Brian Fuehrlein, MD, PhD.
Addiction psychiatrist. Associate professor of psychiatry, Yale University. Director of the psychiatric emergency department at VA Connecticut. West Haven, CT.
Dr. Fuehrlein has no financial relationships with companies related to this material.
CATR: Can you run us through the natural course of alcohol withdrawal?
Dr. Fuehrlein: First, remember that alcohol withdrawal symptoms should be measured from the time since the last drink, not any particular blood alcohol level. Withdrawal symptoms start six to eight hours from the last drink. The CAGE screening is an easy way to remember this. The E stands for “eye opener,” meaning “I drink when I first wake up” (www.tinyurl.com/3nndts8j). This is typically due to the patient waking up in withdrawal after sleeping six to eight hours. The blood alcohol level may not be zero, but that does not mean the patient can’t have withdrawal. The early symptoms of withdrawal are typical ones from the Clinical Institute Withdrawal Assessment for Alcohol, revised (CIWA-Ar, usually referred to as the CIWA): tremors, headache, GI distress, and anxiety (www.tinyurl.com/2p86m5f9). If the withdrawal is mild, these should resolve in a day or so. But if the withdrawal progresses, the usual timeline is hallucinations in 12–24 hours from the last drink, seizures in 24–48 hours, and delirium tremens (DTs) in 48–72 hours from the last drink. It’s typically a linear progression. It’s rare that someone with low CIWA scores early develops DTs 48 hours later.
CATR: How should clinicians assess withdrawal severity early on?
Dr. Fuehrlein: CIWA is a validated scale, typically administered by nurses. Some questions are subjective: “Do you have a headache?” or “Are you anxious?” Others are objective: Nurses judge the degree of tremor or diaphoresis. Typically, a score of 8 or more indicates an alcohol withdrawal that should be addressed.
CATR: How do vital signs play into this?
Dr. Fuehrlein: Vital signs are not a part of CIWA, but they should be checked routinely as well. Usually, it’s an elevated CIWA score or an elevated pulse or blood pressure that triggers a call to the doctor or a dose of medication. Vital signs can be tricky, though. Patients can be dehydrated, be anxious, or have pain, and that can cause an elevated pulse. If someone has primary hypertension and isn’t taking their blood pressure meds, they can have elevated blood pressure. So it takes some clinical judgment to interpret vital signs accurately. If the pulse remains elevated despite good PO intake, that might indicate withdrawal. On the flip side, if the patient continues to have high blood pressure, but CIWAs are low and they have had a few doses of benzos, then it’s probably primary hypertension.
CATR: One criticism of the CIWA is that it can be manipulated.
Dr. Fuehrlein: Just like vital signs, CIWA requires interpretation. There are certain physical signs that I look for. Watching a patient drink out of a cup is helpful. If they have big tremors during the CIWA scoring, but can drink easily out of a cup, that suggests exaggeration, maybe in order to get an extra dose of benzos. But those struggling to drink out of a cup probably have significant withdrawal tremors. And there are some physical signs that are difficult to fabricate. Watch for tongue tremors, vomiting, sweating, and looking anxious when they aren’t aware that anyone is watching.
CATR: And what about hallucinations? How do they compare with hallucinations in psychosis?
Dr. Fuehrlein: Alcohol withdrawal hallucinations are different from hallucinations associated with schizophrenia. Hallucinations arising from a primary psychotic disorder are usually auditory and consist of voices. Patients often lack insight, may be responding to internal stimuli, and may present with negative symptoms. Alcohol withdrawal is typically more activating and often presents as odd visual, auditory, or tactile hallucinations (eg, bugs crawling on the skin). Withdrawal hallucinations are more likely to be visual rather than auditory. If a patient is becoming delirious, they’ll often walk over to the wall and pick at it, as if there’s a bug or a sticker on the wall.
CATR: And what if the patient is using other substances?
Dr. Fuehrlein: That’s a very common situation, and it can complicate how to interpret the vitals or CIWA. Symptoms can overlap, and it’s sometimes hard to tease out what is alcohol withdrawal and what is something else. It seems incredibly simple, but a good history is your best tool in this regard. Most of our patients have gone through withdrawal before; they know their bodies better than we do. I’ll ask, “Do you think it’s mostly the opiates affecting you or do you think it’s mostly the alcohol?” Usually, they know.
CATR: It’s particularly important to differentiate alcohol and opioid withdrawal since both could require medication treatment.
Dr. Fuehrlein: That’s true, and although symptoms of alcohol and opioid withdrawal overlap, there are some differences that you can use to tease the two apart. For example, opioid withdrawal tends to have more GI symptoms and feels like the flu with body aches and a runny nose, and pupils dilate. These patients look really run down. Alcohol withdrawal is not quite like that; it’s characterized more by anxiety, tremors, more physical activation, and agitation. And sometimes it’s a bit of trial and error. If someone might be withdrawing from alcohol and opioids, but you’re not sure, you can give some buprenorphine and see if that helps, or some benzos. Alcohol withdrawal is more dangerous, so if the patient seems to be getting sicker and sicker—with vitals trending in the wrong direction, tremors worsening, becoming confused or hallucinating—I’d recommend leaning in favor of treating alcohol withdrawal.
CATR: Buprenorphine and benzos can cause respiratory suppression when combined. Are you worried about this when treating comorbid alcohol and opioid withdrawal?
Dr. Fuehrlein: Not while in the emergency department (ED). Patients who present to the ED for both opioid and alcohol withdrawal management are often using dangerous amounts of opioids, most likely fentanyl, and maybe drinking a fifth of vodka or more, all at home in a completely uncontrolled environment. Here in the hospital, we might give a few milligrams of lorazepam and buprenorphine. That’s significantly less dangerous than whatever they’re using at home, and we have staff who are monitoring them very closely. I have never given a patient buprenorphine with a few milligrams of lorazepam and had them overdose in the ED. I would be very cautious prescribing these medications in the outpatient setting, but ordering buprenorphine with benzos in the ED does not concern me.
CATR: How do you efficiently triage patients with alcohol withdrawal in the ED? How do you know who can be discharged and who needs inpatient-level care?
Dr. Fuehrlein: It’s all about risk stratification. In general, if someone has experienced complicated withdrawal in the past, defined as DTs or seizures, they are at an elevated risk of experiencing them again. There’s a phenomenon called “kindling,” which says the more times a patient has gone through alcohol withdrawal syndrome or had complications, the more susceptible they become to reexperiencing complications in the future. Each time the patient comes back for withdrawal, all other things being equal, the risk of complications is a little bit higher. You can consider taking a quantitative approach to risk stratification with the Prediction of Alcohol Withdrawal Severity Scale (www.tinyurl.com/22e6cuzs; Maldonado JR et al, Alcohol 2015;50(5):509–518).
CATR: What other risk factors do you pay attention to?
Dr. Fuehrlein: Important risk factors to consider are older age, multiple medical problems, and certain medications, like central nervous system depressants, drugs that lower the seizure threshold, or drugs that increase the chances of delirium such as anticholinergics. The amount the patient drinks is also important; the rough marker we use for increasing complication risk is 20 standard drinks or more per day. For context, 750 mL of liquor—what patients often call a fifth—contains 17 standard drinks. We also look at the big picture. If someone doesn’t have any of these risk factors, they can likely be managed in the outpatient setting. The more risk factors a patient has, the more likely I am to recommend inpatient-level care and possibly even the ICU.
CATR: Are there specific signs and symptoms that lead you to admit a patient?
Dr. Fuehrlein: There are some straightforward signs that warrant an admission. Seizures, dangerously high blood pressure, sustained tachycardia, inability to take anything PO. These are definite medical admissions. Severe abdominal pain may lead to medical admission because it can be a sign of pancreatitis. Hallucinations aren’t dangerous in and of themselves, but they usually are a warning sign that DTs are ahead, and those patients usually need admission as well.
CATR: What medications are the best for alcohol withdrawal management?
Dr. Fuehrlein: Benzos are the gold standard. As far as I know, no definitive studies show that one is superior to the others. I favor lorazepam, which is a bit more rapid acting than other agents, has no active metabolites, and is familiar to most providers. An advantage of the shorter half-life is that we can easily hold a dose if a patient is overly sedated. But I work in an environment that has the staffing to check CIWAs and vitals every four hours. A very busy ED might prefer a longer-acting agent like chlordiazepoxide or diazepam so patients don’t need to be dosed and monitored as frequently.
CATR: What dosing protocol do you use?
Dr. Fuehrlein: We have a point-based algorithm, built upon the risk factors we discussed earlier. Each risk factor counts as a certain number of points, and the total number of points determines whether we consider a patient low, moderate, or high risk. Low-risk patients receive 1 mg of lorazepam every four hours as needed for elevated vitals or CIWA >8, with no standing lorazepam. Moderate-risk patients receive a standing 1 mg of lorazepam every four hours, and on top of that get lorazepam as needed for elevated vitals or CIWAs. The highest-risk patients get 3 mg every three hours standing—that’s 24 mg of lorazepam per day. That’s the maximum that we can do safely in the ED. For us, any patient needing more than 1 mg of lorazepam per hour should be admitted to inpatient medicine. We put a lot of thought into this protocol. I’ve worked at other institutions where all patients got the exact same dose of a long-acting benzo, which also worked well. Are we overthinking it? Maybe, but we see a complicated group of patients, and this protocol works well for us. However, this isn’t the only way to do it. And of course, we also give thiamine, folate, and multivitamins to everyone.
CATR: Does the presence of certain medical illnesses affect your medication choice?
Dr. Fuehrlein: It can, particularly when it comes to choosing a benzo. Lorazepam, along with oxazepam and temazepam, are safer for people with severe liver disease. We all remember the mnemonic from medical school: “They don’t take a TOL on your liver.” You should avoid benzos like diazepam and chlordiazepoxide in these patients (Editor’s note: See CATR March/April 2022 for more on treating addiction in patients with liver disease).
CATR: And what about phenobarbital?
Dr. Fuehrlein: Phenobarbital is an older medication that is becoming popular again. I don’t use it much in the ED setting myself, but I see it especially being used in the ICU. Studies show it can be useful either combined with a benzo-driven CIWA protocol or on its own (Nguyen TA and Lam SW, Alcohol 2020;82:23–27; Hawa H et al, Cureus 2021;13(2):e13282). It’s an intriguing approach, but we don’t have the evidence to definitively say when we should use phenobarbital versus when we should use a benzo. And phenobarbital’s long half-life and narrow therapeutic index warrant caution.
CATR: Is there evidence for adjunctive anticonvulsants?
Dr. Fuehrlein: There is some. We will use adjunctive anticonvulsants in the ED if there is a history of a seizure disorder. But remember, benzodiazepines are anticonvulsants themselves, so as long as a patient doesn’t have preexisting epilepsy, there shouldn’t be a need for other anticonvulsants as long as the patient is adequately treated with benzos. However, anticonvulsants have their place, particularly in an ambulatory setting when you want to minimize benzo prescribing. In those settings, anticonvulsants can be useful, particularly gabapentin at doses of 900–1200 mg per day, which can be a good substitute for lorazepam and might even reduce return to drinking (Myrick H et al, Alcohol Clin Exp Res 2009;33(9):1582–1588).
CATR: What other work do you do with patients in the ED besides treating with benzos and vitamins?
Dr. Fuehrlein: It’s important to keep in mind that withdrawal management is not treatment. You can use lorazepam or diazepam; check CIWAs frequently or infrequently; give the same dose to everybody or dose according to risk factors. These are all valid approaches, but the important thing is that patients are referred to treatment for their underlying substance use disorder. Motivational interviewing can be done right there in the ED. And whenever possible, refer patients to addiction services. Consider administering a dose of long-acting naltrexone before they leave. Withdrawal management is a necessary step, but it’s not where treatment ends—it’s where treatment begins.
CATR: Thank you for your time, Dr. Fuehrlein.
PO Box 626, Newburyport MA 01950