseline risk stratification and disease outcome.

PATIENTS AND METHODS: All adult patients with CD were retrospectively identified and their medical records were reviewed. Clinical, endoscopic, laboratory, and radiological data were collected. Patients were divided into "low" and "high" risk groups according to the presence or absence of penetrating disease, perianal involvement, foregut involvement, extensive disease seen on endoscopy or cross-sectional imaging, young age at the time of diagnosis (<40), persistent cigarette smoking and frequent early requirements for corticosteroid therapy. Initial treatment selection and treatment approach ("step-up" vs. "accelerated step-up" vs. "top-down") within 6 months of diagnosis were recorded. Rates of CD-related bowel resections and hospitalization within 5 years of diagnosis were calculated. Logistic regression analysis was used to examine the association between "discordance" of early treatment selections and risk stratification categories with outcomes.

RESULTS: Eighty-five CD patients were included. The median age and duration of disease were 25 (interquartile range [IQR] 19-32) and 5 (IQR 4-6) years, respectively. Sixty five percent were females and 66% were native Saudis. Smoking was reported in 12% of patients and perianal disease in 18%. "High-risk" features were identified in 43 (51%) patients, of which only 6 (14%) were treated with "top-down" therapy and 7 (16%) with "accelerated step-up" care. The risk of requiring a bowel resection, and hospitalization was higher for "high-risk" patients compared to "low-risk" patients (risk ratio [RR] 13.67, 95% CI 1.88-99.41; p = 0.003, and RR 1.86, 95% CI 0.03-0.43; p = 0.0312, respectively). "Discordance" occurred in 34% of cases. Bowel resection was required in 15/85 (18%) patients and 32/85 (38%) required at least one hospitalization within 5 years of diagnosis. Logistic regression analysis identified a statistically significant association between "discordance" and need for bowel resections (OR 6.50, 95% CI 1.59-26.27, p = 0.009), and hospitalizations (OR 3.01, 95% CI 1.08-8.39, p = 0.035) within 5 years of diagnosis.

CONCLUSIONS: "Discordance" between patient risk-profile and treatment selection early in the course of CD has a significant impact on disease outcome, specifically need for bowel resection and hospitalization, which are more likely to occur in the presence of "high-risk" features. Early identification of "high-risk" features could help prevent long-term complications.