790: ALCOHOL WITHDRAWAL MANAGEMENT IN THE INTENSIVE CARE UNIT: A RETROSPECTIVE PRE-PROTOCOL ANALYSIS

Abstract

Introduction: Alcohol withdrawal syndrome (AWS) has been associated with increased hospital and intensive care unit (ICU) length of stay (LOS), longer duration of mechanical ventilation (MV), higher incidence of delirium and hospital acquired infections, higher health care costs and increased mortality. It has been well defined in the literature that standardized protocols for treatment of AWS are associated with improved outcomes; however, to date there is no universally accepted standard of care. The goal of this study was to characterize the management and associated outcomes of patients treated for AWS in the ICU prior to implementation of a standardized AWS treatment protocol. Methods: A single center retrospective analysis of the management of AWS in the ICU was performed. Patients 18 years of age and older admitted to the ICU between 2012 and 2018 who developed AWS based on International Classification of Diseases, Ninth and Tenth Revision codes and administration of at least one dose of a benzodiazepine were included. Outcomes assessed included benzodiazepine utilization, need for and duration of mechanical ventilation, incidence of delirium and hospital acquired infections, ICU and hospital LOS and mortality. Results: A total of 163 medical and surgical patients were included in the analysis. Mean total lorazepam equivalents required during ICU admission was 64 mg (SD 135.9 mg) which amounted to an average daily lorazepam equivalent of 9.45 mg (SD 12 mg) per patient. Sixty-nine percent of patients required MV with the most common indication being refractory agitation (25.7%) or over-sedation (29.2%). Mean ICU and hospital LOS were 6.4 days and 12.9 days respectively. Complications assessed included development of delirium (76.1 %), hospital acquired pneumonia (38.7 %), and seizure during admission (14.7 %). Conclusions: This data describes the management of alcohol withdrawal patients at our institution prior to initiation of a standardized AWS treatment protocol. Our pre-protocol benzodiazepine utilization, need for MV, incidence of delirium, ICU and hospital LOS are significantly higher compared to previously published data. Future endeavors include comparison of pre- versus post-protocol data and creation of a predictive analytics model to predict ICU LOS based on patient specific variables.