There were 6 studies (n=745) of quinidine, 5 (n=2,415) of warfarin, and 5 (n=255) of amiodarone.
Specific drugs.
Quinidine: the RR for atrial fibrillation reversal was 0.67 (SE=0.16). Intolerable side-effects led to discontinuation of quinidine therapy at the rate of 14% (SE=3%) per patient-year; risk of sudden death was 3.7% (SE=2.4%) per patient-year. No patient in the placebo group suffered intolerable side-effects or sudden death.
Warfarin: the RR for thromboembolism was 0.27 (SE=0.18). The rate of fatal intracranial haemorrhage during warfarin therapy was 0.35% (SE=0.91%) per patient-year. The rate of withdrawal because of intolerable side-effects was 10% (SE=3%) for warfarin, 6% for placebo (p<0.001). Amiodarone: the RR for atrial fibrillation was 0.20 (SE=0.10), assuming all patients would remain in atrial fibrillation without treatment. One patient suffered a fatal adverse reaction.
Comparisons between drugs.
If the baseline risk is less than 11% per patient-year, the total risk during quinidine therapy exceeds the risk without therapy. In contrast, warfarin treatment is associated with a reduced total risk over the whole baseline range of thromboembolism risk. The total risk of therapy with amiodarone did not differ from that with warfarin, and was lower than total risk of quinidine therapy.