Anticoagulation therapy and related outcomes among Asian patients after bioprosthetic valve replacement.

IMPORTANCE

An inadequate number of studies focused on Asian populations have investigated the safety of warfarin usage among Asian patients with tissue valve aortic valve replacement (AVR) or mitral valve replacement (MVR).

OBJECTIVE

This study aimed to identify the optimal international normalized ratio (INR) range for Asian patients during a 1-year follow-up after tissue valve replacement.

DESIGN, SETTING, PARTICIPANTS: We conducted a retrospective cohort study of patients who underwent tissue valve AVR, MVR, and double valve replacement (DVR) between January 1, 2001, and December 31, 2018. Data were sourced from the Chang Gung Research Database, an electronic structured medical database covering 4 regional hospitals and 3 medical centers. The exposure of interest was INR level.

MAIN OUTCOMES AND MEASURES

The outcomes of primary and secondary interest were composite thromboembolic events and bleeding events during the 1-year follow-up, respectively. The relationship between INR level and the risk of thromboembolic events was explored using a logistic regression model in which the INR value was treated as a flexible restricted cubic spline. Because having atrial fibrillation (AF) greatly would greatly affect the INR control result, the analysis was stratified by AF status.

RESULTS

A total of 1059 participants were eligible for this study. The mean patient age was 65.5 (11.9) years; 592 (55.9%) participants were men, and 467 (44.1%) were women. A total of 447 had AF and 612 did not. The lowest bleeding risk was observed at an INR level around 1.9 to 2.0. An INR level of 1.84 (hazard ratio [HR], 0.49; 95% confidence interval [CI]: 0.36-0.67) and 1.7 (HR, 0.78; 95% CI: 0.62-0.99) corresponded to the lowest risk of thromboembolic events in patients with pre-existing AF and those without, respectively. The INR level corresponding to the lowest risk of thromboembolic events was approximately 1.7 in patients without AF but with MVR, DVR, or isolated AVR.

CONCLUSIONS AND RELEVANCE

For patients who underwent tissue valve replacement, the bleeding risk was elevated when the INR was greater than 2.0, but the risk of thromboembolic event increased only when the INR was lower than 1.84 in the AF group and 1.7 in the non-AF group, regardless of whether the patient received AVR, MVR, or DVR.