Does non-persistence with optimised oral anticoagulant therapy post-discharge increase the risk of stroke in patients with non-valvular atrial fibrillation?
Approximately 1 in 8 patients with NVAF are non-persistent with hospital-optimised oral anticoagulant therapy within 12 months, which significantly increases their risk of stroke.
Abstract Background and aims Hospitals often optimise oral anticoagulant (OAC) therapy for patients with non-valvular atrial fibrillation (NVAF), yet it remains unknown whether this optimised therapy persists after discharge and improves stroke risk. We examined post-discharge persistence with optimised OAC therapy and its association with stroke in an Australian NVAF cohort. Methods We used nationally linked hospital and prescription data (2017–2022) for patients with NVAF whose in-hospital OAC therapy was optimised according to Australian AF guidelines. Non-persistence was defined as a ≥60-day gap in OAC supply after discharge on OAC, or any OAC re-initiation in low-risk patients discharged without OAC. Stroke was assessed from the index date until stroke, death, or 12 months. Predictors of non-persistence were assessed using multivariable logistic regression, and associations with stroke were examined using Cox proportional hazards models. Results Among 22,993 patients (mean age 74.7 ± 11.5 years; 43.2% female), 12.6% (n = 2,903; 95% CI: 12.2–13.1%) were non-persistent with optimised OAC therapy within 12 months. Chronic kidney disease and dementia increased the odds of non-persistence, whereas older age and hyperlipidaemia were associated with greater persistence. Non-persistence was associated with higher risks of stroke (adjusted hazards ration aHR: 1.42; 95% CI: 1.30–1.55). Conclusions One in eight patients was non-persistent with their optimised OAC therapy during hospitalisation, which significantly increased their stroke risk. Enhanced post-discharge follow-up is essential to maintain the benefits of in-hospital-optimised OAC therapy in patients with NVAF Conflict of interest Belayneh Kefale, Gregory M. Peterson, Corinne Mirkazemi, Woldesellassie M. Bezabhe:nothing to disclose
Gelaw et al. (Fri,) studied this question.