Phenobarbital compared to benzodiazepines in alcohol withdrawal treatment: A register-based cohort study of subsequent benzodiazepine use, alcohol recidivism and mortality

Background: Long-acting benzodiazepines such as chlordiazepoxide are recommended as first-line treatment for alcohol withdrawal. These drugs are known for their abuse liability and might increase alcohol consumption among problem drinkers. Phenobarbital could be an alternative treatment option, possibly with the drawback of a more pronounced acute toxicity. We evaluated if phenobarbital compared to chlordiazepoxide decreased the risk of subsequent use of benzodiazepines, alcohol recidivism and mortality. Methods: The study was a register-based cohort study of patients admitted for alcohol withdrawal 1998-2013 and treated with either phenobarbital or chlordiazepoxide. Patients were followed for one year. We calculated hazard ratios (HR) for benzodiazepine use, alcohol recidivism and mortality associated with alcohol withdrawal treatment, while adjusting for confounders. Results: A total of 1063 patients treated with chlordiazepoxide and 1365 patients treated with phenobarbital were included. After one year, the outcome rates per 100 person-years in the phenobarbital versus the chlordiazepoxide cohort were 9.20 vs. 5.13 for use of benzodiazepine, 37.9 vs. 37.9 for alcohol recidivism and 29 vs. 59 for mortality. Comparing phenobarbital to chlordiazepoxide treated, the HR of subsequent use of benzodiazepines was 1.56 (95%CI 1.05-2.30). Similarly, the HR for alcohol recidivism was 0.99 (95%CI 0.84-1.16). Lastly, the HR for 30-days and 1 year mortality was 0.25 (95%CI 0.08-0.78) and 0.51 (95%CI 0.31-0.86). Conclusion: There was no decreased risk of subsequent benzodiazepine use or alcohol recidivism in patients treated with phenobarbital compared to chlordiazepoxide. Phenobarbital treatment was associated with decreased mortality, which might be confounded by somatic comorbidity among patients receiving chlordiazepoxide.