Episode 72- "Alcohol Withdrawal? Try Phenobarbital!": Part 2: Dosing + Monitoring

Hello and welcome to Episode 72 of ER-Rx and Part 2 of the Phenobarbital Mini Grand Rounds Series. Last week in Part 1 of the series, we talked about the discovery and the early uses of phenobarb, hit on some pharmacology (which hopefully you stayed awake for), and briefly touched on why phenobarb works so well for alcohol withdrawal. This week, we’re going to spend the episode hitting some key features of phenobarb, including providing some evidence for its use and then getting into the details of how to actually give it to your patients. I want to be respectful of your time, so I’ll try to pack in as much info as I can as quickly as possible.

Since most of us are very comfortable using benzos and symptom-based treatment of alcohol withdrawal using CIWA scores, some of us may just not want to try phenobarb. But back in 2017-ish there was an IV benzo shortage, and that’s when our hospital, and maybe even yours, started using more IV phenobarb, and kind of forced you to do so.
And this isn’t something completely cowboy-ish we were doing- even though it felt that way at the time. I know the first time I saw it used I really had to step out of my comfort zone. While new to me at the time, phenobarb is supported by a good amount of literature, some of it coming from the ER setting, showing that it works at least as well as benzos, maybe even better.

For example, in 2013 Rosenson, et al published a prospective, randomized, placebo-controlled study in 102 patients in the ER. They compared giving one dose of IV phenobarb (10 mg/kg) vs normal saline along with a lorazepam-based alcohol withdrawal protocol. They found the phenobarb group had less ICU admissions (8% vs 25%, CI 4-32%) with no differences in length of stay or adverse events.
Then in 2018, Tidwell, et al published a retrospective study in a MICU showing that patients who got phenobarb had significantly shorter lengths of stay compared to those who got CIWA-based treatment (2.4 vs 4.4 days, P <0.001), significantly less need for mechanical ventilation (2% vs 23%, P <0.001), and significantly less use of adjunctive agents for symptom control  (7% vs 28%, P =0.002).
There are plenty other studies that I’m not going to bore you with, including the one we’ll discuss next time in the final installation of the series, that also show that phenobarb works well and is very safe to use.

My site’s guideline for phenobarb limits its use to patients < 70 years old without significant liver disease (cirrhosis or ALT/AST > 6x the upper limit of normal) or respiratory disease (chronic hypercapnia or anyone who is oxygen- dependent). When we first started using phenobarb, we used the PAWSS score to determine whether or not patients were good candidates, with a score >/= 4 being someone at high risk for severe alcohol withdrawal who would most likely benefit from getting it. But since that time we’ve moved away from this practice, instead recommending the patient be at a RASS of 0 or higher – which for our purposes typically ends up being a patient scoring in the moderate-severe category based on CIWA scores. The main point here is that it wouldn’t be a great idea to give phenobarb to a patient who is lethargic.

And here’s another kicker that all listeners may not agree with: we strongly recommend against giving it to patients who already received a significant amount of benzos prior to getting phenobarb, because there is a much higher risk of respiratory depression and over-sedation when combining these agents. This goes back to the previous episode where I talked about the increased risk of adverse events when combining phenobarb and benzos given their synergistic effect on the GABA-A receptor. And this isn’t just a hypothetical risk—this risk has been described at hospitals that do use the two agents in combination. Make sure to hear more about this in Part 3 of the series.

As far as dosing, we give a 10 mg/kg IV loading dose based on ideal body weight (IBW) over 30 minutes, then we allow 65 mg IV q2H as needed based on symptoms. To put this into perspective, this loading dose of 10 mg/kg based on IBW is only half of the dose we use to treat seizures. And remember last time when I mentioned the linear relationship between dose and drug level? Well, this 10 mg/kg loading dose should get your patient to a level of ~ 15 mcg/mL, which is within the lower end of the typical range of 10-40 mcg/mL we shoot for when treating seizures and is much lower than the toxic level of ~ 60 mcg/mL. This means that with this loading dose, the level we can expect shouldn’t be overly sedating and is very safe- especially if your patients don’t have other sedating meds like benzos on-board, and of course again assuming they had at least moderate severity withdrawal or a RASS of at least 0 prior to getting it.

But there are of course other ways of giving phenobarb, which makes it very flexible to use. Other sites give 65, 130, or 260 mg IV as needed until symptomatic control, then just use 65 mg or so as needed. Remember that we should be tracking how much phenobarb we gave, with a limit of about 20 mg/kg, and definitely not more than 30 mg/kg as a total dose. If you have to use this much phenobarb and your patient is still in withdrawal, then it may be a good time to start thinking about other adjuncts like dexmedetomidine (Precedex). Shout out to the EmCrit Blog for a couple of very detailed posts regarding the use of phenobarb—make sure to check those posts out.
In case you’re wondering why the weird doses—it’s because phenobarb comes as a 65 or 130 mg/mL injection. We store these vials in our Pyxis machines in the ER, so they are readily available for as needed dosing. Another nice thing about keeping the vials near is that if the infusion is taking a while to come from IV pharmacy, we can bolus using the vials, then just have pharmacy send up the rest of the load in an IV piggyback.

We recommend patients are on progressive or ICU level of care for loading doses, most of which are given in the ER at my site, and we do admit patients to general care from the ER only as long as the patient hasn’t received a significant amount of benzos along with the phenobarb prior to transfer.
Our amazing nurses monitor vitals q15 minutes for the first hour, then q4h for days 1 and 2. Basically, they think of this as giving a strong benzo with very similar things applying in terms of adverse effects and things to monitor.
Levels should and can be checked in patients with kidney or hepatic dysfunction, if there are significant drug interactions (for example, HIV meds), or signs of phenobarb toxicity which can manifest as slurred speech, nystagmus, or of course respiratory or cardiac depression.

One final point before we end the episode; for the most part, we aren’t super comfortable with discharging a patient after giving IV phenobarb in the ER. Of course, there are objections, and this decision in part depends on the total mg/kg dose your patient got and how confident you are in the patient’s willingness to not drink alcohol or take other sedatives once you discharge them.
If you’re at a site where they do discharge after giving a phenobarb load- please make sure to educate the patient on the risk of adverse events, including death, if they do drink or use other sedatives with phenobarb still in their system, and I probably wouldn’t send patients home with any other meds except maybe some anti-emetics for symptomatic control.
If your patient does need more than a total dose of 10 mg/kg or still don’t look good, please consider admission, especially if you gave more than 20 mg/kg total or used a combination of benzos and phenobarb to control their symptoms.

In conclusion, we have a good amount of data showing phenobarb to be very safe and effective at treating alcohol withdrawal- even in the ER and ICU settings. If your patients are otherwise healthy and are in moderate-severe alcohol withdrawal, consider trying IV phenobarb, either as a bolus load of 10 mg/kg based on IBW, or giving smaller doses, like 65, 130, or 260 mg until you and the patient feel comfortable. Look out for over-sedation or phenobarb toxicity with constant patient monitoring, and try to avoid giving benzos in combination with the phenobarb—just try to stick to one agent or the other.

As always, thank you so much for your time, and thank you for wanting to learn more about pharmacotherapy. If you have any comments or anything you’d like to add to this episode, or if you want to see my site’s phenobarb protocol, please give me a shout out on the @errxpodcast Instagram page, or reach out to me on errxpodcast.com.

And finally, I’d also like to shout out two listeners- who chose to stay anonymous- for their donation on buymeacoffeee.com. Thank you so much for supporting the show and keeping it free for everyone. Stay tuned for next week’s episode and the final installment of the 3-part phenobarb series, where we bring on the author of a study that discusses phenobarb and/ or benzos for recurrent alcohol withdrawal in the Emergency Department.