450

Abstract

Introduction: Benzodiazepines (BZD) are standard of care for management of alcohol withdrawal syndrome (AWS). Valproic acid (VPA) avoids many of the undesirable effects of BZD; several studies have demonstrated efficacy of VPA for AWS however they contained methodological flaws and small patient populations. This study assessed if VPA combined with BZD improved outcomes related to management of AWS compared to BZD alone. Methods: This was a retrospective cohort analysis. All adult patients from a mixed intensive care unit population who had a Clinical Institute Withdrawal Assessment for Alcohol Revised (CIWA-Ar) assessment performed and received either BZD or VPA were retrospectively identified. Patients were stratified into two groups: BZD alone and BZD combined with VPA. The primary endpoint was amount of BZD in lorazepam equivalents. Secondary endpoints included: total mortality, hospital and ICU lengths of stay (LOS), treatment duration, mean CIWA-Ar score, and percentage of CIWA-Ar <8. Results: A total of 127 patients met inclusion criteria; 90 (70.9%) received BZD alone while 37 (29.1%) received combination therapy. Median total BZD use was 12.8 mg compared to 17 mg with combined therapy (p=0.81). No difference was found in mortality (p=1). Both hospital and ICU LOS were significantly lower with standard therapy (p<0.01,0.00). Treatment duration was similar between groups (p=0.95). The mean percentage of CIWA-Ar<8 was 65.7% and 55% (p=0.04) for BZD alone and combined therapy patients; median average CIWA-Ar scores were 6 and 8.5 (p=0.01) respectively. Conclusions: No benefit was seen for addition of VPA to BZD however it was not found to worsen outcomes. VPA may lower CIWA-Ar scores however optimal dosing strategies will require further study.