Vitamin K antagonists (VKAs) reduce cardiovascular events (CVEs) in atrial fibrillation (AF) when a time in therapeutic range (TiTR) >70% is achieved. Factors affecting the time to achieve the TR (TtTR) are unknown.

Prospective observational study including 1,406 nonvalvular AF patients starting VKAs followed for a mean of 31.3months (3,690 patient/year); TiTR, TtTR, and SAMe-TT₂R₂ score were calculated, and CVEs were recorded.

Median TtTR was 8.0days (interquartile range 5.0-18.0). Patients with high TtTR (ie, >75th percentile) were more likely to be in AF than in sinus rhythm at entry (odds ratio [OR]: 1.423, P=.011).

Median TiTR was 60.0%; low TiTR (below median) was associated with SAMe-TT₂R₂ score (OR: 1.175, P=.001), high TtTR (>75th percentile, OR: 1.357, P=.017), and number of international normalized ratio checks (OR: 0.998, P=.049). We recorded 113 CVEs (3.1%/y), with a higher rate seen in patients with TtTR >75th percentile compared to those below (log-rank test, P=.006). A multivariable Cox regression analysis showed that SAMe-TT₂R₂ score (hazard ratio [HR]: 1.331, P<.001), TtTR >75th percentile (HR: 1.505, P=.047), TiTR <70% (HR: 1.931, P=.004), number of international normalized ratio checks (HR: 0.988, P<.001), digoxin (HR: 1.855, P=.008), and proton-pump inhibitors (HR: 0.452, P<.001) were independently associated with CVEs.

High TtTR is associated with poorer long-term quality of VKAs therapy. Patients with TtTR >18days or with high SAMe-TT₂R₂ score should be considered for treatment with non–vitamin K oral anticoagulants.