<![CDATA[Bleeding complications are a major contributor to hospitalization among cardiovascular patients on warfarin therapy, yet the causative factors remain incompletely characterized. This retrospective case-control study aimed to identify significant risk factors and their potential interaction in bleeding incidence among cardiovascular disease (CVD) patients at RSUD Kabupaten Tangerang. Medical-record data from CVD patients on warfarin hospitalized with or without bleeding between January 2023 and December 2024 were analyzed using JAMOVI, encompassing bivariate, multivariate, and interaction analyses. A total of 132 patients were enrolled, comprising 44 (33.3%) cases with bleeding and 88 (66.7%) controls without bleeding. Fisher's exact test, employed due to sparse and unbalanced data, identified significant associations with herb use (p<0.001, OR=460), INR ≥3 (p<0.001, OR=8.33), and hypofibrinogenemia (p<0.001, OR=179). Firth's penalized logistic regression confirmed herb use (aOR=25.33, p<0.001) and hypofibrinogenemia (aOR=19.10, p<0.001) as independently and strongly associated with bleeding, with a strong model fit (Nagelkerke R²=0.854). However, the likelihood ratio test for the synergistic interaction between these two factors was not statistically significant (χ²=2.78, p=0.095). Overall, herb use and hypofibrinogenemia are independently risk factors for bleeding in warfarin-treated cardiovascular patients; the interaction between them was not statistically significant, so clinicians should screen for herb use and monitor fibrinogen levels.]]>
