Comparison of bleeding risk scores and evaluation of major bleeding predictive factors in patients with major bleeding due to vitamin K antagonist use

BACKGROUND: Major bleeding in the treatment of atrial fibrillation is closely associated with an increased risk of death and major adverse outcomes in both the short and long term, but all bleeding events are associated with a reduced quality of life. Bleeding events are also known to reduce medication adherence. In this sense, bleeding risk scores are important tools to help predict major bleeding. However, it is not clear which scoring system is superior. AIM: In this study, our aim was to compare bleeding risk scores and to examine the factors associated with bleeding in patients with major bleeding while using vitamin K antagonists. METHODS: In this retrospective and single-center study, scoring, laboratory and demographic data were analyzed with SPSS 20.0 statistical program. RESULTS: The mean age of a total of 1434 patients included in our study was 68.2 ± 11.3 years, range was 39–93 years and 769 (53.6%) of these patients were male. Of 588 patients with major bleeding, 93 (15.8%) had intracranial hemorrhage. Logistic regression analysis comparing the scoring systems among themselves revealed that the GARFIELD-AF scoring system had a predictive effect on major bleeding independent of the effect of other scoring systems (OR: 1.532, 95% CI 1.348–1.741, p < 0.001). The area under the curve (AUC) for GARFIELD-AF was 0.690 (0.662–0.718) as a result of the ROC analysis considering the best cut-off point of 3.2% calculated for 2 years. AUC 0.659 (0.630–0.687) for HAS-BLED, AUC 0.636 (0.606–0.665) for ORBIT and AUC 0.611 (0.5810.642) for ATRIA. When we compare the patient group with the control group, it can be said that intracranial hemorrhage occurred independently of INR and TTR values, unlike in the major bleeding group (p:0.129, p:0.545). CONCLUSION: In patients using vitamin K antagonists for atrial fibrillation, the GARFIELD-AF risk score was found to be superior to important bleeding risk scores such as HAS-BLED, ORBIT and ATRIA in terms of predicting major bleeding. It is an important result that intracranial hemorrhages, which have a special place among major hemorrhages, were independent of INR and TTR levels. It is noteworthy that 8.2% of patients with major bleeding had a history of minor bleeding in the last year.