Abstract
Background
Atrial fibrillation (AF) and heart failure (HF) are common co-existing conditions. Randomized trial data suggests a reduction in all-cause mortality with catheter ablation (CA) in selected patients, however, whether these results are replicable in a real-world population and persist in the long-term remains to be shown.
Purpose
To evaluate the long-term effectiveness of CA in AF-HF patients in reducing the incidence of: a) all-cause mortality b) HF hospitalizations, and c) major morbidities (stroke/transient ischemic attack (TIA) and major bleeding).
Methods
A population-based administrative cohort was created of AF-HF patients with government prescription coverage in Quebec, Canada (1999–2015). Patients who underwent CA (cases) were matched 1:2 to controls using risk-set sampling. Cases were matched on time in the cohort and frequency of hospitalizations. Measured time-invariant confounders were controlled for using inverse probability of treatment weighting (IPTW) and included age, sex, clinical characteristics, presence of cardiac implantable electronic devices, and medication use. Multivariable Cox models adjusted the association of CA with the outcomes for the time varying confounders of the presence of an implantable cardioverter defibrillator (ICD) or cardiac resynchronization therapy (CRT), anticoagulation use (warfarin or direct oral anticoagulation), and any antiarrhythmic (AAD) use during follow-up. For non-fatal outcomes, the competing risk of death was accounted for using the Lunn-McNeil approach.
Results
Of the 87,676 AF-HF patients, 298 underwent CA and were matched to 591 controls. After IPTW, the distribution of covariates was balanced between cases and controls [age 65.6±11.0 vs 61.6±11.6; women 24% vs 20%; CHA2DS2-Vasc score 3.2±2.3 vs 2.9±2.1; CA vs non-CA, respectively; standardized mean differences <0.1 for all]. Over a median follow-up of 3.3 (IQR 1.1–6.4) years, 19 (7.3%) of CA patients died compared to 144 (24.6%) non-CA patients. After weighting and adjustment, CA was associated with a statistically significant reduction in the incidence of all-cause mortality [adjusted HR 0.5 (95% CI 0.3–0.9)]. In addition, there was no statistically significant difference in the incidence of HF hospitalizations over the follow-up [CA: 22.5% vs non-CA: 27.1%; adjusted HR 0.9 (95% CI 0.6–1.2)]. The incidences of stroke/TIA (1.7% vs 6.8%) and major bleeding (1.7% vs 4.9%) for CA vs non-CA were not statistically different.
Conclusion
In a matched population-based AF-HF cohort, CA was associated with a reduced risk of all-cause mortality compared to patients who did not undergo CA. Although no difference in the risk of HF hospitalizations, stroke/TIA, and major bleeding was detected between CA and non-CA patients, larger studies are warranted.
Acknowledgement/Funding
Canadian Institute of Health Research; Fonds de recherché du Quebec-Santé, Clinical Research Scholar Award (V. Essebag) and Doctoral Award (M. Samuel)
Effectiveness of structured, hospital-based, nurse-led atrial fibrillation clinics: a comparison between a real-world population and a clinical trial population