Another Strike for Digoxin in AF: Mortality Goes Up in Cohort Study

March 10, 2013

SAN FRANCISCO, California — Patients with atrial fibrillation (AF) starting digoxin for the first time showed an independently significant jump in mortality over an average of one year compared with nondigoxin users in an observational study reported here today at the American College of Cardiology 2013 Scientific Sessions. Stratified analysis showed the effect to be consistent in men and women and by age.

Digoxin was given for rate control; patients with heart failure had been excluded from the study.

Although mortality doubled with digoxin use, hospitalizations appeared unaffected, suggesting that the people died at home, according to lead author Dr James V Freeman (Stanford University School of Medicine, CA). And though there were no data on cause of death, he told heartwire that ventricular arrhythmias are a hazard from digoxin toxicity, "so our hypothesis is that maybe [digoxin users] had an increase in arrhythmic death and sudden cardiac death. But there's no way to know that for certain."

Digoxin in one form or another is one of the oldest still-used drugs in medicine and was never tested in substantial clinical trials for efficacy or safety in AF. Still, Freeman observed, the drug "is much more commonly used than people think, and it's not just that people who were using it 20 years ago are still using it."

The current findings support a recent propensity-adjusted secondary analysis of the AFFIRM trial, which found a 41% jump in all-cause mortality in AF patients taking vs not taking digoxin. As previously reported by heartwire , the increased hazard was seen in both men and women and in patients with and without underlying heart failure.

The group looked at 23 272 digoxin-naive patients newly diagnosed with AF from 2006 to 2009 within Kaiser Permanente Northern and Southern California. Of that group, 12.9% were started on digoxin for rate control.

The adjusted hazard ratio (HR) for all-cause mortality over a median of 0.8 years was 2.06 (95% CI 1.73–2.45) and for hospitalization was 1.05 (95% CI 0.98–1.13). The mortality finding remained significant in subgroup analysis by sex and three age groups.

Adjusted HR (95% CI), digoxin vs no digoxin use, for death from any cause in adults with incident AF (2006–2009)

Subgroup HR (95% CI)*
Male 2.03 (1.58–2.60)
Female 2.13 (1.67–2.71)
Age (y)  
21–74 2.31 (1.68–3.17)
75–84 1.66 (1.23–2.23)
>85 2.50 (1.85–3.38)

*Adjusted for age, race, income, laboratory parameters, prior CV disease and procedures, hypertension, dyslipidemia, cancer, lung disease, and cardiovascular medications

"This is about as strong an observational cohort study as you can get. I say that for a number of reasons," Freeman noted. It's large and from a "closed" healthcare system that is responsible for all patients' medications and lab results, and it would be known whether a patient might be noncompliant with meds, for example.

And there is potential from residual confounding due to differences in comorbidities and medications, which could--though they were adjusted for--suggest the study's results are conservative rather than overstated, he observed. Interestingly, nondigoxin users were sicker and showed a significantly greater baseline history of MI, stroke, and coronary revascularization, as well as more dyslipidemia and hypertension. Understandably, they were also on more CV medications.

The strong observational data for a hazard and thin evidence base supporting it, according to Freeman, is enough to recommend a reassessment of digoxin for AF rate control and maybe to give more consideration to certain patients--such as those who are highly symptomatic--for catheter ablation.

Neither Freeman nor his coauthors had disclosures.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.