J Thromb Thrombolysis. 2026 Jul 23. doi: 10.1007/s11239-026-03361-w. Online ahead of print.
ABSTRACT
Severe sepsis-associated thrombocytopenia creates a difficult clinical trade-off between thrombosis prevention and bleeding risk, yet evidence to guide pharmacological prophylaxis when platelet counts are 30,000-50,000/µL remains limited. We conducted a target trial emulation using the MIMIC-IV database, classifying eligible adults with sepsis upon first intensive care unit admission based on whether a predominantly unfractionated heparin (UFH)-based prophylaxis strategy was administered within a 24-hour grace period after platelet counts initially entered this range. Among 938 eligible patients (189 prophylaxis strategy, 749 control strategy), venous thromboembolism (VTE) occurred in 14 and 86 patients, respectively. In the inverse probability-weighted Cox analysis, early administration of this UFH-based prophylaxis was associated with a lower 28-day VTE hazard (adjusted hazard ratio 0.51, 95% confidence interval 0.28-0.92; P = 0.025), though this association was attenuated in certain prespecified sensitivity analyses. Electronic health record-derived major bleeding occurred in 22 and 95 patients, respectively, and no statistically higher hazard was observed under the prophylactic strategy (adjusted hazard ratio 0.74, 95% confidence interval 0.46-1.20; P = 0.220). These findings suggest that early administration of a predominantly UFH-based prophylactic strategy may be associated with a lower VTE risk, without a statistically higher hazard of the electronic health record-derived major bleeding endpoint. Given the observational design, this efficacy signal remains hypothesis-generating rather than definitive and warrants prospective validation.
PMID:42489962 | DOI:10.1007/s11239-026-03361-w
