913

Abstract

Introduction: Benzodiazepines (BZDs) are commonly used as first-line agents for the management of alcohol withdrawal syndrome (AWS). A subset of withdrawal patients do not respond to treatment despite escalating doses of BZDs. Resistant alcohol withdrawal (RAW) has been defined as the requirement of greater than a benzodiazepine-equivalent of 40 mg of diazepam in one hour for management of AWS. RAW has been associated with a higher incidence of mechanical ventilation and nosocomial pneumonia and a longer intensive care unit stay. Strategies that have been studied to manage RAW include high dose BZDs, dexmedetomidine, ketamine, phenobarbital and propofol. The objective of this study is to characterize the pharmacological management of RAW to help formalize a protocol for the management of these patients. Methods: A retrospective study was completed at a single center in a large, academic medical center. The cohort of patients was identified utilizing ICD-9 codes for severe alcohol withdrawal. Patients were included if they met the criteria for RAW. Data on pharmacologic management were collected for seven days after meeting RAW criteria, including the initial, median and maximum maintenance doses for newly added medications, incidence of specialty consultation and the time to documented resolution of RAW. Results: A total of 447 patients with an ICD-9 code for severe alcohol withdrawal were screened for RAW criteria. One-hundred patients (22.4%) meeting RAW criteria were evaluated. A total of seventeen individual medications and multiple combinations were used for patients meeting RAW criteria, varying from BZDs to antipsychotics for management of agitation. Not surprisingly, BZDs were utilized to treat 100% of patients for AWS and lorazepam was the most commonly was BZD (n = 76). Propofol was the most commonly used adjunct agent (n = 66) after determination of RAW, with a median initial dose of 30 mcg/kg/min and median maintenance dose of 29.55 mcg/kg/min. Dexmedetomidine was utilized in fifteen patients, with a median initial dose of 0.4 mcg/kg/hour and a median maintenance dose of 0.54 mcg/kg/hour. The median maximum maximum dose of dexmedetomidine was 0.7 mcg/kg/hour, which was consistent with the dosing recommendations at our institution. The addition of phenobarbital (n = 9) was considered only with the addition of a toxicology consult. Specialty consultations were obtained in 37 patients and all patients were managed by the critical care team. The median time to resolution of RAW was 7 days. Conclusions: The management of patients meeting RAW criteria indicates that prescribing patterns are quite variable. The introduction of a formalized protocol for the management of RAW patients may be beneficial for streamlined management. Future directions include the comparison of different medications on clinical outcomes.