Digoxin for atrial fibrillation and atrial flutter: A systematic review with meta-analysis and trial sequential analysis of randomised clinical trials

Background During recent years, systematic reviews of observational studies have compared digoxin to no digoxin in patients with atrial fibrillation or atrial flutter, and the results of these reviews suggested that digoxin seems to increase the risk of all-cause mortality regardless of concomitant heart failure. Our objective was to assess the benefits and harms of digoxin for atrial fibrillation and atrial flutter based on randomized clinical trials. Methods We searched CENTRAL, MEDLINE, Embase, LILACS, SCI-Expanded, BIOSIS for eligible trials comparing digoxin versus placebo, no intervention, or other medical interventions in patients with atrial fibrillation or atrial flutter in October 2016. Our primary outcomes were all cause mortality, serious adverse events, and quality of life. Our secondary outcomes were heart failure, stroke, heart rate control, and conversion to sinus rhythm. We performed both random-effects and fixed-effect meta-analyses and chose the more conservative result as our primary result. We used Trial Sequential Analysis (TSA) to control for random errors. We used GRADE to assess the quality of the body of evidence. Results 28 trials (n = 2223 participants) were included. All were at high risk of bias and reported only short-term follow-up. When digoxin was compared with all control interventions in one analysis, we found no evidence of a difference on all-cause mortality (risk ratio (RR), 0.82; TSA-adjusted confidence interval (Cl), 0.02 to 31.2; l(2) = 0%); serious adverse events (RR, 1.65; TSA-adjusted Cl, 0.24 to 11.5; l(2) = 0%); quality of life; heart failure (RR, 1.05; TSA-adjusted Cl, 0.00 to 1141.8; l(2) = 51%); and stroke (RR, 2.27; TSA-adjusted Cl, 0.00 to 7887.3; l(2) 17%). Our analyses on acute heart rate control (within 6 hours of treatment onset) showed firm evidence of digoxin being superior compared with placebo (mean difference (MD), -12.0 beats per minute (bpm); TSA-adjusted Cl, -17.2 to -6.76; l(2) = 0%) and inferior compared with beta blockers (MD, 20.7 bpm; TSA-adjusted Cl, 14.2 to 27.2; l(2) = 0%). Meta-analyses on acute heart rate control showed that digoxin was inferior compared with both calcium antagonists (MD, 21.0 bpm; TSA-adjusted Cl, -30.3 to 72.3) and with amiodarone (MD, 14.7 bpm; TSA-adjusted CI, -0.58 to 30.0; l(2)= 42%), but in both comparisons TSAs showed that we lacked information. Meta-analysis on acute conversion to sinus rhythm showed that digoxin compared with amiodarone reduced the probability of converting atrial fibrillation to sinus rhythm, but TSA showed that we lacked information (RR, 0.54; TSA-adjusted Cl, 0.13 to 2.21; l(2) = 0%). Conclusions The clinical effects of digoxin on all-cause mortality, serious adverse events, quality of life, heart failure, and stroke are unclear based on current evidence. Digoxin seems to be superior compared with placebo in reducing the heart rate, but inferior compared with beta blockers. The long-term effect of digoxin is unclear, as no trials reported long-term follow-up. More trials at low risk of bias and low risk of random errors assessing the clinical effects of digoxin are needed.