General practice-related variation in oral anticoagulant treatment of atrial fibrillation: a nationwide cohort study

Background Guideline-adherent oral anticoagulant treatment (OAC) in atrial fibrillation (AF) remains a challenge. In Denmark, most patients with AF are treated in general practice. Nevertheless, determinants of OAC prescription in primary care are poorly understood.

Aim To investigate variation in OAC adherence between general practice clinics and identify clinic characteristics associated with a lower propensity to prescribe OAC.

Design & setting Nationwide register-based cohort study including prevalent and incident patients with AF and CHA₂DS₂-VASc score≥2 (n=165,731) listed with Danish general practice clinics (n=1666) in 2021.

Method The main outcome was OAC adherence assessed as proportion of days covered. We used clinic OAC propensity to evaluate variation. OAC propensity was quantified as ratios between observed and expected adherence. Expected adherence was estimated based on the composition of the clinic patient populations. Sampled reference populations were constructed to account for random variation. Linear regression models examined associations between OAC propensity and clinic characteristics.

Results The proportion of days covered with OAC in the AF-population was 78%. OAC propensity in clinics in the 90^th percentile was 20% higher compared to clinics in the 10^th percentile, however this difference was reduced to 3% when accounting for random variation. Modest associations were observed between clinic characteristics and OAC propensity. The most significant difference was in the correlation between geographic location and OAC propensity, showing an 8% gap between top- and bottom-performing regions.

Conclusion The study suggests persistent underutilisation of OAC in AF patients and little variation in OAC prescription patterns across general practice clinics.