[9] In addition, this multicenter study included patients with CD from 13 hospitals nationwide. Most patients were referred by primary care physicians and were subsequently diagnosed and treated by IBD specialists at each hospital. These factors may attenuate the recruitment bias. Second, the follow-up period of CD patients was relatively short and variable. It may affect the cumulative rate of CD-related surgery, which was presented in Figure 1. Most censorings happened during

the first 10 years and the steep rise of Kaplan–Meier curve occurred after that. However, we did not anticipate that it would greatly influence the results

selleck chemical in assessment of predictors for the clinical outcomes because Kaplan–Meier and Cox regression methods measure the proportion of patients over a period of time for each group. Additionally, only patients with a follow-up period of more than 6 months were included in our cohort. Considering a part of CD patients presented with severe disease requiring biologics in the initial course of disease, the inclusion criteria of our study may cause potential bias, which can influence on the results. Finally, because this study was conducted in a retrospective manner, we could not control all confounding factors in the analysis. Especially, variety of confounding factors such as different indication or timing for starting cAMP immunosuppressants (azathioprine AZD0530 or 6-mercaptopurine) or biologics (infliximab)

according to doctors might influence the results when analyzing factors related to use of these agents. Despite these limitations, the strength of this study is that it was a large, multicenter cohort study to identify predictive factors associated with clinical outcomes in the Korean population. In addition, we identified variables associated with three different end-points (first surgery, need of immunosuppressive agents, or biological agents) in one study. In conclusion, the present study identified stricturing, penetrating disease behavior, and smoking habits at the time of diagnosis as independent predictors for a first CD-related surgery. In addition, we also found that younger age (< 40 years), ileal involvement, and perianal disease at diagnosis are associated with the need for immunosuppressive or biological agents. Given the differences in pathophysiology and clinical aspects with different ethnicities, our results may characterize the natural disease course in Korean CD patients and be useful to assess risk, predict the clinical outcomes, and determine optimized treatment plans for these patients.