Clinical Impact of Initial Low Molecular Weight Heparin Use in Hospitalised Patients with Venous Thromboembolism Treated with Direct Oral Anticoagulants: A Propensity Score Matched Study.

Hsu, Jung-Chi; Huang, Chen-Yu; Lin, Donna Shu-Han; Chen, Zheng-Wei; Liang, Huai-Wen; Lee, Jen-Kuang · Eur J Vasc Endovasc Surg · 2026

retrospective_cohort · Level III

Where this comes from

Abstract

Treatment of venous thromboembolism (VTE) with direct oral anticoagulants (DOACs) follows established, agent specific, low molecular weight heparin (LMWH) initiation protocols. However, adherence to these protocols may be suboptimal in real world practice. This study evaluated the association between protocol adherence and thirty day clinical outcomes in hospitalised patients with VTE receiving DOAC therapy. A retrospective cohort study using Taiwan's national hospitalisation database from 2012 to 2022 was conducted. Adults hospitalised for VTE and newly prescribed DOAC were included. Propensity score matching was performed by DOAC type based on LMWH initiation status. Among 29 182 patients, 6 683 received dabigatran/edoxaban and 22 499 received apixaban/rivaroxaban. After matching, 2 135 pairs in the dabigatran/edoxaban cohort and 9 858 pairs in the apixaban/rivaroxaban cohort were analysed. In the dabigatran/edoxaban group, LMWH lead in was not associated with recurrent VTE (2.2% vs. 1.5%; subdistribution hazard ratio [SHR] 1.41, 95% confidence interval [CI] 0.90 - 2.21), major bleeding (0.9% vs. 0.7%; SHR 1.33, 95% CI 0.69 - 2.54), or all cause mortality (5.3% vs. 4.5%; hazard ratio [HR] 1.16, 95% CI 0.89 - 1.53). In the apixaban/rivaroxaban group, LMWH lead in was associated with a higher risk of recurrent VTE (2.6% vs. 1.4%; SHR 1.81, 95% CI 1.48 - 2.23), but not with major bleeding (0.7% vs. 0.6%; SHR 1.18, 95% CI 0.83 - 1.69) or all cause mortality (4.4% vs. 4.2%; HR 1.04, 95% CI 0.91 - 1.19). The effect of LMWH vs. non-LMWH did not differ significantly between the two pharmacological groups. In this nationwide observational cohort, initiation with LMWH was not associated with differences in short term thrombotic or bleeding outcomes among dabigatran or edoxaban users. In contrast, among apixaban or rivaroxaban users, LMWH initiation was associated with a higher risk of recurrent VTE without increasing major bleeding. However, the estimates were not significantly different between the two pharmacological groups. Prospective studies and external validation are warranted.