This is one of two related, and complementary posts as part of a special series about ethanol in the ER. The first part can be found here.
Old-fashioned physician wisdom teaches us everything we need to know about managing intoxicated patients.
I have a cloth-bound edition of Austin Flint’s Medical Ethics and Etiquette (published 1883) on my desk. The hardcover, sea-green, handbook is 6 inches tall and about 100 pages long. I often thumb through it when I need to clear my head. Dr. Flint tackles topics that are still relevant today. Topics like dealing with the strain of long work hours, doctor-patient confidentiality (don’t talk in public about a lady’s tapeworm) and how to testify in court. Curiously, a controversial topic like euthanasia is handled in one sentence (he casually drops that this is part of his practice) while the question of what to do when the attending and consultant disagree is hashed out over four pages.
As the title suggests, the book mostly focuses on physician behavior; how to comport oneself in the presence of patients and other doctors. However, in rare moments, Flint wades into the waters of the practice of medicine. Much of this clinical advice doesn’t so much as tell you what to do, but cautions against doing doing too much. Flint tells us that the management of many diseases “. . . consists in close observation and watching for complications or untoward events, meeting indications as they arise in particular cases, palliating symptoms, relieving pain or distress, and sustaining the powers of life . . .” Flint might as well have said the words of that often-quoted doctor whose name is lost to time: “Don’t just do something, stand there.” Flint’s advice was prudent because many therapies of that generation were ineffective, or even dangerous. Nowadays, sitting back and not intervening may be bad advice for, say, bacterial meningitis, but the advice holds up well for alcohol intoxication. Flint’s admonition not to take unnecessary therapeutic actions came almost 80 years before Carl Clemmesen’s landmark “Scandinavian Method” paper told the world that barbiturate-poisoned patients would all do fine if we started caring for them and stopped trying to wake them up.1
Alcoholism was a problem in Austin Flint’s time as it is now. Most of our intoxicated patients do well without any intervention. If you are like me, you see a lot of these patients on Friday and Saturday nights, and also during the weekend days, and also during the week. Ok, you see them all the time. You may have a similar approach to me. I diagnose alcohol intoxication when the history and physical is consistent with it. Patients have a physical exam consistent with alcohol use – mild to moderate sedation, mildly depressed breathing or heart rate, and an odor that most of us would say smells like alcohol. (Ethanol is odorless. In most cases, we are smelling the so-called “congeners”, the other flavors and additives that are carried up to your olfactory bulb with the volatile ethanol.) By history, many patients will plainly tell you that they drank too much. I check a bedside glucose, perform a focused physical exam, and then park them somewhere with the promise to check on them every now and then so I can go see my next patient.
Don’t Routinely Check the Ethanol Concentration
Some clinicians check ethanol concentrations to determine if the patient is sober or to predict when they will become sober. Don’t do this. Sobriety is a clinical diagnosis. Patients are ready for discharge when they can ambulate safely and have a reasonable plan to get to their next destination.
The serum ethanol concentration – aka the alcohol level – is the D-Dimer of toxicology. I don’t obtain this lab unless I have doubts about the diagnosis. If I have a glimmer of doubt that the patient is “just drunk” because I can’t get a good history, I may check it. An alcohol concentration of 440 mg/dL would reassure me that intoxication is the cause, but a concentration of zero would send me running for the hills in search of another diagnosis.
Ethanol Elimination Rates Vary
We observe that most people tend to eliminate ethanol at constant rate of 15-25 mg/dL/hr. This zero-order elimination occurs because alcohol dehydrogenase becomes saturated at relatively low serum ethanol levels. In practice, there is little value to exploiting this fact clinically. First of all, many patients are still absorbing alcohol when they arrive, so their levels may initially seem to decrease very slowly or even go up.
Second, people do not eliminate at the same rate. Slow alcohol metabolizers – malnourished, people on low-protein diets, advanced cirrhosis with portal hypertension – may decrease their serum levels as by as little as 9 mg/dL/hr. On the other hand, ultra-rapid metabolizers – alcoholics with very high ethanol concentrations or genetic CYP2E1 polymorphisms – may be metabolize as much as 40 mg/dL/hr.2
What Would Be the Right Level to Wait for, Anway?
Even if someone had completely absorbed their alcohol and you knew their rate of elimination, what level would you target for discharge? If you use zero as your target, all of your patients will stay longer than they need to and some of your patients will withdraw, requiring further therapy and time. Some clinicians use a target of 80 mg/dL because of its significance as a DWI threshold throughout the US. The figure is actually somewhat arbitrary; the National Transportation Safety Board has suggested it be lowered to 50 mg/dL and other nations use thresholds from > 0 (any detectable alcohol) to 150 mg/dL. There are a few flaws with using a legal driving threshold to determine sobriety for your patient. First, drunk driving statutes use blood alcohol and most hospitals use serum alcohol. These are not the same. Ethanol does not distribute evenly between plasma and red cell fractions of whole blood and concentrations in serum are 15-20% higher than whole blood. However, the real issue is that there is nothing special about a blood alcohol concentration of 80 mg/dL. The use of this number is historical and political, not scientific.
When to Check Ethanol Level
So, I mentioned a few reasons why ethanol concentration is a waste of time, but when is the test clinically useful? One excellent use of the ethanol concentration is when you have a patient with unexplained metabolic acidosis and you want to rule out toxic alcohol ingestion. The presence of positive serum ethanol, which competes with a higher affinity for alcohol dehydrogenase, just about proves that your patient’s metabolic acidosis is not caused by a toxic alcohol like ethylene glycol or methanol. (To see more detail on labs in the diagnosis of toxic alcohol poisoning, check out fellow hound Meghan Spyres’s post.)
Leave the Intravenous Fluids Alone, Too
Most alcohol-intoxicated patients don’t need IV fluids. That bag of saline will not speed elimination of ethanol. This may seem counterintuitive, because some renally-cleared drugs clear faster with IV fluids. Lithium is an excellent example, as it competes with sodium for reabsorption in the proximal renal tubule.3 When patients are hypovolemic, the proximal tubule is salt-avid and pulls lithium back into circulation. Ethanol elimination, in contrast, is not volume-sensitive. In a small volunteer study, people were given ethanol on two separate occasions. The second time around, they received IV fluids right after the ethanol In both cases, the elimination rate was the same – 15 mg/dL/h.4
Give the Vitamins a Rest
I can understand the allure behind the “banana bag.” In addition to fluids, this concoction can contain magnesium and vitamins like thiamine and folate. It feels helpful to watch an IV bag drip futuristic translucent yellow goodness into a patient’s venous circulation. The usefulness of this therapy has been mythbusted here. In theory, it seems nice to give people vitamins (it has become a cottage industry in Vegas), but in practice most of our patients don’t need them. In the US, mandatory vitamin supplementation of food products has made vitamin deficiencies rare.
Rather than give everyone vitamins that they may not need, target therapy to your patient. If an alcoholic patient has some other reason for needing a venipuncture, you can check their magnesium concentration and look for megaloblastic anemia, a sign of folate deficiency.
I administer thiamine in select alcoholic patients. Thiamine deficiency can manifest with prominent neurologic (‘dry beriberi’ or Wernicke-Korsakoff syndrome) or cardiovascular (‘wet beriberi’) manifestations. Wernicke’s encephalopathy can be a challenge to diagnose because the presentation overlaps typical alcohol intoxication. The Caine criteria can aid in the diagnosis of Wernicke’s Encephalopathy:
Caine Criteria5 for Diagnosis of Wernicke’s Encephalopathy
Patients meet criteria if 2 out of 4 are present:
- Dietary deficiency
- Oculomotor abnormalities (e.g. ophthalmoplegia, nystagmus, gaze palsy)
- Cerebellar dysfunction (e.g. ataxia, abnormalities of past pointing, dysdiadochokinesia, impaired heel-shin testing)
- Altered mental state or mild memory impairment
“Wet beriberi” patients may have hemodynamic signs. The end stage of this condition is high-output heart failure and circulatory collapse. An early clinical clue may be a type B lactic acidosis (i.e., lactic acidosis without evidence of poor tissue perfusion). This is because thiamine is a cofactor in the pyruvate dehydrogenase complex, which is the gateway into the TCA cycle, converting pyruvate to acetyl-CoA. If that reaction can’t happen, pyruvate is converted to lactate, and less energy is available for cellular activity.6 So, consider the diagnosis of thiamine deficiency in alcoholic patients with lactic acidosis. You shouldn’t get labs on most of your intoxicated patients, but be on the lookout for those “drunk” patients that have tachypnea, even if it’s mild. Ethanol is a sedative, and should slow breathing. Tachypnea may be your first clue that a metabolic acidosis is afflicting your patient.
Brief Interventions
I just mentioned 3 things to avoid, but what can we do to help our intoxicated patients? You may not like the answer, but it turns out that talking to patients may be more useful than poking them with needles.
Austin Flint’s 1898 text discusses the therapeutic value of the doctor-patient relationship:
“The evils resulting from the abuse of alcohol, the dangers connected with the habitual use of opium… may be pointed out more effectively by a physician who has the confidence of his patients than by anyone else.”
Dr. Flint’s words are borne out in modern research. Investigators at Yale studied the impact of a “brief negotiation interview” in emergency department patients who screened positive for harmful drinking behavior.7 The structured interviews, which took a median of 9.7 minutes, were performed by emergency practitioners. Perhaps most significantly, the intervention group reported reducing driving after drinking >3 drinks by about a quarter, while the standard care group recorded no change. This single center’s experience is not unique. A large meta-analysis in the primary care setting showed that a patients who received a brief counseling intervention consumed less alcohol than minimal- or no- intervention patients after one year.8
If these results sound modest, remember that the intervention required 10 minutes of practitioner time and works better than CPR. The type of interview used in the Yale study, often called a “brief motivational interview,” usually involves asking a few structured questions that help the patient link his behavior to his emergency department visit and make a goal for changing future behavior. The magic trick behind motivational interview techniques is that patients do most of the talking, and patients are ultimately the ones who express the need for change. People respond best to revelations that they come to on their own.
Like any procedure, these techniques improve with study and practice. SAMHSA provides help obtaining training in Screening, Brief Intervention, Referral to Treatment programs (SBIRT) that use motivational interview principles to help patients become engaged in specialized treatment.
“It is undoubtedly true that persons will often listen more considerately to counsels or remonstrances from a medical adviser than to those of relatives, friends, and neighbors, or even to the admonitions of the clergy.”
Drawing blood may make the clinician feel better, but it won’t improve the patient’s outcome. Most of our alcohol intoxicated patients don’t need alcohol levels, IV fluids, or supplemental vitamins. We are trained to do things, but sometimes there is nothing we can do other than waiting and listening.
As in Austin Flint’s era, we must be willing to give patients a bit of our time. A new generation of research has shown us that if we are willing to just stand there and listen, even for just a few minutes, patients can tell us what they need to do to change.
Andrew – fantastic work, as always. We trained together and have similar attitudes but I’m curious as to which alcoholic patients to not give thiamine to? I don’t bother with the 19-year-old college student, or the tourist who had too many martinis. But any patient I see with chronic etoh abuse, I give 100 mg thiamine IM. I know IV better but popping in an IV on all these patients seems like overkill. Wernicke’s is so easy to miss, I feel like I’m better off focusing on prevention.
Interested to hear your thoughts.
Excellent point. I don’t disagree with the idea of giving the chronic alcoholic a big thiamine supplement from time to time. They only absorb 25-50% of what they eat and some don’t “eat” much more than alcohol.. I’m sure some of those Bellevue patients probably really benefited from thiamine! I’m not sure what the ideal dosing interval is, though.
Healthy adults are recommended to get 1-1.5 mg/thiamine per day, but we probably only need about 0.33 mg/ 1000 kcal consumed, so the rec is probably more than is needed for normal diet. The average sized person can store about 30 mg, Based on those figures, I suppose a person with a “full tank” of thiamine would last at least a month to 6 weeks without receiving any other thiamine in their diet, but this is unproven and doesn’t account for other mechanisms of thiamine loss that we don’t know about. So, the ideal dosing interval is definitely longer than a day and probably longer than a week, but two months seems too long.
Wernicke’s Encephalopathy is underdiagnosed, undertreated, and preventable in the first place. One study shows that on autopsy, over 12% of alcoholics have findings consistent with Wernicke’s, and over 80% of cases are not diagnosed during life.
Furthermore, the dosing of thiamine for treatment or prophylaxis is not well established, and the “standard” dose of 100 mg is not evidence based and is primarily historical. Some authors suggest a dose of 500 mg or higher for patients at risk (which, by the way, includes not just alcoholics but any population at risk of malnutrition: those with eating disorders, cancer, hyperemesis gravidarum, etc.)
Given that thiamine is so benign and the benefit of preventing severe neurologic disability is so clear, I find it difficult to argue that we need to be more strict rather than more liberal in deciding who we ought to prescribe it to.
https://twitter.com/JamesMaloyMD
E. Isenberg-Grzeda, H. E. Kutner and S. E. Nicolson, “Wernicke-Korsakoff-Syndrome: Under-Recognized and Under-Treated,” vol. 53, no. 6, 2012.
M. W. Donnino, J. Vega, J. Miller and M. Walsh, “Myths and Misconceptions of Wernicke’s Encephalopathy: What Every Emergency Physician Should Know,” vol. 50, no. 6, 2007.
Thank you for this interesting post. I am an ER RN and I work in both a large city ER and casually in a rural ER. In the city we do not tend to give IV fluid boluses to our intoxicated patients however at the rural site it is not uncommon to see an order for up to 2L of fluid per patient. I have long questioned the efficacy of this practice and am often told “it’s to flush them out” or “to clear the alcohol”. This post sheds some much needed wisdom on the matter. Thank you!