Available bleeding scoring systems poorly predict major bleeding in the acute phase of pulmonary embolism

Abstract Background Bleeding prediction scores may help to guide acute management of patients with pulmonary embolism (PE). However, existing scoring systems have not been validated for in-hospital assessment. We aimed compare 6 available bleeding scores, in a real-life cohort the major bleeding. Methods recorded characteristics 2,754 PE included prospective observational multicenter study contributing 18,028 person-days follow-up. assessed VTE-BLEED, RIETE, ORBIT, HEMORRA2HAGES, ATRIA, and HAS-BLED at baseline. ISTH-defined events were independently adjudicated. The accuracy was evaluated compared. Results observed 82 first (3.0% (95% CI, 2.4–3.7)). Overall, predictive power poor, C index ranging from 0.57 0.69 (Figure 1). RIETE score had numerically highest model fit best discriminatory capacity, but without reaching statistical significance versus HEMORR2HAGES, ATRIA scores. VTE-BLEED significantly lower indices, integrated discrimination improvement, net reclassification improvement compared four others. Conclusion Currently insufficient PE. development acute-PE-specific risk is needed optimally target that warrant bleeding-prevention strategies. FUNDunding Acknowledgement Type funding sources: None. Figure 1