Inpatient alcohol withdrawal: a new approach (maybe) A small study contributes to the scant evidence by Phyllis Maguire
Published in the April 2012 issue of Today's Hospitalist
GIVEN THE NUMBER OF INPATIENTS who experience alcohol withdrawal, you would think there would be ample evidence on what works to curb seizures and delirium tremens while cutting down on excessive doses of benzodiazepines. While benzodiazepines remain the keystone of withdrawal treatment, high doses can cause a host of complications that include respiratory depression and more ventilator days.
But the reality is that hospitalists don't have much data on alcohol withdrawal in inpatients. Much of the available evidence has been drawn from outpatient populations or inpatient detox centers, and those studies exclude
"Probably the best treatment is going to be small doses of more than one drug."
–Jeffrey E. Lyon, MD Essentia Health
patients with significant medical comorbidities.
A study in the October 2011 Journal of Hospital Medicine makes a small but significant contribution to the literature. Researchers in two hospitals in Duluth, Minn., enrolled 79 patients deemed to be at high risk of withdrawal based on their withdrawal history or on patient-reported drinking patterns.
All of those patients received symptom-triggered benzodiazepine treatment. However, those who hit a score of 11 (out of a possible 67) on CIWA scoring were also randomized to receive either 10 mg of oral baclofen every eight hours or placebo. Out of the 79 patients enrolled, 31 ended up being randomized.
Researchers found that only 6% of patients in the baclofen arm needed higher doses of lorazepam (defined as 20 mg or more over 72 hours), compared to 54% of those taking placebo. Because these patients needed much lower benzodiazepine doses, the authors hailed the patient-safety benefits of baclofen as an adjunctive therapy and called for larger studies.
But that doesn't mean that they're now prescribing baclofen for all patients in withdrawal, says Jeffrey E. Lyon, MD, the study's lead author and associate chief of the hospital medicine division at Essentia Health.
"Baclofen is not FDA-approved for the treatment of alcohol withdrawal, although more people are using it in detox centers and in outpatient settings to decrease both symptoms and recidivism," Dr. Lyon says. At the same time, he points out, "we've certainly not seen any significant harm from low-dose oral baclofen in those who can take it."
Dr. Lyon spoke to Today's Hospitalist about the study results.
What made you decide to study the impact of baclofen on alcohol withdrawal in inpatients?
My colleagues and I were intrigued by a series of studies published years ago by researchers in Rome. One was a study of five outpatients who showed a dramatic improvement in their CIWA scores within a couple of hours after oral baclofen.
I thought the Italian studies really hit on something important in looking at baclofen, which is a GABA2-receptor specific drug. Our study tried to expand on that.
Have there been recent advances in treating alcohol withdrawal in the hospital?
The standard of care hasn't changed all that much in terms of what medications we use. The biggest change we've seen in the last 10 years has been going from a fixed-dose benzodiazepine regimen to a symptom-triggered regimen based on withdrawal symptoms and scores.
But that's not been uncontroversial, which probably has to do with the fact that the withdrawal assessment tools we have aren't perfect. There are a lot of other things that mimic alcohol withdrawal, from sepsis to other kinds of autonomic activation. So that leads to the overuse of benzodiazepines to treat alcohol withdrawal.
How often are physicians caught off guard by patients who suddenly go into withdrawal?
Hospitalists are certainly taken by surprise by significant withdrawal symptoms in some patients. These aren't patients who are in the hospital specifically for withdrawal. They're the patients who present with a broken leg or some type of unexpected event that lands them in the hospital and interrupts their usual drinking.
Some patients aren't very forthcoming about either the fact that they drink daily or how much they drink because there's a lot of shame involved. But there are also alcoholics who, though not exactly proud of drinking heavily, make no attempt to hide it from you. I always appreciate honest patients like that.
But the question of how to predict who is going to withdraw significantly is not easily answered. Obviously, the amount you drink is an important predictor, but even people who drink heavily don't always go into withdrawal—or one time they may and another time they may not. Having gone through severe withdrawal before is a positive predictor, as is concurrent illness. People with bad infections, heart attacks or even something as general as a high Apache score, which gives you an idea of how acutely ill they are, are more likely to withdraw as well.
I was struck by the number of patients in your study who were also on chronic narcotics for pain control: 41% among those who were randomized.
We don't know if someone who's taking narcotic medications is more or less likely to experience withdrawal, and we also don't know if narcotics make withdrawal worse or better. That's not something that we studied, but it's certainly something to look at—and it speaks to the reality that we face as physicians that so many people are on chronic narcotics.
Your study mentions that several drugs are being considered as adjunctive therapy. Is baclofen really the frontrunner?
It's one of a number of drugs being tried in a variety of settings related to alcohol addiction. I don't want to over-reach with our results; ours was a very small study, and we certainly do not advocate abandoning standard treatment in favor of substituting baclofen. The results were significant, but the role of baclofen in treating alcohol withdrawal still remains to be defined.
Our next study will hopefully look at baclofen as a way to prevent rather than treat alcohol withdrawal. We want to enroll people we think are at high risk for withdrawal, then randomize them to baclofen or placebo and see if the baclofen group doesn't go into withdrawal.
I understand that the literature on many issues related to alcohol withdrawal is sketchy, such as whether to use long-acting vs. short-acting benzodiazepines. What other areas of withdrawal research would you like to see pursued?
I think we need to understand the neurochemistry of alcohol addiction better. Baclofen has shown a lot of promise, and I'm hopeful that it's going to at some point be the standard of care to treat acute alcohol withdrawal more safely. But alcohol, which does not have a particular single receptor, has multiple effects on different neurotransmitter systems.
Probably the best solution and the best treatment is going to be small doses of more than one drug, rather than being able to treat it with just one.
Phyllis Maguire is Executive Editor of Today's Hospitalist.