Increased bleeding risk with bridging and early anticoagulation in traumatic brain injury patients with intracranial hemorrhage and mechanical heart valves.

BACKGROUND: The role of bridging anticoagulation with unfractionated heparin (UFH) in traumatic brain injury (TBI) patients with mechanical heart valves (MHV) remains uncertain. While bridging has traditionally been used to reduce thromboembolic risk, emerging evidence suggests it may increase bleeding complications without clear benefit. This study evaluates the impact of bridging and anticoagulation resumption timing on clinical outcomes. METHODS: A retrospective observational study was conducted at Memorial Hermann–Texas Medical Center, including 44 adult patients with MHV and acute TBI previously on warfarin. Patients were categorized into bridging (n = 22) and non-bridging (n = 22) groups. Anticoagulation resumption was also compared between early (< 14 days, n = 37) and delayed (> 14 days, n = 7) initiation groups. The primary outcome was hematoma expansion leading to clinical deterioration or death. Secondary outcomes included thromboembolic events, extracranial bleeding, and anticoagulation-related laboratory abnormalities. RESULTS: Hematoma expansion was more frequent in the bridging group (27.3%) compared to the non-bridging group (13.6%), and four bridging patients (18.2%) required blood transfusions. Thromboembolic events occurred in one bridging (4.5%) and two non-bridging patients (9.1%). Among those who resumed anticoagulation within 14 days, four experienced hematoma expansion and four had extracranial bleeding, whereas none in the delayed resumption group (> 14 days) had these complications. Laboratory findings showed higher supratherapeutic INR (> 3.5) and PTT (> 90) in the bridging group. CONCLUSION: Bleeding complications were more frequent in the bridging group, while thromboembolic events were slightly more common in the non-bridging group. Early anticoagulation resumption (< 14 days) appeared to carry a higher complication rate, whereas delaying beyond 14 days was associated with fewer observed events. These findings should be interpreted cautiously given the small sample size.