859

Abstract

Introduction: The incidence of alcohol withdrawal syndrome (AWS) is approximately 2 million patients per year in the United States. The primary pharmacologic therapy for AWS is benzodiazepines, yet escalating doses are associated with respiratory depression. Therefore, dexmedetomidine is commonly utilized as an adjunct therapy due to lack of effect on respiratory drive; however, literature comparing efficacy and safety profiles of these agents is currently lacking. Methods: This retrospective study reviewed all adult patients who received continuous infusions of benzodiazepines (lorazepam or midazolam) or dexmedetomidine for severe AWS. Exclusion criteria included intubation prior to continuous infusion, history of non-alcohol withdrawal seizures, and admission for AWS in the previous 30 days. Results: A total of 61 patients met all study criteria (benzodiazepine n=33, dexmedetomidine n=28). There were no differences in the primary composite endpoint between the benzodiazepine and dexmedetomidine groups (9.1% and 7.1% respectively, p=1.0), nor its individual components of respiratory distress (9.1% vs 7.1% respectively, p=1.0) and alcohol withdrawal seizures (0% vs 3.6% respectively, p=0.46). During dexmedetomidine infusions, episodes of bradycardia (46.4% vs 15.2%, p=0.008) and hypotension (42.9% vs 12.1%, p=0.009) occurred more often. After initiation of dexmedetomidine, one third of patients did not continue to receive benzodiazepines, potentially increasing the risk of alcohol withdrawal seizures. The average cost of hospitalization in the dexmedetomidine group was $17,014.62 ± 2,180.62 compared to $11,467.60 ± 1,568.48 in the benzodiazepine group (p=0.11). Conclusions: No difference was identified in AWS endpoints, but more cardiovascular complications were noted in the group that received dexmedetomidine. Off-label use of dexmedetomidine for AWS should be done with caution due to patient safety concerns until randomized, controlled trials can accurately identify its appropriate role in treatment.