Net benefit of anticoagulation in subclinical device-detected atrial fibrillation: a Markov decision model analysis

Abstract Background The role of anticoagulation for stroke prevention in patients with device-detected atrial high-rate episodes, also known as subclinical atrial fibrillation (AF), is a subject of equipoise. Purpose To assess the net benefit of oral anticoagulant therapy in patients with device-detected subclinical AF. Methods A Markov decision model was constructed to estimate net impact of non-vitamin K antagonist oral anticoagulant (NOAC) therapy on the quality-adjusted life years (QALYs) of a patient with subclinical AF. The model incorporated new trial-based data on stroke, bleeding and mortality outcomes, as well as prior evidence on their effects on quality of life. The effect of anticoagulation on the severity of stroke and bleeding events was considered in the analysis (Figure 2). Incidence of clinical AF was also included in the model, with all patients exhibiting overt AF assumed to have the average stroke, bleeding, and mortality rates observed in real-world contemporary patients diagnosed with clinical AF. The main outcome measure was cumulative QALYs during a ten-year simulation. Deaths, ischemic strokes, and bleeding events were also calculated. Results Initiating NOACs resulted in 233 (21.7%) fewer ischemic strokes, 55 (1.1%) fewer deaths, and 453 (37.3%) more major bleeding events per 10 000 patients over a ten-year simulation period, compared to withholding anticoagulation in the same-sized cohort. Anticoagulation resulted in slightly fewer permanently disabling events. Per patient, these differences translated to approximately one additional quality-adjusted week of life (0.024 QALYs) with NOAC treatment during the ten-year simulation (Figure 1). When the confidence intervals of treatment effects were considered in probabilistic sensitivity analysis, there was a 65.8% probability that NOAC treatment leads to more QALYs than withholding treatment. Conclusions This Markov decision modeling study demonstrated that initiating NOACs in patients with device-detected subclinical AF may lead to a minimal increase in QALYs. However, the benefits are uncertain, and the effect size is not clinically significant.Cumulative quality-adjusted life years Markov model input parameters