Comparative Effectiveness and Safety of Interventions for Atrial Fibrillation in Heart Failure: A Network...

Patients with HF and AF face high risks of death and HF decompensation, yet the comparative effectiveness of commonly used AF management strategies remains uncertain. This study is aimed at comparing the efficacy and safety of different interventional and medical approaches using NMA of RCTs.

PubMed, Embase, CENTRAL, Web of Science, and Google Scholar were searched from inception to November 30, 2025, without language restrictions. Parallel-group RCTs enrolling adults with clinically diagnosed HF and documented AF were eligible if they compared at least two AF management strategies, including CA, AVNA with CRT or RV pacing, MRC, drug-based rhythm control (rhythm drug), or CON. Primary outcomes were all-cause mortality and HF hospitalization; secondary outcomes were worsening HF, CV mortality, QoL, and LVEF. Random-effects NMA was performed in Stata 17.0, reporting ORs and SMDs with 95% CIs; treatments were ranked using SUCRA.

Twenty-three RCTs (5159 patients) were included. For all-cause mortality (21 trials; 5095 patients), AVNA_CRT (OR, 0.30; 95% CI, 0.12-0.78) and CA (OR, 0.53; 95% CI, 0.35-0.82) were associated with lower mortality than CON; AVNA_CRT also outperformed MRC (OR, 0.35; 95% CI, 0.15-0.79) and AVNA_RV (OR, 0.49; 95% CI, 0.25-0.96). For HF hospitalization (9 trials; 3491 patients), AVNA_CRT reduced risk compared with AVNA_RV (OR, 0.20; 95% CI, 0.04-0.94). For worsening HF (7 trials; 2542 patients), AVNA_CRT was superior to AVNA_RV (OR, 0.41; 95% CI, 0.20-0.84). CA improved QoL versus AVNA_CRT (SMD, -0.94; 95% CI, -1.83 to -0.05) and MRC (SMD, -0.73; 95% CI, -1.22 to -0.24), and improved LVEF versus MRC (SMD, 1.49; 95% CI, 0.57-2.41). No between-strategy differences were clearly demonstrated for CV mortality.

In RCT evidence for HF with AF, AVNA_CRT and CA were associated with more favorable prognostic outcomes, and CA provided more consistent improvements in QoL and LVEF. When AVNA is selected, CRT should be preferred over RV pacing. Strategy choice should be individualized to clinical goals, patient phenotype, and procedural feasibility.