BackgroundEstimation of bleeding risk is an unmet need for individualized therapy in acute pulmonary embolism (PE) patients with increased mortality risk. MethodsWe analyzed the association between various patients' characteristics and occurrence of major bleeding (MB) according to the modified International Society of Thrombosis and Hemostasis (ISTH) criteria (“overt” bleeding is the only modification from the original criteria) at 7 days from admission to the hospital and thrombolytic therapy with a tissue-plasminogen activator (tPA). Pulmonary embolism bleeding score index (PEBSI) was created using multivariate regression analyses, and finely, dichotomous index was used for the discrimination of patients with low risk for MB from those with high risk. ResultsDuring the 6-year period (2015–2021) 367 PE patients were treated with tPA and included in the Regional PE registry. Among them, 29 (7.9%) fulfilled the criteria for MB. Five factors were identified as significantly associated with MB and were used to build the PEBSI score: previous bleeding, recent surgery, diabetes, the use of drugs that could be associated with bleeding, and anemia. PEBSI score showed c-index for 7-day MB 0.794 (95CI% 0.698–0.889). Patients with PEBSI scores of 0 or 1 had a low risk for MB (2.8%) and those with scores>1 had a high risk for MB (18.6%) (p < 0.001). Internal validation of PEBSI score using a randomly, equally split method confirmed the discriminative value of the PEBSI score. ConclusionNovel PEBSI score has significant power to discriminate patients with low risk for MB on thrombolytic therapy from those with high risk.