Safety and effectiveness of phenobarbital vs. benzodiazepines for severe alcohol withdrawal in alcohol-associated liver disease in the ICU.

BACKGROUND &

aimsSevere alcohol withdrawal syndrome (SAWS) in patients with alcohol-associated liver disease (ALD) presents management challenges due to altered drug metabolism and increased toxicity. While benzodiazepines remain standard therapy, concerns about drug shortages, over-sedation, hepatic encephalopathy, and respiratory depression have prompted interest in phenobarbital as an alternative.

methodsWe conducted a retrospective cohort study using the MIMIC-IV database (2008-2019) comparing phenobarbital and benzodiazepines in patients in the intensive care unit (ICU) with ALD and SAWS. Patients receiving either drug as the initial sedative were identified, and outcomes were analyzed using inverse probability of treatment weighting to balance baseline characteristics. Outcomes included in-hospital mortality, need for mechanical ventilation, time on ventilator, ICU length of stay (LOS), and hospital LOS. Subgroup analyses compared phenobarbital with short-acting (lorazepam) and long-acting (diazepam, chlordiazepoxide) benzodiazepines.

resultsAmong 763 phenobarbital-treated and 2,863 benzodiazepine-treated patients, in-hospital mortality was similar (4.5% vs. 5%; odds ratio 0.88, 95% CI 0.64-1.21). Phenobarbital was associated with reduced mechanical ventilation (32.7% vs. 39.3%; odds ratio 0.82, 95% CI 0.74-0.91), and shorter ICU LOS (3.6 vs. 4.1 days; p = 0.02) with comparable hospital LOS. When stratified by benzodiazepine type, benefits were driven by improved outcomes vs. short-acting benzodiazepines (lower mortality, ventilation, and LOS). In patients with decompensated cirrhosis or MELD-Na >14, phenobarbital demonstrated similar mortality (19% vs. 18.9%) and comparable ICU and ventilatory outcomes.

conclusionsIn this first large cohort of patients with severe liver disease, phenobarbital showed comparable mortality, respiratory, and ICU outcomes to benzodiazepines in ALD patients with severe alcohol withdrawal. These results support phenobarbital as a viable alternative in this high-risk population, warranting prospective trials to validate the findings. IMPACT AND IMPLICATIONS: This is the first large study to evaluate phenobarbital vs. benzodiazepines for severe alcohol withdrawal specifically in patients with alcohol-associated liver disease, a population often excluded from prior research. This study provides important evidence showing that phenobarbital yields comparable respiratory and mortality outcomes to benzodiazepines in critically ill patients with alcohol-associated liver disease. These findings are important for hepatologists, intensivists, and hospital clinicians who face therapeutic uncertainty and safety concerns regarding benzodiazepine use in this group. Establishing phenobarbital as a viable alternative could help inform future clinical guidelines and support safer, evidence-based management of withdrawal in patients with advanced liver disease.