Episode 94- “Don’t Set the Precedent”: Precedex for Alcohol Withdrawal
Are you using Precedex for alcohol withdrawal? Hear my take on why I don’t like this idea, with evidence from a recently-published meta-analysis.
“‘Don’t Set the Precedent’: Precedex for Alcohol Withdrawal”:
– Alcohol withdrawal is treated mainly by giving benzodiazepines (or phenobarbital). These work by binding GABA receptors and inhibiting NMDA receptors—basically imitating ethanol’s effects
– Given the side effects of these agents there has been some interest in using adjunct agents – like Precedex (dexmedetomidine) to help treat withdrawal. Precedex causes anesthetic and sedative effects through agonizing alpha-2 adrenergic receptors in the brainstem, inhibiting norepinephrine release. But it can also cause bradycardia through its effects on presynaptic alpha-2-adrenergic receptors and hypotension through peripheral vasodilation
– Importantly, Precedex should never be used alone in alcohol withdrawal. It’s mechanism of action doesn’t affect GABA or NMDA receptors – it doesn’t treat the underlying pathophysiology of the disorder. All it does is reduce the autonomic symptoms of withdrawal and reduces CIWA scores
– In a recent meta-analysis, the authors looked for randomized controlled trials and cohort studies specifically looking at Precedex in combination with benzos for critically ill adult patients with alcohol withdrawal- excluding case reports and case series
– When looking at the primary outcome of ICU length of stay in hours, the cohort group analysis found a mean difference of +48 hours (37.48-58.64, P = <0.01), significantly favoring the benzo monotherapy group. In a post-sensitivity analysis, the RCT group also found a mean difference of +50 hours (38.95- 61.56, P <0.01), again favoring the benzo monotherapy group
– The cohort studies showed a significant difference favoring the benzo monotherapy group in hospital length of stay (53 hours; 18.47-87.89, P=0.003). The RCT group showed no difference in the study arms in a post-sensitivity analysis. There was no difference between groups in total benzo requirements but there was, unsurprisingly, significantly more bradycardia and hypotension in patients who got Precedex
– In conclusion, Precedex had at best no effect on ICU length of stay and at worst increased it by 2 days! It also didn’t have any effect on other secondary outcomes- besides causing more bradycardia and more hypotension. Instead of adding Precedex, let’s treat the underlying issue itself with benzos or phenobarbital. If you are going to use it- remember to schedule benzos along with it
What’s up everyone and welcome to Episode 94 of ER-Rx- a podcast tailored to your clinical needs. I’m your host, Adis Keric. This week, I want to talk about a meta-analysis published in one of my favorite journals—Annals of Pharmacotherapy- that discusses the use of dexmedetomidine (Precedex) in patients going through alcohol withdrawal. This is really timely for me because I feel like I’m seeing an uptick in the use of Precedex for this indication in my own shop. I don’t know about you guys, but for a while I didn’t even think we were allowed to use Precedex in the ED. I used to tell the ED providers that they just couldn’t order it based on an old policy I found that said it was restricted to ICU providers- until some of my coworkers made fun of me and showed me a newer policy saying that we CAN, in fact, use it in the ED—oops. Of course, as a resident and when I used to staff ICUs more, I’d see it often, but I always wondered just how effective it truly was. To be honest, I was never really impressed, and maybe that’s why I subconsciously thought it just wasn’t allowed in the ED. Was my hunch correct? Let’s look- starting with some background information.
Nearly 15 million people in the US has alcohol use disorder, and about 40% of ED visits had it listed as a comorbidity. Half of all ED trauma patients have withdrawal as evidenced by CIWA scores of > 20. The pathophysiology of alcohol withdrawal isn’t fully understood, but essentially, when we expose ourselves to alcohol repeatedly, our brains undergo some pretty significant changes. These include up-regulation of excitatory NMDA receptors along with inhibitory GABA receptor down-regulation and insensitivity- meaning that we need higher levels of GABA (or ethanol) to maintain the body’s baseline. When we suddenly stop drinking, the scales are tipped in favor of excitatory glutamate’s binding to NMDA receptors and an inability of our own down-regulated GABA receptors to compensate. This is how patients develop the symptoms of withdrawal, delirium tremens, and seizures. To treat symptoms of withdrawal and prevent these things from happening, the gold-standard treatment is giving benzodiazepines. Benzos work by tipping the scales back in favor of inhibitory tone. They bind GABA receptors and inhibit NMDA receptors—basically imitating ethanol’s effects. As a side note, along with benzos, another – and maybe even better- treatment option here is phenobarbital. It works very similarly to benzos – and to save us time I’ll just direct you to my previous talks on it back in Episodes 71, 72, and 73.
The issue is that benzos (and phenobarbital) are associated with side effects like respiratory depression, and some have alcohol withdrawal that is refractory to benzos (another side note—it’s very rare to have withdrawal that is refractory to phenobarbital). So of course, since we don’t have the absolutely perfect agent without any side effects whatsoever to treat something, people get curious about alternatives.
This is where Precedex comes in. Precedex is a cool drug. It’s a continuous IV medication that causes anesthetic and sedative effects through agonizing alpha-2 adrenergic receptors in the brainstem, inhibiting norepinephrine release. The bummer is that it can cause bradycardia through its effects on presynaptic alpha-2-adrenergic receptors and hypotension through peripheral vasodilation. The nice thing about Precedex is that it can reduce autonomic symptoms while still allowing the patient to be arousable and cooperative without causing respiratory depression. But a very important thing to remember here is that Precedex should never be used alone in alcohol withdrawal. It’s mechanism of action doesn’t affect GABA or NMDA receptors at all- so it doesn’t treat the underlying pathophysiology of the disorder. All it does is reduce the autonomic symptoms of withdrawal and reduces the CIWA score, but that isn’t necessarily a good thing—more on this later.
Given these effects, there’s been some who use it with benzos for alcohol withdrawal. Some studies show lower ICU and hospital lengths of stay, a decrease in total benzo usage, and better CIWA scores- but results are very mixed, and we’re still not sure if these outcomes are actually beneficial to our patients. To try to make sense of all of this, the authors of the study I’m going to talk about today wanted to run a meta-analysis to try to weed out its true effects. The authors looked for randomized controlled trials and cohort studies specifically looking at Precedex in combination with benzos for critically ill adult patients with alcohol withdrawal- excluding case reports and case series and only included studies with specific outcomes of interest like ICU and hospital LOS, total benzo usage, and the incidence of bradycardia or hypotension. Out of almost 2,000 studies screened, they ended up including only 12 in the systematic review and 7 in the meta-analysis. Another cool thing they did was separate randomized controlled trials (RCTs) and retrospective cohort studies when analyzing data- so we have pooled results from both types of studies.
Let’s get into the stuff you’ve been waiting for. When looking at the primary outcome of ICU length of stay in hours, the cohort group analysis found a mean difference of +48 hours (37.48-58.64, P = <0.01), significantly favoring the benzo monotherapy group. However, the RCT group had a result of -20 hours (-36.86- -3.28, P=0.02), significantly favoring the Precedex-in-addition-to-benzo group. But, in a post-sensitivity analysis they found a mean difference of +50 hours (38.95- 61.56, P <0.01), again favoring the benzo monotherapy group. What does this mean? When looking at ICU length of stay, we can with some confidence say that adding Precedex doesn’t shorten the number of hours someone spends in an ICU, and that based on low-quality evidence of data, it may even increase it by about 2 days.
Let’s quickly look at some secondary outcomes. When talking about hospital length of stay, again the cohort studies showed a significant difference favoring the benzo monotherapy group (53 hours; 18.47-87.89, P=0.003). The RCT group showed no difference in the study arms in a post-sensitivity analysis. When looking at total benzo requirements in mg, there was no difference between groups. When looking at side effects, there was, unsurprisingly, significantly more bradycardia and hypotension in patients who got Precedex.
In conclusion, Precedex had at best no effect on ICU length of stay and at worst increased it by 2 days! It also didn’t have any effect on other secondary outcomes- besides causing more bradycardia and more hypotension.
There are many reasons why Precedex can increase the length of stay. One confounder is that maybe sicker patients, or those with higher CIWA scores, are the ones who the team thought would benefit from the addition of Precedex. But unfortunately, the studies included didn’t routinely list CIWA scores in the groups- so we’ll never know. Another potential reason is its cardiovascular side effects- patients who develop bradycardia or hypotension may need to be monitored more or may need treatment to reverse these effects. The last, and scariest explanation is that, as mentioned in the beginning of the episode, Precedex only masks withdrawal symptoms without treating the underlying physiology of the disease. This means that patients on Precedex score lower on the CIWA and potentially get less symptom-triggered benzos. Studies included in the analysis actually found that patients who got Precedex had higher rates of intubation and seizures; not because Precedex lowers respiratory drive or the seizure threshold, but because these patients had disease progression of their withdrawal symptoms, possibly because they were getting less benzos.
The other thing I want to touch on is this outcome of total benzo requirements. Why are we even looking into this? It’s not the cost of benzos- which despite being on shortage recently are pretty cheap. The argument is that if we are using less benzos, we’ll have less side effects like respiratory depression and intubation. But this is such a strange outcome to care about. Patients should get as much benzos as needed to make them feel comfortable and to prevent seizures. If you actually wanted to just not give any benzos (which is not recommended), but did happen in some of these studies, patients can develop seizures and irreversible encephalopathy.
This, to me, is very similar to our fear of using high-dose, guideline-recommended benzos in the setting of status epilepticus. Despite our fears of benzos causing more intubation in seizing patients, studies show the opposite: giving high dose benzos leads to less intubation than an inadequately treated seizure.
My take on all of this is that I was kind of right to not let providers order Precedex for alcohol withdrawal in the ED- I just didn’t know why. Just because we can use it based on hospital policy, doesn’t mean we should. Instead, let’s treat the underlying issue itself with benzos or phenobarbital. If for whatever reason you are going to use it- remember to schedule benzos along with it. I’ve seen things like diazepam 10-20 mg Q6H, in addition to nurse-driven, as needed benzos based on CIWA scores. Also, just know that you won’t be affecting their length of stay, and you’re potentially causing more side effects, seizures, and intubation, along with increasing drug costs by about $500 per day. Sounds like a pretty shitty precedent to me.
As always, thank you so much for your time, and thank you for wanting to learn more about pharmacotherapy. If you have a second, please follow and share the show on Apple Podcasts, Spotify, YouTube, and wherever else you listen to podcasts. Following, sharing, and rating the show are great ways to help the podcast grow and get more of our community involved. Also, if you are enjoying the podcast, please take a second to donate on buymeacoffee.com. Besides actually using the donation to buy coffee to power through an episode, these donations also help maintain the podcasting fees I pay and keep the podcast free, and ad-free, for everyone.
Also, if you have any comments or anything you’d like to add to this episode, please give me a shout out on the @errxpodcast Instagram page, or reach out to me personally on errxpodcast.com- I’d love to respond to all comments and criticisms.