New predictive model of rebleeding during follow-up of patents with obscure gastrointestinal bleeding: A multicenter cohort study.

Niikura, Ryota; Yamada, Atsuo; Nagata, Naoyoshi; Kobayashi, Yuka; Okamoto, Makoto; Mitsuno, Yuzo; Ogura, Keiji; Yoshida, Shuntaro; Shimbo, Takuro; Hirata, Yoshihiro; Koike, Kazuhiko

BACKGROUND AND AIM The long-term recurrence rate of patients with obscure gastrointestinal bleeding (OGIB) who underwent video capsule endoscopy (VCE) remains unknown. Our study aimed to identify the cumulative incidence of rebleeding, develop a predictive model of rebleeding, and evaluate whether the model can be applied in other outcomes among patients with OGIB. METHODS We conducted a multicenter retrospective cohort study of 320 patients with OGIB who underwent VCE between 2009 and 2014. A rebleeding model was developed using multivariate Cox proportional hazards analysis and evaluated using Harrell's c-index. RESULTS Rebleeding occurred in 43 patients (13.4%) during a mean follow-up period of 18.3 (standard error, 0.9) months. The rebleeding sources were the small intestine (n = 17), extra-small intestine (n = 13), and unknown (n = 13). The cumulative incidence of rebleeding was 11.0% at 12 months and 35.3% at 60 months. Multivariate analysis revealed that female gender, liver cirrhosis, warfarin use, overt bleeding, and positive VCE findings were significant predictors of rebleeding. The rebleeding rate was 0% in patients with no predictors and 40% in patients with full predictors (P < 0.01). The model presented a high predictive accuracy (c-statistic, 0.733). Patients with higher predictors exhibited higher transfusion requirements, longer length of stay, and higher mortality (P < 0.01). CONCLUSION The cumulative incidence of rebleeding at 12 and 60 months was 11.0% and 35.3%, respectively. Five factors enable prediction of not only rebleeding but also transfusion requirements, length of stay, and death in OGIB patients.