Phenobarbital and/or benzodiazepines for recurrent alcohol withdrawal: A self-controlled, retrospective cohort study

Abstract

Background

While there is ample data supporting the use of barbiturates and benzodiazepines (BZDs) for the treatment of alcohol withdrawal, there is a paucity of information on treating recurrent withdrawal among high healthcare utilizing patients. The purpose of this study was to assess the efficacy and safety of phenobarbital (PB), with or without adjuvant BZDs, for treatment of acute alcohol withdrawal in the emergency department (ED) in patients with high rates of recurrent withdrawal.

Methods

This non-matched, self-controlled, retrospective cohort study evaluated patients seen in the ED of an urban trauma center and safety-net teaching hospital between July 1st, 2018, and July 31st, 2019. Patients treated for alcohol withdrawal were included if they had at least one visit where they received intravenous PB with or without BZDs, then during a separate encounter received BZD only. Each encounter was then assigned to a treatment group based on administration of PB only, BZD only, or the combination of PB and BZD. The primary outcomes were admission to hospital or discharge and return to the ED for any reason within 48 h of disposition.

Results

A total of 137 unique patients were included, with 642 encounters composed of 245 PB only, 293 BZD only, and 104 combination visitations. No significant difference was found between the PB, BZD, or combination treatment groups for rates of admission (36.7%, 38.9%, and 46.1% respectively) or for return within 48 h (17.1%, 15.0%, and 13.5%). There was a significantly longer ED length of stay for the combination group (8.6 h) compared to either the PB or BZD only groups (6.4 and 7.0 h, respectively, p < 0.05) but not between the monotherapy groups. There were significantly higher rates of ICU admission and hypotension when PB and BZDs were used together (8.6% and 15.4%) versus either agent alone (PB 2.9% and 5.7%, BZD 3.8% and 4.5%, p < 0.05).

Conclusion

Among patients with multiple visits presenting with alcohol withdrawal, treatment with PB, BZDs, or both did not result in significantly different rates of admission or readmission within 48 h. Receiving a combination of PB and BZDs was associated with significantly longer ED length of stay, more ICU care, and increased incidence of hypotension as compared to either PB or a BZD alone.

Introduction

Emergency department (ED) visits related to alcohol are a common and increasing occurrence. ED visits in the United States related to alcohol consumption increased more than 60% from 2006 to 2014 [1]. This increase is related to both acute and chronic complications of alcohol consumption, including withdrawal. Pharmacological treatment of alcohol withdrawal syndrome (AWS) has traditionally relied on benzodiazepines (BZDs), specifically lorazepam and diazepam. However, increasingly literature has supported use of phenobarbital (PB) either as an adjunct or primary agent for AWS [[2], [3], [4], [5], [6], [7], [8]]. In one ED trial, a single dose of PB in addition to symptom driven lorazepam was shown to reduce intensive care unit (ICU) admission rates and less overall BZD doses [3]. As monotherapy, symptom-triggered PB was shown to be equivalent to symptom-triggered lorazepam followed by an outpatient BZD taper for reducing withdrawal symptoms during the ED visit and for 48 h after discharge [4]. Other trials have demonstrated benefit with phenobarbital in inpatient and critical care settings but there is continued debate about its role for AWS in the ED. [[5], [6],[8], [9], [10], [11]].

Diazepam and PB have long half-lives (20–50 h and 53–140 h) that exceed the duration of ED visits posing the potential for adverse events once patients leave the ED. [12] Studies have shown that up to 44% of AWS patients have return visits for withdrawal, but to-date no study has looked at the optimal treatment for the acute management of patients with high-recidivism [[11], [12], [13]]. The purpose of this study was to assess the efficacy and safety of PB, with or without adjuvant BZDs, for treatment of acute alcohol withdrawal in the ED caring for a large urban patient population with high rates of recurrent ED visits for withdrawal.