Atrial fibrillation and the risk of death in patients with heart failure: a literature-based meta-analysis

Intern Med J. 2010 May;40(5):347-56. doi: 10.1111/j.1445-5994.2009.01991.x. Epub 2009 May 21.


Background: Heart failure (HF) and atrial fibrillation (AF) are common, associated with significant morbidity and mortality, and frequently coexist. It is uncertain from published data if the presence of AF in patients with HF is associated with an incremental adverse outcome. The aim of this study was to combine the results of all studies investigating prognosis for patients with HF and AF compared with those in sinus rhythm (SR) to asses the mortality risk associated with this arrhythmia.

Methods: Electronic databases were searched (Biological Abstracts, Current Contents, EMBASE, Medline, Medline In-progress, PubMed and Scopus), to 31 December 2006, using the key words congestive heart failure, heart failure, ventricular dysfunction, atrial fibrillation, atrial flutter, sinus rhythm, prognosis, outcome, death and hospitalization. Bibliographies of retrieved publications were hand searched. Studies were eligible if they included a HF population and if outcomes were reported by cardiac rhythm (AF or SR). Studies were reviewed by predetermined protocol (including quality assessment). Data were pooled using a random effects model.

Results: Twenty studies were included (from 3380 initially identified) representing 32946 patients (10819 deaths). Nine randomized controlled trials (RCT) were included. The prevalence of AF was 15%, crude mortality rates were 46% (AF) and 33% (SR). The odds ratio for death was 1.33 (95% confidence interval (CI) 1.12-1.59) for AF compared with SR. Eleven observational studies were included. The prevalence of AF was 23%, crude mortality rates were 38% (AF) and 25% (SR). The odds ratio for death was 1.57 (95% CI 1.20-2.05) for AF compared with SR.

Conclusion: This meta-analysis demonstrates that AF is associated with worse outcomes for patients with HF compared with those with SR. Further research is required to determine whether the adverse outcome associated with AF is related to the arrhythmia itself, or to variables, such as HF severity, patient age and comorbidity.

Publication types

  • Comparative Study
  • Meta-Analysis

MeSH terms

  • Atrial Fibrillation / complications
  • Atrial Fibrillation / mortality*
  • Heart Failure / complications
  • Heart Failure / mortality*
  • Humans
  • Mortality / trends
  • Randomized Controlled Trials as Topic / methods
  • Risk Factors