Bayesian analysis of amiodarone or lidocaine versus placebo for out-of-hospital cardiac arrest

Objective Clinical trials for patients with shock-refractory out-of-hospital cardiac arrest (OHCA), including the Amiodarone, Lidocaine or Placebo (ALPS) trial, have been unable to demonstrate definitive benefit after treatment with antiarrhythmic drugs. A Bayesian approach, combining the available evidence, may yield additional insights. Methods We conducted a reanalysis of the ALPS trial comparing treatment with amiodarone or lidocaine with placebo in patients with OHCA following shock-refractory ventricular fibrillation or ventricular tachycardia (VF/VT). We used Bayesian regression to assess the probability of improved survival or improved neurological outcome on the 7-point modified Rankin Scale. We derived weak, moderate and strong priors from a previous clinical trial. Results The original ALPS trial randomised 3026 adult patients with OHCA to amiodarone (n=974, survival to hospital discharge 24.4%), lidocaine, (n=993, survival 23.7%) or placebo (n=1059, survival 21.0%). In our reanalysis the probability of improved survival from amiodarone ranged from 83% (strong prior) to 95% (weak prior) compared with placebo and from 78% (strong) to 90% (weak) for lidocaine - an estimated improvement in survival of 2.9% (IQR 1.4%-3.8%) for amiodarone and 1.7% (IQR 0.84%-3.2%) for lidocaine over placebo (moderate prior). The probability of improved neurological outcome from amiodarone ranged from 96% (weak) to 99% (strong) compared with placebo and from 88% (weak) to 96% (strong) for lidocaine. Conclusions In a Bayesian reanalysis of patients with shock-resistant VF/VT OHCA, treatment with amiodarone had high probabilities of improved survival and neurological outcome, while treatment with lidocaine had a more modest benefit.