Return Encounters in Emergency Department Patients Treated with Phenobarbital Versus Benzodiazepines for Alcohol Withdrawal

Background: The emergency department is frequently visited by patients suffering from symptomatic alcohol withdrawal, and the traditional management has been dominated by repeated doses of benzodiazepines.1, 2 Phenobarbital has been studied as an alternative or adjunct to benzodiazepines with encouraging results, but primarily among inpatient services.3, 4 Much of the appeal of phenobarbital lies in its ability to stave off severe alcohol withdrawal symptoms days after administration, without the necessity of daily redosing.5  

Paper: Lebin J et al. Return Encounters in Emergency Department Patients Treated with Phenobarbital Versus Benzodiazepines for Alcohol Withdrawal. J Med Toxicol 2022. PMID: 34697777

Clinical Question: In patients with moderate to severe alcohol withdrawal being discharged from the emergency department (ED), does treatment with phenobarbital alone and phenobarbital plus benzodiazepines compared to benzodiazepines alone decrease the odds of returning to the ED within 3 days after initial discharge?

What They Did:

• Retrospective cohort study at a single academic medical center between July 1, 2016 to June 30, 2019.
• Electronic Health Record (EHR) database review using keywords for demographics, triage vital signs, and administered medications.  
• History of liver disease, history of substance abuse disorder, and history of delirium tremens were also extracted from the hospital EMR.
• 470 patients found that were treated with phenobarbital, a benzodiazepine, or a combination of both and then discharged.
  • For comparison’s sake, the medications administered were converted to benzodiazepine equivalents in milligrams of lorazepam.
• Authors assessed how many of the 470 had return ED visits (for any reason) within 3 days of discharge, within 3-7 days, and after 7 days of discharge.

Inclusion Criteria

• Adult patients discharged with diagnosis of “alcohol withdrawal” from the ED

Exclusion Criteria

• Incarcerated 
• Transferred or discharged to another facility
• Patients who were not treated with any pharmaceuticals for alcohol withdrawal, interpreted by the authors as indicating “mild” withdrawal or predominance of alternative diagnosis.

Outcomes

Primary

• Return encounter within 3 days to any ED 

Secondary 

• Return encounter between 3-7 days to ED 
• Return encounter within 7 days to ED
• Diagnosis and disposition for follow-up encounter 
• Survival to beyond 3 days after index ED visit

Results 

• Return rate (BZD vs PB vs PB+BZD) to the ED
  • @ 3 days: 25% vs 13% vs 10% (p=0.001)
  • >3 to 7 days: 18% vs 7% vs 10% (p=0.36)
  • @ 7 days: 38% vs 18% vs 20% (p=0.003)

• PB and PB+BZD group had lower odds ratio of return ED visit at three days compared to BZD group
• No statistically significant difference in odds ratio of return ED visit at three days between PB and PB+BZD groups
• No statistically significant difference in odds of return visit among treatment groups between 3-7 days 
• This result persisted when the odds ratio was adjusted for age, history of delirium tremens, history of substance use disorder, and history of liver disease.
• There were no significant differences in the prevalence of liver disease, history of substance use disorder, history of delirium tremens, mean age, triage heart rate, or triage blood pressure across treatment groups.
• No statistically significant difference in survival beyond 3 days 
• 92 of the total 120 patients who returned to the ED were associated with a diagnosis of alcohol withdrawal or intoxication. 11 of those were admitted to the hospital.

Strengths

• Investigators asked a clinically valuable question that is patient-oriented.
• Patients selected consecutively and not a convenience sample, reducing the chance of selection bias.
• Patients in the study groups were similar regarding demographics.
• Possible confounders such as pre-existing liver disease, history of delirium tremens, history of substance use disorder were accounted for in results
• Patients being studied were ED patients
• One of only a few studies looking at outpatient efficacy of phenobarbital after administration in the ED

Limitations

• Endpoints collected did not include things such as resolution of withdrawal symptoms in the ED.
• Retrospective in design
• No randomization, and therefore possibly unbalanced groups
• Did not include enough possible confounders, such as socioeconomic status, unhoused status, presence of other intoxicating substances on drug screen, coexistence of complicating acute medical illness.
• Did not account for patients with polypharmacy use, which would have potentially altered the required dosing of benzodiazepines
• Did not account for withdrawal severity determining choice to use phenobarbital or not, leading to confounding by indication
• Single center 
• Clinical Institute Withdrawal Assessment (CIWA) scores – or some other more universal diagnostic tool for alcohol withdrawal – not routinely used, so dose variations were secondary to physician choice rather than objective measurement
• Alcohol withdrawal is a clinical diagnosis, and therefore could have been inappropriately applied to some of the patients in the study
• No protocolized dosing or choice of medications – therefore patients with more severe alcohol withdrawal may have been selected to receive phenobarbital vs benzodiazepines alone, allowing for potential bias in selection
• Did not examine patient use of discharge prescriptions or undocumented outpatient medication
• Patients may have presented to EDs outside of the included catchment area, and thus were not included in the study

Discussion

• This study is a useful addition to the body of literature supporting the treatment of alcohol use disorder in the ED with phenobarbital, but it is unable to establish a causal relationship between the use of phenobarbital for alcohol withdrawal and decreased ED bounce backs in the short-term.
• Phenobarbital also has some other benefits making it a great first-line agent for alcohol withdrawal in the ED.
  • Long half-life could obviate the need for discharge medications.
  • Provides ongoing GABA agonist therapy after ED discharge during the time period of highest odds of withdrawal symptoms.
  • Could be used as a therapeutic bridge during the initial days of treatment until patients are able to get into community care or other medication assisted therapy treatments.
• It is unclear if the difference in return encounters is due to differential use of benzodiazepine equivalents, or the inherent pharmacodynamics of phenobarbital. There was a massive difference in benzodiazepine equivalents between treatment groups: 26 mg and 28 mg in the PB and PB+BZD groups, respectively, and 6 mg in the BZD group. Accounting for outpatient prescriptions or out-of-hospital use of other undocumented GABA agonists may have accounted for this vast divergence in dosing, but the authors justifiably remark that this information, even if gathered, is notoriously inaccurate. 
• CIWA scores or some other more universal diagnostic tool for alcohol withdrawal were not routinely used so we have no idea why one patient got more than another.
• Tracking the success or failure in the treatment of alcohol withdrawal in patients discharged from the ED is a difficult task given the inherent tendency of substance use disorders to create instability in the lives of those afflicted by them. However, it is precisely because of this instability that phenobarbital presents an attractive alternative: it’s given as a single dose in a monitored setting, and it covers the first few days of treatment, obviating the need for frequent redosing and complicated medication management. 

Author’s Conclusion

“In this cohort of patients with acute alcohol withdrawal who were discharged from the ED, those who received phenobarbital were less likely to return to the ED within three days of the index visit when compared to those who received benzodiazepines alone.”

Our Conclusion

Phenobarbital appears to be an effective treatment of alcohol withdrawal when given alone or in combination with benzodiazepines. Although hampered by its retrospective design and lack of protocolized regimen, this study supports using phenobarbital as a first-line agent in the emergency department for treatment of alcohol withdrawal. 

Clinical Bottom Line:

Phenobarbital is a useful alternative or adjunct to benzodiazepines in the treatment of alcohol withdrawal in the emergency department. 

Guest Post By:

J.D. Cambron, DO
Assistant Professor
Department of Emergency Medicine
Christus Health/Texas A&M University School of Medicine

References:

(1) Kosten, T. R.; O’Connor, P. G. Management of Drug and Alcohol Withdrawal. New England Journal of Medicine 2003, 348 (18), 1786-1795. DOI: 10.1056/NEJMra020617.

(2) Schuckit, M. A. Recognition and management of withdrawal delirium (delirium tremens). N Engl J Med 2014, 371 (22), 2109-2113. DOI: 10.1056/NEJMra1407298  From NLM.

(3) Ibarra, F., Jr. Single dose phenobarbital in addition to symptom-triggered lorazepam in alcohol withdrawal. Am J Emerg Med 2020, 38 (2), 178-181. DOI: 10.1016/j.ajem.2019.01.053  From NLM.

(4) Rosenson, J.; Clements, C.; Simon, B.; Vieaux, J.; Graffman, S.; Vahidnia, F.; Cisse, B.; Lam, J.; Alter, H. Phenobarbital for acute alcohol withdrawal: a prospective randomized double-blind placebo-controlled study. J Emerg Med 2013, 44 (3), 592-598.e592. DOI: 10.1016/j.jemermed.2012.07.056  From NLM.

(5) Stehman, C. R.; Mycyk, M. B. A rational approach to the treatment of alcohol withdrawal in the ED. Am J Emerg Med 2013, 31 (4), 734-742. DOI: 10.1016/j.ajem.2012.12.029  From NLM.

Post Peer Reviewed By: Salim R. Rezaie, MD (Twitter: @srrezaie)