B55-31 Clinical Outcomes of Benzodiazepine vs Phenobarbital EMR-Based Order Sets for Alcohol Withdrawal in a Community Hospital

Rationale

Phenobarbital has gained increasing attention as an alternative to benzodiazepines for inpatient alcohol withdrawal syndrome (AWS), yet most comparative studies focus on severe AWS or ICU-level care and rarely evaluate implementation within an electronic medical record (EMR)-based treatment pathway. Evidence is limited regarding how phenobarbital performs across an entire hospitalized AWS population in a community setting.

Methods

We conducted a retrospective cohort study of adult inpatients treated for AWS at a university-affiliated community hospital using a standardized EMR-based AWS order set. Patients were assigned to benzodiazepine- or phenobarbital-based management based on the initial order set selected. Baseline demographics and comorbidity burden were similar between treatment groups. The primary outcome was early clinical decompensation, defined as ICU transfer or endotracheal intubation within 24 hours of treatment initiation. Secondary outcomes included delirium, seizures, and hospital length of stay.

Results

Among 399 included patients, 296 received benzodiazepines, and 103 received phenobarbital. Early ICU transfer or intubation occurred in 10 patients (3.4%) in the benzodiazepine group and in none in the phenobarbital group. Treatment-related delirium occurred in 6.3% of benzodiazepine-treated patients and in no phenobarbital-treated patients. Seizure rates were similarly low in both groups. Mean hospital length of stay was longer in the phenobarbital cohort (5.41 vs 4.13 days), contrasting with prior studies suggesting shortened stays with phenobarbital.

Conclusions

In this real-world, hospital-wide evaluation of an EMR-based AWS order set, phenobarbital-based treatment was associated with no early ICU transfers or intubations and no delirium events, with seizure rates comparable to benzodiazepines. The longer length of stay observed in the phenobarbital cohort could reflect social factors common in safety-net settings rather than inferior clinical effectiveness. These findings support the use of phenobarbital as a safe and effective early management strategy for AWS in community hospitals.

This abstract is funded by: None