The Bad Reputation of Digoxin in Atrial Fibrillation—Causality or Bias? Nationwide Nested Case-Control Study

Abstract

Aims

Studies have reported excess risk of mortality associated with digoxin in atrial fibrillation (AF).

This study sought to investigate if these findings could be replicated and whether a potential association could be explained by bias.

Methods

Using Danish Nationwide registers, a nested-case control study from 2012 to 2022 was conducted in a cohort of patients with AF. Cases were defined as death of any cause and the exposure was treatment with digoxin compared with beta blockers/verapamil. To investigate bias, additional analyses with negative control outcomes as case definitions—in which we would not expect a plausible association (eg, nursing home admission)—were employed. Associations were reported as hazard ratios (HRs) with 95% confidence intervals (95% CI).

Results

A total of 59,748 cases were identified and matched 1:10 with controls (53% men, median age: 84 [IQR: 77-89]). Digoxin was associated with increased rates of mortality in the entire cohort (HR 1.85, 95% CI 1.78-1.92) as well as subgroups such as patients with heart failure (HR 1.84, 95% CI 1.65-2.06), diabetes (HR 1.85, 95% CI 1.6-2.14), and kidney disease (HR 1.37, 95% CI 1.04-1.8). Significant associations with all negative control outcomes were also found, most notably nursing home admissions (HR 1.79, 95% CI 1.67-1.93).

Conclusion

Digoxin use was associated with increased mortality in AF. However, negative control outcomes were also associated with digoxin use indicating that the described association between digoxin and mortality is likely not causal and being prescribed digoxin is merely a marker of more advanced disease and frailty.