Establishing the safety of phenobarbital treatment of alcohol withdrawal syndrome on general medical wards: A retrospective cohort study

• Published in: Alcohol (Fayetteville, N.Y.) N.Y.), 2024-05, Vol.116, p.29-34
• Main Authors: Ronan, Matthew V., Ganatra, Rahul B., Saukkonen, Jussi
• Format: Article
• Language: English
• Subjects: Alcohol withdrawal; alcohol withdrawal syndrome; Benzodiazepines; Cohort analysis; Delirium; Drug dosages; Emergency medical care; general medical wards; Hospitalization; Intensive care; Intubation; Length of stay; Medical records; Patients; Phenobarbital; Pneumonia; Psychiatry; Seizures
• ISSN: 0741-8329; 1873-6823; 1873-6823
• DOI: 10.1016/j.alcohol.2023.10.005

Concern about adverse effects from phenobarbital limits its use in treating alcohol withdrawal syndrome (AWS) on general medical wards. Benzodiazepines are the recommended treatment for inpatient management of AWS, yet a subset of patients have an inadequate response or experience complications of AWS despite treatment with benzodiazepines. Data supporting an alternative treatment are needed. We set out to estimate the rate of serious adverse events (SAEs) of phenobarbital treatment for AWS on general medical wards. Retrospective cohort study of all general medical ward patients hospitalized at a single tertiary urban VA Medical Center from October 2018–May 2021 who received phenobarbital for treatment of AWS. Primary outcomes were SAEs attributed to phenobarbital and treatment failure. SAEs were defined as ICU transfer or intubation for over-sedation, pneumonia, and death. Treatment failure was defined as progression of withdrawal resulting in seizure, ICU transfer, behavioral emergencies, or death. During the study period, phenobarbital was administered in 29% (244) of all AWS hospitalizations. Among them, 93% had a history of AWS hospitalization and 68% had a history of complicated AWS. Fifty-three percent of patients met criteria for moderate, severe, or complicated withdrawal prior to phenobarbital initiation. The mean cumulative dose of phenobarbital per patient was 966.5 mg (13.6 mg/kg). SAEs occurred in 1 of 244 hospitalizations (0.4%): there were no intubations, ICU transfers for oversedation, or deaths due to phenobarbital or AWS. One case of pneumonia was possibly attributable to phenobarbital. Treatment failures (6 ICU transfers, 9 behavioral emergencies) were identified during 12 of 244 hospitalizations (4.9%). SAEs and treatment failures were infrequent among 148 patients treated with phenobarbital across 244 hospitalizations with a mean cumulative dose of 966.5 mg per patient. Our findings suggest that phenobarbital is a safe alternative treatment of AWS in general medical ward patients. •There were no seizures, deaths, or intubations in patients treated with phenobarbital for alcohol withdrawal.•Nearly 70% of the cohort had a history of severe, complicated alcohol withdrawal.•Transfer to the ICU occurred in only 2.5% of hospitalizations treated with phenobarbital.•Adjunctive medication use for symptom management was high (53% of cohort).