Dexmedetomidine for reduction of atrial fibrillation and delirium after cardiac surgery (DECADE): a randomised placebo-controlled trial

Abstract

BACKGROUND Atrial fibrillation and delirium are common consequences of cardiac surgery. Dexmedetomidine has unique properties as sedative agent and might reduce the risk of each complication. This study coprimarily aimed to establish whether dexmedetomidine reduces the incidence of new-onset atrial fibrillation and the incidence of delirium. METHODS A randomised, placebo-controlled trial was done at six academic hospitals in the USA. Patients who had had cardiac surgery with cardiopulmonary bypass were enrolled. Patients were randomly assigned 1:1, stratified by site, to dexmedetomidine or normal saline placebo. Randomisation was computer generated with random permuted block size 2 and 4, and allocation was concealed by a web-based system. Patients, caregivers, and evaluators were all masked to treatment. The study drug was prepared by the pharmacy or an otherwise uninvolved research associate so that investigators and clinicians were fully masked to allocation. Participants were given either dexmedetomidine infusion or saline placebo started before the surgical incision at a rate of 0·1 μg/kg per h then increased to 0·2 μg/kg per h at the end of bypass, and postoperatively increased to 0·4 μg/kg per h, which was maintained until 24 h. The coprimary outcomes were atrial fibrillation and delirium occurring between intensive care unit admission and the earlier of postoperative day 5 or hospital discharge. All analyses were intention-to-treat. The trial is registered with ClinicalTrials.gov, NCT02004613 and is closed. FINDINGS 798 patients of 3357 screened were enrolled from April 17, 2013, to Dec 6, 2018. The trial was stopped per protocol after the last designated interim analysis. Among 798 patients randomly assigned, 794 were analysed, with 400 assigned to dexmedetomidine and 398 assigned to placebo. The incidence of atrial fibrillation was 121 (30%) in 397 patients given dexmedetomidine and 134 (34%) in 395 patients given placebo, a difference that was not significant: relative risk 0·90 (97·8% CI 0·72, 1·15; p=0·34). The incidence of delirium was non-significantly increased from 12% in patients given placebo to 17% in those given dexmedetomidine: 1·48 (97·8% CI 0·99-2·23). Safety outcomes were clinically important bradycardia (requiring treatment) and hypotension, myocardial infarction, stroke, surgical site infection, pulmonary embolism, deep venous thrombosis, and death. 21 (5%) of 394 patients given dexmedetomidine and 8 (2%) of 396 patients given placebo, had a serious adverse event as determined by clinicians. 1 (<1%) of 391 patients given dexmedetomidine and 1 (<1%) of 387 patients given placebo died. INTERPRETATION Dexmedetomidine infusion, initiated at anaesthetic induction and continued for 24 h, did not decrease postoperative atrial arrhythmias or delirium in patients recovering from cardiac surgery. Dexmedetomidine should not be infused to reduce atrial fibrillation or delirium in patients having cardiac surgery. FUNDING Hospira Pharmaceuticals.