Phenobarbital for inpatient alcohol withdrawal (Bowen, 4/3/2024)

 A recording of this presentation can be found HERE.

***

Thanks to Dr. Anna Bowen, our future addiction medicine fellow (2024-2025), who gave an excellent talk this week on the use of Phenobarbital for Inpatient Alcohol Withdrawal. Dr. Bowen compiled the evidence for phenobarbital as well as a variety of phenobarbital protocols from different health systems. She urged us to create/modify one of our own at SSRRH, which seems like an excellent idea. 

My notes on her presentation:

• Alcohol use disorder is very common in the US! 10.5% of the US population (+29 million people) over age 12 quality has having an AUD
• In 2021, there were an estimated 178K deaths related to excessive alcohol use in the US. Alcohol is one of the leading causes of preventable deaths in our country
• Complicated alcohol withdrawal syndrome (AWS) occurs in 5-20% of hospitalized patients
• Alcohol use is responsible for nearly double the number of deaths per year as opiates

Most of us are well aware of the time course for AWS. See image below for a reminder. Remember that alcoholic hallucinosis, which often occurs early in the course, is a distinct entity from the more life-threatening delirium tremens (DTs). 

Our standard treatments for AWS include: 1) Benzodiazepines (the historic gold standard), 2) anti-epileptic drugs (including valproic acid, gabapentin, carbamazepine), and 3)Alpha-adrenergic blockers (clonidine, precedex and guanfacine).

Phenobarbital is a new old drug! It has recent (and growing) evidence for non-inferiority to the historic goal standards -- benzodiazepines-- as treatment for AWS. And in several studies, it is superior. It acts at both the GABA and NMDS receptors.

Studies of phenobarbital in AWS demonstrate:

• shorter ICU stays, decreased ICU admissions
• shorter hospital length of stay (LOS)
• lower rates of mechanical ventilation
• lower rates of adjuncts (including precedex, benzos)

Granted these studies, most recent -- between 2017 and 2023-- are all small (~100 patients each) and there are variable comparisons between different protocols, BUT in taken in totality, phenobarbital seems to have similar and/or better evidence for treatment of AWS compared to traditional treatment. If the goal is to decrease ICU admits and/or use of benzodiazepines in at-risk populations, phenobarbital is probably the tool of choice. 

So, how should it be used?

Dr. Bowen showed us 5 protocols from 5 different healthcare systems: Swedish (WA), BMC (MA), Yale (Connecticut), Contra Costa (CA) and ZSFGH (CA). While they have their own unique differences, there are three main steps to using phenobarbital in AWS:

1) Assessing risk for AWS (i.e. risk stratifying patients who screen positive for AUD for high risk AWS)

2) Understanding the risks/benefits of phenobarbital in different populations

3) Dosing protocols

Assessing for risk of Severe AWS

The benzo-sparing protocol that we currently use in the hospital -- often referred to as the Stanford Protocol --  uses the PAWSS score to risk-stratify for high risk AWS. Alternate protocols use other criteria to designate patients as high risk. These high risk characteristics include 1) a history of intubation or prior ICU stay for AWS 2) BAL>200 with signs of withdrawal despite high BAL, and/or 3) a personal history of seizures or DTs.

Whatever the methodology, the first step in treating AWS is the determination of this risk stratification. This will lead you to either treat or not treat prophylactically for AWS.

Understanding Risks/Benefits and Contraindications of Phenobarbital

Those of us who didn't "grow up" using phenobarbital may be less familiar with the pharmacology and/or risks and contraindications for phenobarbital. Here are a few important ones:

• Concomitant high dose administration of benzodiazepines. Several protocols use <8mg lorazepam as an acceptable cut-off for liberal admin of phenobarbital. If a patient has received >20mg of lorazepam, phenobarbital is contraindicated
• Respiratory depression (as a result of benzos or otherwise) makes patients high risk
• Advanced cirrhosis WITH active hepatic encephalopathy and/or AMS is considered a contra-indication. But a prior history of HE or well controlled HE is not. Of note, cirrhosis itself is NOT a contraindication to use of phenobarbital
• Patients on chronic seizure meds (including phenobarbital) is a relative contraindication
• Pregnancy
• Renal dysfunction (CrCl<30)
• History of allergy and/or Steven's Johnson syndrome

It is important to check for drug-drug interactions with phenobarbital (warfarin, OCPs, furosemide). 

Also important to know that the half-life of phenobarbital is 40-140 hours, this is VERY long, which is a big reason it is desirable to use in AWS (which can go on for days), but also to know it will stay in a patient's system and can have an impact over several days time. 

Patients with cirrhosis will clear phenobarbital more slowly, so a PO taper and/or repeated doses are less needed

AND that if administered slowly (either via slow IV or IM), the risks of respiratory depression as a negative side effect are decreased. 

Phenobarbital Protocols

Actual administration and dosing protocols vary widely across systems. Variations include: 

• Weight based vs. standard dosing
• 5 to 10 to 15mg/kg IDEAL BODY WEIGHT loading doses vs. 130mg vs. 260mg doses
• Single dose vs. subsequent 
• as often as q30-60 minutes dosing with respiratory effects noted prior to additional dosing
• IV vs. IM administration (vs. PO)
• +/- PO taper after initial admin
• Location of administration
• ED vs. ICU vs. step-down units vs. medical floor patients

See images below for some of the protocols. Note variations in dosing, admin, etc. The final image is a table comparing the main components of the protocols:

summary Table comparing different phenobarb protocols in AWS

Here's what I took home from Dr. Bowen's talk:

• Phenobarbital is a safe tool in our toolbox for management of AWS in many of our high risk patients
• We should risk assess all patients for AWS and consider phenobarbital for patients who we deem "high risk"
• Phenobarbital should ideally be administered early -- within the first 24 hours of that risk assessment
• We should be careful about the use of all adjuncts -- including benzos, as well as benzo-sparing protocols -- when using phenobarbital due to its long half-life and risks for respiratory issues
• We need a system for monitoring respiratory depression in patients who receive phenobarbital
• We can give a large loading dose (10-15mg/kg IBW) in a highly monitored environment (e.g. ED or ICU) or lower amounts (e.g. 130-260mg x 1) with frequent rechecks for respiratory depression
• Phenobarbital is safe in cirrhotic patients (except those with active HE) but with the understanding that it probably stays around even longer for these patients. No need for PO taper. 

I look forward to collaborating with Dr. Bowen and our ED in the coming months so that we can use phenobarbital more systematically for our high risk patients for AWS.