Baseline cognitive function and clinical outcomes with integrated care management based on the ABC pathway in patients with atrial fibrillation: insights from MIRACLE-AF Trial
M Li, Y Chen, M Chu, S Zhang, S Yang, Y Shen, G Y H Lip, M ChenAbstract
Background
The MIRACLE-AF trial was a prospective cluster-randomised study evaluating a telemedicine-based, village doctor-led implementation of the Atrial fibrillation Better Care (ABC) pathway in older patients with atrial fibrillation (AF) in rural China, demonstrating a significant reduction in the primary composite outcome compared with usual care. Cognitive impairment is common in patients with AF and may affect adherence and delivery of integrated care. Whether baseline cognitive function modifies treatment effects remains unclear.
Purpose
We examined clinical outcomes across levels of baseline cognitive function in the MIRACLE-AF trial.
Methods
Baseline cognition was assessed using the Mini-Mental State Examination (MMSE). Interaction between continuous MMSE and treatment allocation was examined using Cox models with restricted cubic splines, followed by linear interaction analyses. MMSE was categorised as <18, 18–23, and ≥24. Clinical outcomes were prespecified MIRACLE-AF endpoints. Cox proportional hazards models with a frailty term for clustering were used to estimate adjusted hazard ratios (aHRs) with adjustment for AF type, CHA2DS2-VASc score, and educational attainment. Interaction terms were examined across cognitive strata.
Results
Among 1,028 participants with complete MMSE data (median follow-up 2.8 years), 37.7% had MMSE <18 (mean age 78.0 ± 5.5 years; 260 [67.0%] female), 29.7% had MMSE 18–23 (mean age 74.9 ± 5.5 years; 142 [46.6%] female), and 32.6% had MMSE ≥24 (mean age 74.0 ± 5.0 years; 53 [15.8%] female). Restricted cubic spline analyses showed no significant nonlinear interaction between continuous MMSE and treatment allocation. In linear models, each 3-point increase in MMSE was associated with an approximate 5% relative enhancement in treatment effect, though not statistically significant (P_interaction=0.459). For the primary outcome, the aHR was 0.86 (95% CI 0.54–1.35) in MMSE <18, 0.56 (95% CI 0.34–0.92) in MMSE 18–23, and 0.55 (95% CI 0.32–0.93) in MMSE ≥24 (interaction P=0.336). Cumulative event rates were consistently lower with intervention across strata.
Conclusions
The MIRACLE-AF integrated care intervention was associated with favourable outcomes across cognitive strata. Greater relative risk reductions were observed among patients with preserved or mildly impaired cognition, with attenuated estimates among patients with severe cognitive impairment.Baseline MMSE and Treatment InteractionBaseline MMSE-stratified analysis