This is one of two related, and complementary posts as part of a special series about ethanol in the ER. The second part can be found here.
It was the end of the first World War. In a small corner of England, Sir Dr. Edward Mellanby (Doctor Sir?) performed one of the most important experiments in the history of modern medicine. Hyperbole aside, the results of this series of experiments would have far reaching implications for the as-of-yet not even conceived of fields of Medical Toxicology and Emergency Medicine. What did he do? What rigorous investigation of such import did the discoverer of Vitamin D1 perform that would change our modern practice?
He got four dogs2 drunk.
Actually, hammered, with serum ethanol concentrations ranging from 128-445 mg/dL. And after he gavaged them full of spirits, he observed them and took serial measurements of their ethanol concentrations. What he found impacts daily practice in the ER. Let’s explore how this almost century old data influences what we do today.
Consider this scenario. It is the day of a very popular music/dance/knitting/mahjong/art/culture festival and the Emergency Services van backs up with 100 presumed ethanol intoxicated patients. Some are reported to have fallen, others not. You don’t know which is which. You have one CT scanner. How do you determine if a patient is head injured, intoxicated, or both? What do you do? To help determine that, let’s go back to our drunk dogs.3The ‘X’-axis is time. The ‘Y’-axis is the serum ethanol concentration. As dogs (people) drink, the ETOH is absorbed, and the serum concentration rises. At some point, they stop drinking, absorption of the last drink occurs, and subsequently the concentration peaks. After the peak, metabolism and elimination now dominate absorption and the concentration starts to fall.
If you look at the curve, there are two points, separated by time, where the serum concentration is the same numerical value. One on the ascending arm of the concentration curve and one on the descending arm of the concentration curve. Mellanby observed an important distinction. Although the number was the same, the clinical appearance, or evaluation of the patient (dog), at those two points was vastly different.
Mellanby found that the dog was clinically more intoxicated on the ascending limb of the curve than on the descending limb. Now, it’s hard to tell if a dog is drunk. They don’t randomly launch into song. Instead, his observations recorded evidence of intoxication as “. . . scraping the toe-nails on the floor in walking, weakness in the hind legs as is evident on standing still, a rolling gait, stumbling, difficulty of getting on to feet after falling, and complete inability to walk . . . ”4,5
To understand the Mellanby effect we need to attempt to understand the concepts of intoxication and tolerance. Despite the fact that alcohol has been around and used by humans for well, forever, no specific receptor for ethanol has been identified, and the mechanism behind intoxication is still uncertain and debated.6 Measurement of intoxication is something that is very difficult to do, as this requires observation, not hard science. There have been a number of studies that have attempted to quantify or qualify this phenomenon. The difficult aspect of this is knowing what to measure and how to measure it. Research must involve more than a Judge Potter Stewart definition of alcohol intoxication (“I know it when I see it.”) Mellanby himself, in his report from 1919, addresses this limitation upfront, stating :
“. . . although there is an unquestionable relationship between alcohol in the blood and intoxication, it is not certain that the relation between the two is close enough to allow the latter to be strictly determined from the former. This fact would not be of much importance if there were as accurate a measure of intoxication as there is of alcohol in the blood, but this is not the case.”
One of the difficulties in studying alcohol is that it has different clinical effects at different blood concentrations. How does the same substance cause people to dance shirtless at one concentration and be profoundly unresponsive at another? While ethanol affects the whole brain, it appears to affect different brain regions at different concentrations. At low concentrations, ethanol is a selective CNS depressant, appearing to depresses areas in the frontal cortex. Areas that normally would remind you that pants are NOT optional, and where and when you can dance. This “cortical release” leads to animated behavior and a loss of restraint. This is often described as “paradoxical stimulation”, when it is in fact due to inhibiting inhibitory regions of the brain (i.e., disinhibition, or a kind of double negative . . . ). As the dose increases you see the effects of the inhibition on more brain regions. Patients go from initial excitation (e.g., dancing shirtless/pants-less), to somnolence, incoordination, emotional lability (e.g., “I Love You Man”), to unresponsive. These phases of intoxication are similar to those with general anesthetics. Further complicating intoxication research is that these signs are quite variable. Interestingly, subjective intoxication (e.g., “How drunk do you feel?”) is a measurement that seems to be relatively stable and reflects the experience of intoxication better than other measures, and is a variable that is often used in alcohol research.7
The acute effects of drinking also depend on the habits of the drinker. This is described by the development of tolerance, which has both a metabolic (pharmacokinetic) and a functional (pharmacodynamic) component.8 The kinetics of tolerance are not clearly understood, but studies have shown that the development of tolerance objectively is not just from repetition or practice.9–13
Tolerance can be defined as needing more of something in order to achieve the same effect. Looked at in another way, tolerance decreases the effect of a fixed amount of something, or reduces sensitivity to a particular stimulus. Alcohol tolerance is a diagnostic and clinical feature of alcoholism, and may lead to (or be a sign of) alcohol dependency. Studying tolerance is difficult, as sensitivity to alcohol has different mechanisms and tolerance varies in populations studied. Tolerance is often confused with addiction, which is a disorder of the brain’s reward system characterized by a stimulus that reinforces reward and the pursuit of repeated exposure, in spite of suffering adverse consequences. Acute tolerance (also called “within session tolerance”) is a decrease in response to alcohol within a single exposure, occurring independently of changes in concentration, or chronicity of use. Chronic tolerance is an acquired decrease in response to alcohol over multiple exposures. Acute and chronic tolerances are correlated and acute may be predictive of future chronic tolerance and alcohol-use disorders.14 The difference in observed intoxication reported by Mellanby, and supported by multiple subsequent investigations over the years (using multiple different measures), is thought to be explained by acute tolerance.15,16 An extremely detailed and well done meta-analysis of available literature supports the Mellanby effect for subjective intoxication (and willingness to drive), while also demonstrating that objective measures of skills (such as skills measured on driving simulators) were generally worse on the descending part of the curve.17 This is particularly dangerous and may be responsible for countless drunk driving cases. When you are hammered, 2 hours will not sober you up – no matter how “sober” you feel. . .
How does the Mellanby effect help guide our practice? What can we gather from this that will impact our management of our busload of inebriates?
A patient who is intoxicated can have a myriad of clinical signs and symptoms ranging from mild disinhibition to coma. Each patient in this regard should be evaluated with a thorough physical exam and an eye towards assessment for cryptic traumatic and metabolic disorders. The most important lesson that Mellanby teaches us is that clinical intoxication improves with time (again, presuming that the patient has stopped drinking). So serial examination should help us to define the patient that just needs more time versus the patient that needs more resources.
How do we assess our busload of patients?
- It starts with a detailed physical examination. Full set of vital signs, including rapid assessment of glucose. Get ‘em naked (often the most difficult and overlooked part of the entire process). Evaluate closely for signs of trauma (particularly the head). Any signs of trauma or abnormalities should be assessed or evaluated. You are establishing a baseline against which your subsequent exams will be measured.
- No trauma or abnormality found? Great. Watch the patient, and perform serial exams looking for objective signs of improvement. Repeat until clinically sober.
- Not improving after one or two exams? Pause and reconsider what could be missing? Is this someone who needs imaging or labwork?
Also, if you are considering getting an ethanol level, giving fluids, or vitamins then go no further and read this . . . But always remember that a few drunk hounds can change your practice.