Characterization of Phenobarbital Use for Alcohol Withdrawal in the Emergency Department

This study aims to characterize phenobarbital (PHB) utilization for alcohol withdrawal syndrome (AWS) within NewYork-Presbyterian Hospital (NYPH), assess adherence to institutional guidelines, and evaluate associated clinical outcomes. The current practice at NYPH is to reserve PHB as an adjunctive agent for benzodiazepine-refractory withdrawal, defined as ≥ 40 mg of IV diazepam in 1 hr, ≥ 200 mg in 3 hrs, or ≥ 400 mg in 8 hrs. The findings will inform practice optimization, support evidence-based guideline refinement, and potentially improve patient outcomes in AWS management.

This retrospective chart review included adults ≥18 years presenting to the Emergency Department (ED) at Columbia University Irving Medical Center from July 1, 2023, to June 30, 2025, who received PHB for suspected AWS. Medication data included PHB and benzodiazepine doses, timing, and adherence to the NYPH AWS guideline. CIWA-Ar assessments, monitoring, and symptom resolution (CIWA-Ar ≤8) were recorded. Safety outcomes included changes in hemodynamics and respiratory status. The primary outcomes were escalation of care, defined as vasopressor use, increased respiratory support from baseline, and time to CIWA-Ar ≤8. Secondary outcomes included ED and ICU length of stay. Descriptive statistics were used to analyze outcomes.

A total of 50 patients were included. PHB was administered at an average cumulative dose of 314 mg over a mean of 3 doses with a total cumulative ideal body weight (IBW)-based dose of 5.5 mg/kg. Only 16% of patients adhered to the NYPH AWS guideline, with the most common deviation being early initiation. The average initial CIWA-Ar score was 13, and the average score before PHB was 20. The average time to symptom resolution after PHB administration (CIWA-Ar ≤8) was 400.2 minutes. One patient required vasopressor support for hypotension. Escalation in respiratory support occurred in 30% of patients, with the most common being initiation of nasal canula. 78% of patients were admitted to the ICU. The mean lengths of stay were 36.1 hours in the ED, and 77.4 hours in the ICU.

These findings suggest that PHB is commonly used earlier than intended in the AWS treatment pathway at NYPH and may be safely administered in a high-risk population likely to experience worsening withdrawal symptoms during admission, highlighting opportunities for protocol refinement and standardization of care.