How drunk is too drunk—another foray into medical ethics

The best ethical questions are real ones. Sure, it’s fun to play the lifeboat game, but when you’re dealing with flesh and blood human beings on a daily basis, games aren’t all that helpful. So here’s a non-life-and-death question: if a patient comes to see you and smells of alcohol, can you add an alcohol level to their blood work without specifically informing them?

Ethical discussions are best held as, well, discussions, so I’ll lay out some ethical principles and let you discuss before I weigh in further.

First, any patient who comes to see a doctor signs a “general consent for treatment” which usually contains a phrase such as:

I request and authorize Health Care Services by my physician, and his/her designees as may deem advisable. This may include routine diagnostic, radiology and laboratory procedures and medication administration.

Second, for your reference, here is the summary of the AMA’s code of medical ethics.

And finally, a brief list of the most agreed-upon basic principles of medical ethics:

    Beneficence – acting in the best interest of the patient.
    Non-maleficence – avoid harm to the patient
    Autonomy – the patient has the right to refuse or choose their treatment
    Justice—fair distribution or resources
    Truthfulness/informed consent

Remember that ethics aren’t a checklist. Real life situations are just that—real, with real people.

OK, the thread is now open.


  1. Hmmm, looks to be violating these three,

    Autonomy – the patient has the right to refuse or choose their treatment.

    Hard to refuse or consent if you aren’t asked.


    Rather undignified to be assumed a lush and have tests secretly ordered to prove it.

    Truthfulness/informed consent.

    That one is kind’a obvious isn’t it?

    Why not just ask the patient if he/she has been drinking and then ask them if you can test their blood alcohol for medical reasons.

    Uh, what exactly would those medical reasons be? I know that alcohol abuse has health risks but wouldn’t the patient have to have a specific condition, like diabetes or a liver problem, for you to have therapeutic reasons to find their exact blood alcohol level?

  2. If a BAL is necessary for patient care, I don’t see why you would have to ask permission beyond what you’d need for any other screening.

    For example, if you have an unresponsive patient who smells of alcohol, it’s important for the patient’s own good to determine whether they’re drunk enough to explain their altered level of consciousness, or whether the liquor just part of a more complicated situation.

    I don’t think it’s the doctor’s place to be ordering tests for the benefit of law enforcement. So, if someone comes in from a car accident and there’s no compelling medical rationale for doing a BAL, I’d ask permission, or skip it–unless an investigator with appropriate legal authority directed me to administer the test.

  3. Lance: the big thing here is that alcohol can screw with a lot of tests and treatments. It can create false positives or negatives: you can’t count on dizziness as an indicator of inner ear problems, blurry or double vision as an indicator of eye problems, excessive reaction to light or sound as… whatever causes those, and lack of pain as lack of damage. You can’t give the patient a great many drugs that are interfered with by alcohol – that includes pretty much everything that says “don’t drink while taking this drug”, and there’s a lot of those. There’s also a difference between using Listerine or rubbing alcohol and drinking too much – the former two don’t do that stuff, but certainly can make a person smell of drink.

  4. D. C. Sessions

    Why is the patient seeing you?

    If it’s related to a liver complaint, the answer to your question is rather different than if he’s seeing you about an arthritic knee, and different again from if she thinks she might be pregnant.

    Questions are always more interesting than answers.

  5. CaladanGuard

    I generally wouldn’t have a problem with it. (I’m Australian, not American, for reference) Again, depending on the condition that the patient arrives to be treated for. But anything major, especially trauma or anything that requires invasive treatment it would nearly be unthinkable not to test BAL.

    As far as I am aware all patients admitted from vehicle accidents are routinely given BAL tests as well. It’s required for the Police to include in their reports.

  6. I left this thread open on purpose, and I’ll weigh in later with some thoughts, but I do want to point out some important points raised by you so far…

    The context I left intentionally vague in order to examine the general principles, and the situations you’ve brought up are excellent.

    Remember, Lance my friend, that it’s not a checklist. The patient did sign a general consent. I certainly wouldn’t ask the patient if they mind if their CBC includes a mean corpuscular volume.

  7. Remember, Lance my friend, that it’s not a checklist. The patient did sign a general consent. I certainly wouldn’t ask the patient if they mind if their CBC includes a mean corpuscular volume.

    (Glad we’re friends.) You were the one that presented the “basic principles of medical ethics”. Even if it isn’t a checklist I assume it was meant to give some guidance.

    As you say each situation is unique and so long as you were going to keep the information confidential between you and your patient I suppose their really is no harm in a blood alcohol test. Although I’m not sure why you couldn’t at least tell them you were testing them.

    My doctor, with whom I have an excellent long term relationship by the way, knows that I want to be very involved in my medical care and that I expect that any tests or treatments will be explained to me before hand.

    If I happened to come to an appointment after knocking back a few drinks I would be a bit miffed if I found out later that he had ordered a BAL test without at least telling me.

    Then again I’m fairly confident that he wouldn’t do any such thing so I guess it comes down to individual doctor-patient relationships.

  8. D. C. Sessions

    By the way, what’s this “without telling me” part?

    I’m halfway literate, and besides I live with a med tech. Those lab orders are, after all, quite readable to those who want to read them — and can’t legally be hidden from the patient even when said patient isn’t carrying them in hand.

  9. Whoa, hold on. I hope those aren’t doctors in the crowd, saying it’s perfectly okay to give someone a form of examination against their will, especially if you are going to share that info.

    If the patient wants it, asking doesn’t do any harm (well I suppose it’s inconvenient)
    If they don’t, well, that’s privacy for you. You can try to persuade by explaining, but if they understand why it would be in their interest, and they decide it’s not worth it, Hands Off, pal.

    This giving someone a blood test without even their knowledge, then handing over their medical records to the police is especially nasty, and I feverently hope it’s confined to BAL for which I suppose there is a pretty strong utilitarian argument.

    I’ve had some significant problems with the police arising from because they knew a bit about my medical state. It’s a lot like confirmation bias, as soon there is something medicalish about you, you’re that much less of a person, in they eyes of a dumb cop. I’ve seen first hand that privacy does matter, bigtime, even if you haven’t done any thing wrong.

  10. In addition to the above-quoted general consent for treatment is the government-mandated privacy policy, which generally doesn’t allow sharing results outside of clinical situations. If police want to see my patient’s record, they need a subpoena.

    That being said, the data point now exists, privacy or no…

    That being said, what makes an EtOH level different (if it is) from any other tests i might routinely run, such as a TSH or CBC?

  11. D. C. Sessions

    That being said, what makes an EtOH level different (if it is) from any other tests i might routinely run, such as a TSH or CBC?

    You mean, besides it being more directly under the patient’s control?

  12. I have a minor concern about the patient being billed for a questionable procedure.

    I am more concerned about what purpose is served. You have not suggested how alcohol is significant to diagnosis or treatment in the hypothetical case.

  13. Well, so long as you genuinely think they won’t mind, I suppose it’s the same as any other test, then. But lots of people don’t want their family or someone to know they’ve been drinking, for reasons that are their own business.

    If the privacy safeguards work, I suppose the test is fine whether or not you mind other people knowing, but I was always under the impression that the family tends to get information like that….. secretaries bending the rules, things like that.
    Doing it for law enforcement is exceptionally questionable, though. Suppose it’s a car accident, and both parties were intoxicated – one gets their alcohol tested without their permission, and the other doesn’t. It goes to court, and now one party is at an unfair disadvantage, the other party can claim they were sober. Where’s your justice now?

  14. If blood alcohol level would affect treatment then I think it’s reasonable. I know there are a number of medications that interact with alcohol and if the complaint is something that might be treated with on of those drugs it’s very important to know it.

    I’m not a doctor so my hypothetical case may be sketchy it’s going only on my knowledge as an interested layman. Say a patient comes in with a blood clot, the usual treatment is an anticoagulant. Alcohol already thins the blood, so giving an anticoagulant to someone with a few drinks in them that medication could cause hemorrhaging.

  15. CaladanGuard

    Here in Australia, we have a legal responsibility under various laws to test any and all passengers, drivers, people who were walking along the street and got hit by said cars or bikes or skateboards -and- present to our hospital for treatment of injuries sustained in aforementioned incident for blood alcohol, and in some states now full drug screens and deliver our reports to the state police.

    Whether the patient consents or not, it’s my ass on the legal line if I don’t do the tests. So they get their blood taken under sedation, or a large orderly if necessary.

    In private practice, or non motor vehicle related incidents, the lines get grayer, but I’ll stick with the one case for now.

  16. LanceR, JSG

    In regards to the efficacy of the current privacy measures, there is a joke making the rounds of nurses:

    Knock knock!
    Who’s there?
    HIPAA who?
    I can’t tell you that.

    My mom loves that one! (Old school nurse. You have to know one to understand.)

  17. llewelly

    So, if someone comes in from a car accident and there’s no compelling medical rationale for doing a BAL, I’d ask permission, or skip it–unless an investigator with appropriate legal authority directed me to administer the test.

    It would be much, much better for it to be illegal for law enforcement to access medical tests not specifically made for a law enforcement purpose.

  18. One more little wrench/spanner in the works…

    What if I suspect the patient is intoxicated, but they leave my office and drive into a kindergarten?

  19. D. C. Sessions

    Knock knock!
    Who’s there?
    HIPAA who?
    I can’t tell you that.

    Ain’t that the truth. I can’t tell you how many people I’ve packed up and shipped off to various hospitals and when I’ve tried to follow up (as in, “tell me if I screwed up, Doc”) been told that HIPAA prohibits them from telling me if the patient lived or died.

    In one case where the splinting was particularly challenging, all I wanted to know was whether the splint had been adequate. Fat chance. I found out later purely by chance (I rode the lift with the kid) and he still has the use of his hand. He never knew we were sweating that.

    Funny thing is, I actually got better followup this week when we stopped to help at a nasty roadside accident. DPS called to tell us how that one turned out (and notice that DPS isn’t a care provider — maybe that’s how they got around the rules.)

  20. When I read the header I thought the hypothetcal doctor was drunk, not the hypothetical patient.

    I think that worries me far more. And honestly, I don’t think I’ve ever had a doctor explain exactly why he orders the blood tests he does. I’m assuming he is finding out stuff he thinks is important to know. So unless his only motivation is to gossip about my drinking habits, why would I complain?

    I have always found informed consent to be a very fuzzy thing. An expert explains possibilities to a non-expert, who may be suffering from pain, fatigue, dizziness etc etc.

    And says – make a choice. So when I am too sick to decide what channel to watch I need to decide which test and treatment to take. Am I the only one who finds this odd?

  21. HIPAA is widely misunderstood. The basic guidelines are rather simple.

    But back to the ethics of this case.

    Some questions to ask:

    1)Is EtOH level different from other tests? If so, how?

    2)What is the purpose of the test?

    3)Cui bono?

    4)Is deception involved, and if so, did it serve a purpose congruent with the patient’s needs?

  22. So when I am too sick to decide what channel to watch I need to decide which test and treatment to take. Am I the only one who finds this odd?


  23. CaladanGuard

    From my opinion, on the general question, in PalMD’s numbered list form answers:

    1) I don’t believe it’s different from any other test. In a normally presented case with no legal aspects to it.

    2) To assist with my diagnosis and treatment of the patient. To eliminate any symptoms that may be related to the alcohol and ensure that their are no conflicts with immediate treatments with Ethanol as a contraindication.

    3) Everything we do is to the patients benefit. Financially, adding the cost of the test would hardly be viable either. But in my case, I’m salaried, it certainly doesn’t benefit me.

    4) Maybe. I don’t detail every test I do to a patient, but do they really need to know? In a normal case, I’m not handing the results out to anyone, I’m not doing it so I can judge their wicked ways (I enjoy a drink or…too many) I’m doing it as a step to aid in the diagnosis and treatment.

    Though they may kick up a fuss about the test due to negative connotations surrounding the issue I don’t see that we should treat it any differently than testing WBC, MCH, or Creatinine.

  24. Depending on how drunk a patient is, is he even capable of giving imformed consent? I would say no if very drunk. He might not even remember what happened.

  25. Dr. Kate

    I’m going to echo several posters here and go with this:

    If some aspect of your treatment, diagnosis, etc., could be affected by the person having had a drink (or several), I think you are ethically obligated to do that test, whether the patient consents or not (although I also don’t see any harm in asking the person if they’ve had a drink). If you don’t, you may compromise your ability to benefit (or not harm) the patient.

    If you’re in the (rare) situation in which the person’s alcohol status could not possibly have a bearing on your diagnosis or treatment, then I think you should not order a BAL test any more than you should order any other unnecessary test. Chances are the patient will have to pay for those tests, so you should order only the ones that have diagnostic value.

    If you’re in the US, you should under no circumstances give any test results of any kind to any law enforcement officer (or anyone else) without the patient’s written consent (except in the cases legally recognized as having exemption).

  26. The patient did sign a general consent. I certainly wouldn’t ask the patient if they mind if their CBC includes a mean corpuscular volume.

    You dare measure patients’ MCVs without their explicit consent! I’m shocked! And appalled! How dare you suggest that their corpuscular volume might be abnormal? Hmm…doesn’t play well, does it?

    The main difference, IMHO, is that society places a value judgement on whether and how much alcohol a person drinks that it does not place on a person’s MCV. If we lived in a society where, say, there was severe prejudice against people with thalessemia. Then measuring a patient’s MCV has a certain level of “threat” that it doesn’t in current society: you might find out that they have thalessemia and treat them differently. Similarly, patients who present after drinking might reasonably fear that their doctors may treat them differently if they knew their BAL.

    The question then becomes, “Is there enough danger to a patient from his or her BAL being measured to justify requiring specific permission?” Currently, the legal answer appears to be no: no specific permission is legally required as it is for testing for HIV or genetic testing. Should it be ethically? Er…I don’t know that.

    (Incidentally, I’m assuming that the question is about testing without explicit permission, NOT sneaking the test in against the explicit or implicit refusal by the patient. That would be clearly unethical.)

  27. OK, then let’s make a slightly more specific scenario, which many doctors would recognize.

    Let’s say a patient comes to see you. You know they have a history of drinking. They tell you they’ve quit. They don’t appear visibly impaired. You’ve drawn blood for other reasons (cholesterol, etc.).

    After the patient leaves, a staff member says that the patient smelled a little soused. At that point, would it be ethical to add an ethanol level to the labs?

  28. After the patient leaves, a staff member says that the patient smelled a little soused. At that point, would it be ethical to add an ethanol level to the labs?

    In this scenario, do you have reason to believe that the patient would object to you checking an EtOH level? Suppose you imagine calling the patient at home and asking. Would his/her response be, “Yeah, sure, if you think I need it” or “WHAT? NO!” followed by a 30 minute rant about you, doctors in general, and the world? What will you do if the test is positive? Will the conversation you need to have with the patient suffer because you get side tracked by the issue of whether you should have ever drawn the level? Will the patient be relieved to get it into the open and be able to get help? (not that you can know the answers to these questions ahead of time, but based on your knowledge of the patient, what seems most probable?)

    (In case you’re wondering, the short answer I’m giving is “I don’t know.”)

  29. Layperson weighing in here. In almost every instance where a doctor has ordered a blood test of any kind, that’s exactly how much information I was given; “I’m going to order some blood tests” is about all I’ve heard, except in very specific circumstances. And I’ve never been asked if I wanted the tests done, I’ve been handed a prescription for the lab work and given a map of local labs that can do the tests.

    To the comment that the orders for the tests are readable, not everyone has the motivation to try to decrypt the various codes and notations used to on the prescriptions and then research what they mean. I’m a computer programmer, so I have to keep my own set of jargon stored in my head. I don’t have the time or energy to go learn a whole new one 😉

    As far as whether or not it’s ethical, unless you were doing it for totally nefarious purposes (in which case you wouldn’t give a crap if it was ethical or not), the patient did sign the general consent form. I’d have to read that as giving permission for the test.

  30. Michelle Schatzman

    What about this solution: the doctor informs the patient that his blood alcohol level will be tested, because the diagnostic and the treatment depend on this data. If he refuses, let him sign a form stating his refusal (I am thinking of the guy who could drive into a kindergarten).

    If he seems too drunk to answer intelligibly or to ansswer at all, the test must absolutely be done: I have learnt that some diseases cause symptoms analogous to drunkenness, and therefore, it is essential to know.

    What would happen to the therapeutic relation if the patient is well known to the doctor, and the doctor expects this relation to continue? Can the doctor explain quietly the strictly medical and immediate issues, in order to convince the patient? How much time does a doctor have to convince a reluctant patient?

    The last case : adding the BAL test to the list as an afterthought. I know little about legal issues for US doctors. My guess would be that the doctor should not refrain from prescribing a test if he thinks that the result of this test could change the diagnosis and/or the treatment. He should also tell the patient, explain its rationale and state that the permission has been given in the general consent form.

    Common sense suggests that the doctor would be liable if he refrained from prescribing a BAL in the above circumstances. But is law always consistent with common sense?

    Regarding legibility of prescription of biological analyses : if one is subjected to blood tests on a regular basis, it is definitely a good idea to ask what they mean and why they are prescribed. In particular, it can help the patient to catch an error, and to understand how difficult a medical decision can be. Good for the patient and good for the doctor.

    I sound very brutal… I’m glad that I am only a doctor of… mathematics.

  31. I can’t stop thinking of the David Rosenbaum case.

    Synopsis: Guy has a couple of drinks with dinner, decides to go for a walk to the corner store. He gets mugged and the mugger whacks him on the head, rendering him unconscious and with a severe brain injury. Ambulance picks him up and, smelling the alcohol, assumes he is inebriated, dilly-dallies around. He does not get treated for said head injury and eventually dies in the hospital, because all the while the emergency responders and the ER staff assumed that his symptoms were the direct result of alcohol.

    I guess my question would be, YOU examined him, did YOU think he was drunk? If you did, why not just ask directly if he’s had a beer or three? If you didn’t, why do you think your staff member might know better than you?

  32. Excellent questions. As a random fact, I’m practically anosmic for about half the year, so unless someone is completely trashed, I’m unlikely to smell it, and if they’re a regular drunk, they may have very good behavioral compensation and be able to hide their inebriation.

    I’ll give you some more thoughts shortly.

  33. That being said, what makes an EtOH level different (if it is) from any other tests i might routinely run, such as a TSH or CBC?

    Other than it not really being a “routinely run” test? I think one would easily argue that since the test is not routine (like a CBC), they should be specifically informed of it being requested/drawn/analyzed.

  34. OK, pretty much all the points raised above are good ones, and this is how we approach ethical problems in medicine—more of the time than not, explicit discussions of ethics aren’t necessary, but when they are, there is seldom one correct answer, and the discussion helps guide your future decisions.

    –If I have a patient who I am explicitly treating for substance abuse, they already know that I may test them at any time. It’s usually an explicit agreement, so the answer in that case is clear cut.

    –If I suspect my patient has been drinking but is not an immediate threat to others, the approach that most closely respects a patient’s dignity is to confront them directly. However, the benefit to the patient may preclude this.

    –If I confront the patient, they tell me they haven’t been drinking, and I think they are lying, the therapeutic relationship is in trouble. If I confront them about their lie in the wrong way, they won’t trust me. If they are clearly lying, I won’t trust them. A BAL might actually increase trust. Unless of course it’s seen as going behind their back.

    Substance abuse is a very tough topic, but all doctors will confront it. In general, a straightforward approach is necessary.

    I do think a BAL is qualitatively different from other tests, as the results have a social impact different from, say, a hemaglobin. That being said, I think it is sometimes ethical and sometimes not to add one on, depending on the patient and your relationship with them.

    I hope that is sufficiently muddy for now.

  35. EtOH level test is different from other tests because you have reason to believe the patient might not want you to know his or her blood alcohol levels.

    The same would be true if you had some reason to think the patient wouldn’t want you to run another test, even if it is normally routine. If a patient adamantly didn’t want to find out if they had some condition—say cancer, or a genetic disease—then testing them for it without consent seems rather dodgy.

  36. PalMD wrote:

    One more little wrench/spanner in the works…

    What if I suspect the patient is intoxicated, but they leave my office and drive into a kindergarten?

    If you don’t think your patient is fit to drive, you should tell them so directly.

    I bet doctors frequently encounter patients who think they’re okay to drive when they’re not–whether they’re sick, drunk, emotionally upset, or still woozy from anesthetic.

    It’s the kind of thing that should be dealt with in a forthright, non-judgmental way.

    A BAL probably wouldn’t help anyway. The patient would be long gone before you got the results, wouldn’t they?

  37. If your relationship with your patients is such that you feel the need to run tests instead of talking to them, maybe you need to work on your people skills.

Leave a Reply

Your email address will not be published. Required fields are marked *