Digoxin and Mortality in Atrial Fibrillation

Abstract & Commentary

By John P. DiMarco, MD, PhD, Professor of Medicine, Division of Cardiology, University of Virginia, Charlottesville.

Source: Gheorghiade M, et al. Lack of evidence of increased mortality among patients with atrial fibrillation taking digoxin: Findings from post hoc propensity-matched analysis of the AFFIRM trial. Eur Heart J 2013; Apr 16. [epub ahead of print.]

The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial was completed more than 10 years ago. That study enrolled more than 4000 patients who were randomized to either rate-control or rhythm-control strategies for management of atrial fibrillation (AF). In this paper, Gheorghiade and colleagues perform a post-hoc propensity-matched analysis on the effects of digoxin on mortality in the AFFIRM trial.

The authors used a public use dataset from AFFIRM obtained from the National Heart, Lung, and Blood Institute. They identified 1377 patients who received digoxin as initial therapy at baseline in the trial. Digoxin was used alone in 16% of the entire group and in combination with either a beta-blocker (14%) or with a calcium channel blocker (14%). The primary endpoint for this analysis was all-cause mortality. Secondary endpoints included all-cause hospitalization and non-fatal arrhythmias. A propensity score was estimated based on 59 relevant variables. Using a matching protocol, pairs of patients receiving and not receiving digoxin as initial therapy at baseline were collected. Comparative outcomes in these two groups are then reported. The matched patients receiving and not receiving digoxin as initial therapy had a mean age of 70 years, 40% were women, and 40% had prior hospitalizations due to arrhythmias. All-cause mortality occurred in 14 and 13% of matched patients on and off digoxin as initial therapy, respectively. The results were similar among various subgroups including those with and without heart failure. Digoxin had no association with mortality at any time point when used either as monotherapy or in combination with other rate-control drugs. Propensity-matched and propensity-adjusted hazard ratios for all-cause mortality associated with digoxin therapy at baseline and during the 6 months prior to baseline were all close to one. There was also no significant difference in the incidence of all-cause hospitalization in the matched patients receiving (56%) and not receiving (59%) digoxin as baseline initial therapy. Finally, incident non-fatal arrhythmias were rare with only 1% of matched patients in each group having such arrhythmias.

The authors conclude that digoxin has no association with mortality in patients with heart failure in the AFFIRM study when bias is removed with a careful propensity-matched analysis.


Whether digoxin should continue to be used for rate control in patients with AF has become controversial. Most guidelines, however, still recommend digoxin as a second-line agent in patients with heart failure and as an alternate first-line therapy in sedentary, elderly patients. This paper is in response to another study using the AFFIRM database that reported that digoxin increased mortality.1 An earlier study by AFFIRM investigators had also shown that digoxin was associated with increased mortality. Both of these latter studies used a time-dependent survival analysis and it is likely that digoxin was added as patients became sicker and their heart rates became more difficult to control during the course of the trial. The paper, which used propensity matching based on baseline variables only, supports the continued place of well-monitored digoxin as a second-line agent for rate control in patients with heart failure and in select patients who respond to it as monotherapy.


1. Whitbeck M, et al. Increased mortality among patients taking digoxin — analysis from the AFFIRM study. Eur Heart J 2013;34:1481-1488.