乌鸦传媒

This article is a collaboration between 乌鸦传媒 and:

Analysis of a Taiwanese registry turned up reduced stroke and embolic risks among patients on nearly universally low doses of non-vitamin K oral anticoagulants (NOACs) compared with warfarin (Coumadin).

Reflecting concerns over higher bleeding risk among Asian populations, fully 87% of the 3,916 rivaroxaban users in the country and 90% of the 5,921 dabigatran users there were taking low doses (10 to 15 mg once daily and 110 mg twice daily, respectively), Lai-Chu See, PhD, of Chang Gung University in Taoyuan, Taiwan, and colleagues .

Compared with the 5,251 warfarin users in the nationwide retrospective cohort study of consecutive patients with non-valvular atrial fibrillation in the Taiwan National Health Insurance Research Database from February through December 2013, outcomes were significantly better in both propensity-matched and unmatched analyses with the two NOACS for:

• Ischemic stroke or systemic embolism
• Intracranial hemorrhage
• All-cause mortality

"The consistency is reassuring," of Brigham and Women's Hospital in Boston, wrote in an accompanying editorial, noting that the clinical trials with these drugs had shown stroke, mortality, and intracranial hemorrhage reductions against warfarin overall as well across a wide range of patient subgroups.

"So, do these results support the investigators' conclusion that we should consider routinely prescribing the lower doses of dabigatran and rivaroxaban for Asian patients because of their superior efficacy and safety profile compared with warfarin in this study," he continued, "even though the comparative efficacy findings were discordant with the clinical trial data in Asian patients, in which only dabigatran 150 mg demonstrated significant reductions in ischemic stroke but not dabigatran 110 mg or full doses of rivaroxaban, apixaban, or edoxaban?"

"Unfortunately, we cannot reconcile these differences because the investigators do not provide any data on how well warfarin was used.... But as is often the case with real-world data, we are often left with more questions than answers," Ruff concluded.