Alcohol use disorder and initiation of oral anticoagulant therapy in patients with atrial fibrillation: a nationwide cohort study

Underuse of oral anticoagulant (OAC) therapy is prevalent among patients with atrial fibrillation (AF), and disparities in treatment are noted across various vulnerable demographic groups. Alcohol use disorders (AUD) are common and are often associated with severe multimorbidity, which emphasizes the need for comprehensive management of both alcohol-related and other somatic conditions to improve overall health outcomes. Prior research indicates that patients with AUD receive poorer treatment for their somatic comorbidities compared to those without these conditions. However, it is currently unknown whether AUD affects OAC initiation for stroke prevention in patients with AF.

Our study aimed to fill this knowledge gap by exploring whether AUD influences OAC initiation among patients with AF, using a nationwide dataset to assess treatment disparities.

The Finnish AntiCoagulation in Atrial Fibrillation (FinACAF) registry includes data on 229,565 patients diagnosed with incident AF in Finland between 2007 and 2018, identified from national registries covering all levels of care. The definition of AUD was based on ICD-10 and ICPC-2 codes from all levels of care to enhance the comprehensive identification of patients likely affected by AUD. The main outcome was OAC initiation, and its occurrence was compared between patients with and without AUD.

The mean age of patients was 72.7 years and 50% were female. Altogether, 85.0% had a CHA2DS2-VA score ≥ 1 and were therefore eligible for OAC therapy, and 4.7% had AUD. Patients with AUD demonstrated a higher prevalence of diabetes, liver failure, renal failure, and previous bleeding. OAC therapy was initiated in a smaller proportion of patients with AUD compared to those without (52.5% vs. 71.4%, p<0.001). This disparity remained significant even after adjusting for comorbidities, socioeconomic status, and laboratory values influencing OAC initiation (adjusted hazard ratio [HR] 0.68; 95% confidence interval [0.66-0.71]). The absolute difference in OAC initiation between groups stayed consistent at approximately 20% throughout the study period (Figure 1). Interaction analyses indicated that the association between AUD and lower OAC initiation was stronger in patients with lower income, lower educational attainment, and lower thrombocyte levels, but there was no significant interaction with sex, bleeding history, or hemoglobin levels.

Patients with AUD are significantly less likely to receive OAC therapy compared to those without AUD.

OAC initiation in one year follow-up