Articles

Followup: Benedetti on Placebo Ethics

A few months ago I wrote about Fabrizio Benedetti’s research on the neurobiology of the placebo response, and a discussion about placebos and ethics ensued in the comments. Now Dr. Benedetti has written about that issue in a “Perspective” article in the journal World Psychiatry, “The placebo response: science versus ethics and the vulnerability of the patient.” 

We have learned that verbal suggestions can activate neurotransmitters and modulate pain perceptions, and positive expectations can activate endogenous opioid and cannabinoid systems. A complex mental activity has objective effects on body physiology. Words and drugs can activate the same mechanisms. Drugs are less effective without therapeutic rituals. We are delving deep into human foibles and vulnerable traits at the center of human interactions. What implications do these insights into mind-body interactions have for patient care?

Things used to be simpler. Placebos are inert by definition. Prescribing an inert treatment constituted lying to a patient, implying that it would have a specific effect when the provider knew it would not. It might fool the patient into thinking he felt better, but it would not actually improve the course of illness, and it involved deception. Prescribing placebos was unethical, period. 

But now we are learning that while the placebo itself is inert, the act of giving a placebo is not: it can produce actual physiological effects through suggestion and expectation. So is it legitimate to prescribe sugar pills or lie to patients if we know it can produce objective health benefits? What was once black and white has now become a gray area.

Another question arises: how can we apply this knowledge to mainstream medicine? When we prescribe drugs, should we be using rituals and words to ensure that they are maximally effective? What limitations or guidelines might govern that practice?

Benedetti says

If a syringe filled with distilled water and handled by a doctor may induce expectations of benefit, then the same expectations can be induced by talismans, mascots and bizarre rituals carried out by quacks and shamans…Deception is at the very heart of a placebo procedure, and indeed it makes no difference if this deception comes from a doctor or a quack or a shaman.

New knowledge about neurophysiology is being used to justify using any procedure that increases expectations and beliefs, no matter where it comes from. This is worrisome, to say the least. CAM and quackery are trying to exploit this new knowledge in ways that tend to throw medicine back to its pre-scientific days. Acupuncturists who have acknowledged that their procedures are no better than placebo have argued that we ought to use them anyway, since placebos really do work to make patients feel better. Manufacturers of bogus health products have argued in court that they know their products can only have placebo effects but placebo effects help people and they have to lie in their advertising in order to elicit those placebo benefits. What if this line of reasoning were carried over to pharmaceutical research? If a new drug fails in clinical trials, might the manufacturer try to justify marketing it anyway for the placebo effect?

My personal opinion is that lying to patients or misrepresenting the state of the evidence in any way is always unethical and must be scrupulously avoided. Telling the patient that what you are prescribing is a placebo wouldn’t excuse you, because the very fact that you are giving it has persuasive meaning for the patient. I think attempts to elicit a placebo effect should be only used in conjunction with an effective treatment. Words should be used carefully, and the focus should be on general measures that bolster the doctor/patient relationship and enhance the patient’s trust, like spending more time with the patient and showing a greater interest and sympathy.

We face two major challenges. First, we need to improve the ways in which we communicate the results of placebo research so the public won’t think we are willy-nilly recommending placebo deceptions of any kind in clinical practice. Second, we need to answer some difficult questions about the ethical limits of taking various actions to increase expectations. The answers will not be easy. Let the discussion begin.

Posted in: Medical Ethics

Leave a Comment (91) ↓

91 thoughts on “Followup: Benedetti on Placebo Ethics

  1. kathy says:

    “Let the discussion begin”. Indeed! Medicine isn’t the only field where outside expectations affect people. In sport one can see the effect calm confidence from the “authority figures” – captain and coach – can have on a team’s performance. Likewise the expectations of teachers and parents can affect a child’s performance in schoolwork (I apologise that I have no references for these statements).

    However, if the “healer” knew the medicine/treatment didn’t have any objective effect, could they project that confidence the patient needs to feed their own belief? I suppose if the patient was desperate enough for relief they would believe even someone who was unconvinced and unconvincing …?

  2. wertys says:

    We have Prof Benedetti and Prof Ernst on a panel discussing this exact question next year at a conference I’m organizing ! Unfortunately I’ll have to get back to you all after that happens in May next year!

    At present though, I have to say that if you have a patient (eg an elite athlete) who willingly and with fully informed consent agrees to use their innate placebo response to ensure maximal performance, I can’t see how that’s unethical. If the patient/consumer/end user has misconceptions or false beliefs about the placebo response you are training into them then it is unethical because you are infringing their autonomy.

    While this will not necessarily be applicable in an objective fashion to every patient you have in front of you, as a rule of thumb it seems as effective and rational as any in clinical medicine.

  3. Jimmylegs says:

    The first point wertys makes me wonder, has any studies been done on placebos in which the patient knows full well it is a placebo or inert pill. But I still believe that it is unethical still, to give a patient something that will not work and the patient knows it will not work in order to get some benefit. Why bother give anything at all?

    My personal stance on giving a patient a placebo or treatment known to be a placebo at best (CAM) is unethical, it is not even a gray area for me.

  4. BillyJoe says:

    The placebo should be used only in conjunction with an effective treatment.
    It is also unethical to exaggerate the effectiveness of the treatment in order to enhance the placebo effect.
    On the other hand, ethical ways to enhance the placebo effect include the way the doctor approaches and interacts with his patient, showing genuine interest, concern and empathy.

    In other words, I’m pretty well in agreement with Harriet

  5. phayes says:

    @Jimmylegs
    IIRC, I did find one such study – and not one concerning a conditioned response – once. It might have been via the (long) article/paper by Adam Kolber mentioned here: http://mindhacks.com/2008/05/27/placebo-is-not-what-you-think/

  6. windriven says:

    Are the ethics of using a placebo different for a patient for whom there is no cure? Some medical treatment is entirely palliative. Placebos sometimes have demonstrable palliative effects.

  7. daedalus2u says:

    wertys, I would say that any elite athlete is already able to modulate their “placebo effect”. That is probably the main skill required to be an elite athlete.

    I see the “placebo effect” as a manifestation of the control system that physiology uses to allocate resources, mostly between immediate consumption and doing long term maintenance. Evolution has configured organisms to minimize substrate “wasted” on not generating descendants.

    Substrate wasted on generating a big phenotype when it isn’t needed is substrate that isn’t used to generate descendants. Substrate wasted in keeping a phenotype in better repair than is needed is substrate that isn’t used to generate descendants. Likely much of the regulation of the allocation of substrates evolved when eukaryotes (or pre-eukaryotes) were single celled, but our physiology has been “locked” into that mode because that was there first, it worked, was the major driving force in evolution at the time, and no sufficiently superior allocation scheme has evolved later (except for things like altruism toward conspecifics, offspring and so on) to displace it.

    My hypothesis is that physiology allocates resources according to an allocation scheme which is only partially under voluntary control. There has to be involvement of the CNS to trigger things such as the “fight or flight” state so as to escape from a bear. During the fight or flight state, physiology diverts resources from involuntary maintenance to voluntary use by muscle. In the limit, you can run yourself to death trying to escape from a bear. Physiology has allocated resources to escape that were required for maintaining viability. There weren’t enough to do both, being caught by a bear is certain death, running until you drop dead from exhaustion is “the same” from an evolutionary standpoint, so organisms have evolved to be able to run themselves to death. It isn’t that dropping dead from exhaustion is a feature, but that organisms that had the capacity to do so had more descendants than organisms that were unable to do so, and so all extant organisms are descended from organisms that could.

    Allocation of substrates to maximize descendants is what physiology has evolved to do (and all it has evolved to do). Many of the control systems that physiology invokes are artifacts of a particular time and place in that organisms’ evolutionary history. They are there to cope with the apocryphal “running from a bear”, which very few people have ever experienced, but which all of us could do in a heartbeat.

    The CNS is necessary to turn on the “fight or flight” state (when the retina shows an image of a bear chasing you), and also to turn it off (when the CNS integrates sufficient signals to feel “safe”).

    I completely agree with Dr Hall, that lying to patients is always unethical and no clinician should ever lie to a patient about anything, even (or perhaps especially) if the patient wants you to lie to them. Facing reality is always better than wishful thinking.

  8. daedalus2u says:

    windriven, the palliative effects of placebo pale in comparison with the palliative effects of drugs like morphine.

  9. windriven says:

    @daedalus2u

    I’ve used that argument many times. But this was a philosophical question, not one of the relative efficacies of dilaudid and acupuncture.

  10. qetzal says:

    Suppose a doc told a patient:

    “Look. I don’t have any actual drug that can help you. But I can give you these placebo pills. They don’t contain any active drug. But research has shown that if you tell yourself that they’re going to help, you may actually feel better when you take them. The pills themselves won’t make you feel better. They’ll just be a ‘prop’ to help you make yourself feel better.”

    I don’t know if the placebo effect would still ‘work’ in patients who were told something like that. But it wouldn’t surprise me if it did, at least for some people and for certain conditions (esp. pain). And I think it would be straightforward to test in a clinical trial.

    If it could be shown that that worked, I don’t see any ethical problems. The doc is being completely honest, and the patient still gets some potential benefit.

    Thoughts?

  11. daedalus2u says:

    qetzal, I would say no, I consider the scenario you suggest to be unethical.

    Depending on what the condition is, there are positive things the doctor can suggest that will likely help. A better diet, stress relief, moderate exercise, sufficient sleep, follow-up in a certain time, and then rely on a positive bedside manner to elicit placebo effects.

    If someone is in such a state that a better diet, stress relief, moderate exercise, sufficient sleep and a positive bedside manner won’t help them, then very likely placebos won’t help them either.

    I think you would have a hard time showing clinically significant superior results from a clinical trial comparing offering a placebo as you suggest and lifestyle counseling as I suggest. It is already known that those lifestyle changes will help just about everyone with just about every problem, so there is no justification for not providing such treatments now, without waiting for a clinical trial to show which approach is superior.

  12. fledarmus1 says:

    Lying is bad, placebo effect is good (at least in a clinical setting. It drives the researchers nuts though…). Is there any evidence that ONLY lying to the patient can give the placebo effect?

    What I mean is that, underlying the placebo effect is a real pathway that has measurable effects which are related to the patients’ beliefs about the effectiveness of the treatment they are receiving. Why not study that pathway and learn to activate it without lying to the patient? It might be as simple (in concept anyway) as being perfectly honest with the patient – rather than sticking him full of needles, sugar pills, or water with fancy names, convince him or her that he or she will indeed get better because it is true that a certain percentage of those who believe they will get better, do. In practice, this may resemble psychotherapy more than prescription medicine, but if you can help a patient by making him think that if he takes the woo he will get better, why not do it just by convincing him that there is a biochemical pathway he can activate by his belief in his health, and if he can activate it, he will get better?

    The other possibility, of course, is to find a way to chemically intervene in whatever pathway the placebo effect works through. If that worked, then manipulation of the placebo pathway should be more effective than attempting to trick a person into accessing it by themselves. In cases where drugs to treat the underlying condition are either unknown or ineffective, drugs to activate the placebo effect may be more predictable than lies and placebos to get the patient to self-activate the placebo effect.

    I have seen somewhere (not at my fingertips) research suggesting that certain identifiable groups of people are more susceptible to placebo effects than others. If those groups are truly identifiable (without hindsight bias) and those placebo effects are indeed effects and not simply cases of self-reported “feeling better”, learning biochemically why one population is affected and another is not may lead to ways of stimulating the placebo effect in otherwise unmoved populations. All without lying about what you are giving them.

  13. qetzal says:

    daedalus2u,

    If you’re right – that other interventions would always work better for any relevant situation – then I agree with you. I’m not convinced you’re right, though.

    Getting people to make lifestyle changes is notoriously difficult. Just because they’re counseled to do so doesn’t mean they will. We see this same issue with cardiovascular disease. Sure, patients should be told to exercise more, lose weight, and eat well. But if that’s the best intervention, does that mean prescribing statins is unethical? I don’t think so.

    Besides, suppose a patient needs to reduce stress. The doc tells them to take a sugar pill and imagine that the simple act of taking will help them relax. IF there were data showing that worked, why would that be unethical? Especially since it wouldn’t preclude any other stress-reducing intervention that might be implemented.

    I also don’t think it’s fair to require that placebos should show clinical superiority over other interventions. That’s a very high standard, and not one that standard drugs are typically required to meet (cf the example of statins above).

  14. Scott says:

    Count me as also in agreement with Dr. Hall and BillyJoe. When providing real medicine, by all means provide it in such a manner as to maximize nonspecific benefits (short of deceit).

    If there were convincing evidence that nonspecific benefits cannot be maximized without actual placebo treatments, I would potentially start to shift my position.

  15. Todd W. says:

    @Jimmylegs

    has any studies been done on placebos in which the patient knows full well it is a placebo or inert pill.

    Well, sort of. There was a study by Dr. Ted Kaptchuk that purported to examine the effects of placebo without lying to subjects. Dr. Gorski discussed it here. In short, subjects with IBS were told they were going to get a placebo or no treatment. However, the study including a bit of deception insofar as priming subjects to expect a response from the placebo!

    Patients who gave informed consent and fulfilled the inclusion and exclusion criteria were randomized into two groups: 1) placebo pill twice daily or 2) no-treatment. Before randomization and during the screening, the placebo pills were truthfully described as inert or inactive pills, like sugar pills, without any medication in it. Additionally, patients were told that “placebo pills, something like sugar pills, have been shown in rigorous clinical testing to produce significant mind-body self-healing processes.” The patient-provider relationship and contact time was similar in both groups. Study visits occurred at baseline (Day 1), midpoint (Day 11) and completion (Day 21). Assessment questionnaires were completed by patients with the assistance of a blinded assessor at study visits. [emphasis mine]

    So, while subjects were truthfully told that they would receive either no treatment or placebo, they were still being misled and primed to expect a response. Kaptchuk and his team should have left out that bolded bit in describing placebos to subjects.

  16. Todd W. says:

    Now, on topic to Dr. Hall’s post, although I’m incredibly wary about using placebo to treat, I’m not 100% in my objection to its use.

    To play devil’s advocate, if a subject’s symptoms are entirely subjective, use of real medicines is likely to have little to no effect, other than to generate adverse reactions. Use of real drugs, in such a case, would be unethical, since you are putting the patient at risk without any real, objective benefit. Again, though I’m loathe to condone lying to patients, in this scenario, is there a good argument to made for the use of deception in some degree (e.g., not necessarily outright lying, but perhaps misdirection or omission)?

    Part of the problem with the placebo response is that there’s no consistent way to elicit it. At least at present, there is no known means of knowing beforehand what techniques will cause a placebo response in any given patient.

  17. daedalus2u says:

    fledarmus1, invoking the physiology of the placebo effect pharmacologically is exactly what I am proposing to do.

    http://daedalus2u.blogspot.com/2007/04/placebo-and-nocebo-effects.html

    qetzal your assertion that:

    “I also don’t think it’s fair to require that placebos should show clinical superiority over other interventions. That’s a very high standard, and not one that standard drugs are typically required to meet (cf the example of statins above).”

    is not correct. The Declaration of Helsinki requires that all clinical trials must have clinical equipoise, that is, one of the legs of the trial cannot be known to be clinically inferior to the other leg. One leg has to be “standard of care” treatment. The new treatment is compared against “standard of care”, and if the new treatment is better, it becomes the new standard (after sufficient replication and stuff), if the new treatment is inferior it is abandoned (unless it is CAM, in which case they call for more trials).

  18. mousethatroared says:

    Just wanted to quickly comment that I enjoyed the article, Harriet Hall. I raises a lot of intriguing questions… about ethics, risk/benefit and the possibility of unexpected consequences when changing a process.

    I’m going to let those questions rattle around to see if I have a real opinion about whether this new information should change the approach to placebos, but I do have one observation.

    Everytime I read an anecdote about placebo response…such as one that (I think) WLU relied the other day, where his headache disappears within minutes of taking a Tylenol, I think, DAMN, I want me one of those powerful placebo responses.

    I don’t think I’ve ever had an immediate response to an oral medication. In fact it seems like everytime the doctor prescribes something, it doesn’t work as well or as fast as they expected.* I wonder if those of us with rather negative outlooks on life are placebo resistant.

    If that is the case, I have no idea how one one cope with the variety of patient placebo response.

    *Well, or at least as fast as they lead me to believe I could expect:)

  19. Janet Camp says:

    What does it mean for the patient to “feel better” anyway? Feel better psychologically? Does this effect last?

    Many years ago I asked my gyn if it would be okay to try some herbal concoction for PMS. He said “go ahead, but even if it gives you a placebo effect, it will wear off and then you’ll be back to square one”. I didn’t bother, and it was the beginning of my questioning the woo.

    Can a non-believer get a placebo effect? Is the effect of the meds I take less effective because I don’t get any “placebo enhancement”?

    I guess I would vote in favor of letting docs use this info on a case-by-case basis. There may be some cases with already confirmed wooies where it would be beneficial and help build trust in the doctor. The biggest danger is probably that any placebo effect could end up masking real problems and keep the person from coming back to see the doc.

  20. trrll says:

    So, while subjects were truthfully told that they would receive either no treatment or placebo, they were still being misled and primed to expect a response. Kaptchuk and his team should have left out that bolded bit in describing placebos to subjects.

    I’ve never bought Dr. Gorski’s objection to the study, because the patients can only be said to have been misled if the placebo failed to work. But it did, so the worst that can be said is that the evidence for a true statement might have been overstated. Moreover, while one can quibble about the exact meaning of “rigorous” or “self-healing,” and whether they are the most appropriate adjectives to use, there is certainly evidence in the literature to support that position (although clearly not adequate to convince Dr. Gorski).

    I also cannot agree with Dr. Hall’s argument that ‘Telling the patient that what you are prescribing is a placebo wouldn’t excuse you, because the very fact that you are giving it has persuasive meaning for the patient,” because I don’t see anything inherently wrong with persuading a patient to feel better. Indeed, if that is unethical, one would have to conclude that the entire field of psychotherapy is in violation of medical ethics.

  21. trrll says:

    I don’t think I’ve ever had an immediate response to an oral medication. In fact it seems like everytime the doctor prescribes something, it doesn’t work as well or as fast as they expected.* I wonder if those of us with rather negative outlooks on life are placebo resistant.

    Pharmaceutical companies would love it if they could identify individuals who are “placebo resistant,” because that would reduce the placebo background in clinical trials and make it easier to detect a pharmacological benefit. Sometimes, companies will begin a trial with a “placebo run-in” in which everybody gets a placebo, and dismiss from the trial anybody who reports an effect, in hopes of enriching the population for placebo resistant subjects. However, the evidence does not support the idea that this reduces the magnitude of the placebo “response.”

    Of course you could have a “nocebo” response to oral medications that makes you less likely to feel a real pharmacological effect as well as less likely to feel a placebo effect, in which case there would be little advantage to eliminating you from a clinical trial.

  22. nybgrus says:

    Interesting discussion so far. Can’t wait for pmoran to show up and hear what he has to say (I mean that seriously Peter).

    Todd – thanks for linking the article, I was thinking of exactly that as I read through the comments. I think that is the point here – there is no reason to believe that the placebo effect (which, as Benedetti pointed out in his prior work really should be placebo effects and separate from placebo response) can be elicited without deception being involved at some level.

    Furthermore, placebo effects are transient and ephemeral and do not have evidence of altering objective outcomes vis-a-vis the natural course of a disease. The oft cited apparent contradiction is the placebo activation of dopaminergic pathways in Parkinsonians. This does indeed lead to an objective change in functionality, but it does actually change the natural course of the disease.

    This brings me to a point that fledarmus1 made – what if we could “control” the placebo effect pharmacologically. Well, we do! The whole point is that we are demonstrating, via sophisticated neurobiology experiments, that there do exist ways of inducing the body to release endogenous compounds and/or activate endogenous receptors without any active substance – in other words, just by believing it should be so. Well… that is precisely what pharmacological therapies do. They either stimulate receptors or block them, either by activating (or deactivating) them directly (acting as mimics of the endogenous compound) or by altering some pathway upstream.

    An interesting question arising from this then would be “Can we harness placebo to do what we already do with drugs, but with no side effects?” I do not think there is currently a solid answer on this. I think that side effect profiles would certainly be less (in this hypothetical world where we could activate placebo effects at will) but I don’t think non-existent. Couple that with the ephemeral nature of placebo effects and the highly variable effect sizes and highly variable mechanisms for inducing it and I think we don’t have much here in terms of a clinically applicable notion.

    Clearly, lying to patients is not an ethical option so in my opinion placebo medicine is a question already answered. I also believe that we have an ethical onus to inform and educate our patients beyond merely enough information to cajole them into signing a consent form. So there goes “looking the other way.” And I also think we have an ethical onus to ensure the most reliable, safe, and consistent medicine we can provide, not something that leave a patient dangling in the wind (metaphorically) when they suddenly realize they have been duped with a placebo or, at least equally as likely, have some experience we counteracts via nocebo.

    And that is the point that I will close with – as much as we are all focusing on placebo effects and how interesting and quirky they are and trying to find some use for them clinically, we must also realize that the nocebo effect is just as real and powerful. Something as simple as a bad day – either on the part of patient of practitioner – can immediately erase all the positive placebo effects garnered to date. Stumbling across an article or newsreport, feeling lousy because you and your partner got into an argument, disliking your physician because (s)he had a bad day and was an arse to you, or even just getting cut off in traffic can all obliterate the placebo effects you have been relying on. For a patient with chronic pain, that can leave him/her untreated and in agony with no immediate recourse. I find the concept of that quite unacceptable.

    Placebo effects research is incredibly useful from a neurobiological and medical research standpoint. I do not see it having a future in clinical application and it most certainly does not have any application currently.

  23. nybgrus says:

    @mouse:

    I don’t think I’ve ever had an immediate response to an oral medication. In fact it seems like everytime the doctor prescribes something, it doesn’t work as well or as fast as they expected.* I wonder if those of us with rather negative outlooks on life are placebo resistant.

    Exactly my point. There is no way to predict, titrate, or modulate the effects with any sort of accuracy nor to maintain them with anything but luck.

  24. nybgrus says:

    @trrll:

    I’ve never bought Dr. Gorski’s objection to the study, because the patients can only be said to have been misled if the placebo failed to work.

    What do you mean? I am honestly confused by the statement. The objection that Gorski has (and which I share) is that the patients were primed to expect that they should have a placebo response. They were merely substituting one expectation for another.

    In other words, if I give you a sugar pill and tell you it is morphine you can get a buzz because you expect that the pill will do something. You don’t even need to know what morphine is – I can just say “this pill will make you feel a little loopy and help take the pain away” and that is sufficient to induce a placebo effect.

    But if I tell you “this is a sugar pill with nothing active in it, but we have reason to believe it will still make you feel loopy and take the pain away” I have still done the same thing – primed your expectation. I can see there is indeed a fine distinction there wherein this second scenario is not quite as obviously directly lying, but it is indeed a form of deception nonetheless and identical to the first example – merely separated by one degree of a lie.

    So if the placebo failed to work, then we would actually have concluded that they were not mislead, since we expect that misleading people induces placebo. And telling them that sugar pills have been shown by “rigorous clinical testing to produce significant mind-body self-healing processes” is indeed misleading them and priming them.

  25. trrll says:

    What do you mean? I am honestly confused by the statement. The objection that Gorski has (and which I share) is that the patients were primed to expect that they should have a placebo response. They were merely substituting one expectation for another.

    That can only be said to be a false priming if it did not result in a placebo response. But it did. So the objection boils down to: subjects were correctly informed of the contents of the pill and what it was expected to do.
    Where is the deception in that?

  26. trrll says:

    Furthermore, placebo effects are transient and ephemeral and do not have evidence of altering objective outcomes vis-a-vis the natural course of a disease.

    I know that it is a common belief, but is there real evidence that the effects of placebos are any more transient than the effects of other medications? I’m not aware of any. And of course, there are pharmacological medications that produce transient effects, but for which there are nevertheless valid indications. The same is true for drugs that do not alter the natural course of the disease; in particular, this is true for many drugs given for pain relief, which is one of the most robust effects reported for placebos.

    An interesting question arising from this then would be “Can we harness placebo to do what we already do with drugs, but with no side effects?” I do not think there is currently a solid answer on this. I think that side effect profiles would certainly be less (in this hypothetical world where we could activate placebo effects at will) but I don’t think non-existent.

    Subjects receiving placebos in controlled trials certainly report side effects. Side effect risk would need to be evaluated just as for an active drug.

    Clearly, lying to patients is not an ethical option so in my opinion placebo medicine is a question already answered.

    But that doesn’t apply if patients can exhibit a belief-independent (perhaps conditioned) response to a placebo. Moreover, I suspect that there is a population of patients who would be willing to provide informed advance consent to the use of a placebo, even with misdirection, if in the doctor’s opinion it would yield a benefit. Such consent is considered ethical in the context of a clinical trial. Does it become unethical if done in the context of clinical practice?

  27. Gregory Goldmacher says:

    There are two potential slippery slopes here.

    One is the slippery slope of the quack. Once you hold up placebos as useful treatments, you are implicitly supporting all quackery, because hey, the snake oil might provide benefits from the placebo response, so why not use it therapeutically all the time.

    The other is the slippery slope of cynicism. In the discussions on this blog I have seen that anything that speaks to the patient’s narrative, rather than objectively documented disease manifestations, is dismissed as leading to quackery, because quacks take advantage of that first slippery slope. This means telling patients who benefit psychologically from some “treatment” – no matter how benign (meditation?) – that they aren’t really getting any benefit. Seems kind of heartless to me.

    CAM types are masters of managing people’s illness narratives. In some diseases, where perception and expectation are particularly meaningful (e.g. illnesses with unpredictable course, illnesses with difficult to manage pain, etc.) and no definitive and highly effective medical treatment exists, CAM purveyors can easily convince people that they are doing some good. It does not justify deceiving the patients, but the concept of the illness narrative as something that can be managed independently of the actual effective medical treatment has some explanatory power with regards to what CAM peddlers do. I think managing the illness narrative has some value of its own, and we in the SBM and medical communities should pay more attention to it.

  28. Provi says:

    I may be being overly simplistic here. But I think its pretty cut and dry ethically barring in research (where as I assume people are said ‘you may be on the placebo’ but aren’t told if they are definitely). Saying that medicine is acceptable so long as it elicits the ‘placebo’ effect is basically a catch-all defence that can apply to anything from pharaceutical products to quackery.

    Any actual treatment thats promoted should have to prove its efficiacy. If the limits of a medical treatment is just based on whether it makes a patient feel good thats no standard at all. God knows it coudl even be used as a defense for harmful treatments: ‘the positive placebo effects outweighed any negative symptoms imposed by the treatment’.

    Whether its ok to do that for sugar pills for people with a cold or fever? Thats basically the limit I think. Maybe inducing a placebo effect in terminal patients might be desirable. But this doesn’t seem to be a slippery slope to me if its accepted, rather a sheer drop off a cliff.

  29. DevoutCatalyst says:

    CAM types are masters of telling their clients that physicians haven’t a clue, that CAM healers have special knowledge that doctors couldn’t possibly have because DOCTORS are so lame.

  30. qetzal says:

    daedalus2u,

    Declaration of Helsinki notwithstanding, drugs in the US are rarely approved based on a formal showing of clinical superiority over existing treatments. Quoting Robert Temple, MD, Deputy Center Director for Clinical Science, USFDA/CDER:

    FDA’s experience with comparative effectiveness claims is relatively limited. Our enabling law (FDC Act, as amended in 1962) does not require assessment of comparative effectiveness and the legislative history made it very clear that there was no relative effectiveness requirement. A new drug does not have to be better than, or even as good as, existing treatment.

    See http://www.fda.gov/downloads/Drugs/NewsEvents/UCM209270.pdf, slide 2.

    He further explains that even when there’s an effective treatment for a serious illness, most new drugs are only tested to show non-inferiority (within some acceptable limits).

    And as I asked before, is it unethical to prescribe statins to a patient who would benefit more from improved diet and exercise? Did FDA require trials comparing statins to lifestyle changes? Did any country require such trials before approval? I think the answer to all three questions is no, and I don’t see why a sugar pill should have to meet a standard that doesn’t apply to any other intervention.

  31. Jann Bellamy says:

    Lying or intentional misrepresentation in the sale of a product or service is legally fraud. I don’t see how an act that constitutes fraud could ever be considered anything other than unethical.

  32. nybgrus says:

    @trrll:

    That can only be said to be a false priming if it did not result in a placebo response.

    I apologize but I am still not following. If the priming was working, then the placebo response would be noticed. If the priming was not working, then no placebo. I am not sure what you mean by “false priming.”

    So the objection boils down to: subjects were correctly informed of the contents of the pill and what it was expected to do.
    Where is the deception in that?

    They were correctly informed of the contents of the pill but incorrectly informed as to what to expect. The way they were informed would be the equivalent of me giving my patient a dose of NSAID and saying I expect it to get them rip roaring high and loopy anyways. Also, the quote in the study stated “have been shown in rigorous clinical testing to produce significant mind-body self-healing processes.” That is patently misleading and a lie. There are some clinical tests that have shown significance – but that is statistical significance, not clinical significance. A lay person would be mislead to think along the lines of the common – not scientific – usage of the word. The second part is also false – there is no evidence, rigorous or otherwise – that shows the placebo effect can leading to “healing” processes. It is indeed still a lie to the patients. If it were worded more like “This is a sugar pill that is completely inert. However, studies have shown that if you really believe it is a drug it can still help you with your symptoms” that would be much less misleading and I wager have a much smaller effect size and significance.

    As I said, it is not a direct and overt lie, but it is certainly a lie and telling people that they should expect an effect.

    I know that it is a common belief, but is there real evidence that the effects of placebos are any more transient than the effects of other medications?

    I suppose it would be better to rephrase in order to be more accurate. You are correct that drug effects are transient – after all we don’t give morphine once for life. The key word I used (and will use solely from now on) is ephemeral. We know the pharmacokinetics of drugs so we can reliably estimate the intensity and duration of effect for the population at large. For individual patients we can readily titrate the effect. We can also reverse many drugs but, and this is key, the patient typically cannot. We can give narcan to reverse an opioid but the patient cannot suddenly stop being influenced by the drug spontaneously on his/her own.

    Placebo effects are ephemeral in that we have no way to anticipate, regulate, or otherwise control the “pharmacokinetics” of it. Induction of effect is highly cultural and individually specific – to the point where it becomes nigh impossible to determine the best way to induce effect. It can also be completely nullified by factors outside of anyone’s control. A patient may be racist, for example, and a placebo effect able to be induced by a white doctor would not be able to be induced by a black doctor – by the same or any other method.

    Additionally, the level of effect cannot be reliably titrated. A person in pain, for example, might not be able to have it adequately managed all the time.

    Lastly, it can disappear just as fast as it came about. As I said above, if the patient discovers it really is a placebo, that their doctor has lied or mislead them (in any way), or even if they have just had a bad day and are not as suggestible, then the effect can disappear. In fact, the nocebo can kick in as well. For example, lets say that a patient is informed as in the IBS study – “I am going to give you this sugar pill. It is completely inert but if you believe it works, it will help with your pain and heal your body.” So the patient takes it and feels better. Then he or she comes along to SBM (or any other site) and sees someone else writing about how that sort of stuff is just “all in your head” and “isn’t real” and you just “feel better because you are tricking yourself.” And bam! The patient is angry, thinks their doctor lied, and placebo effect is gone.

    But that doesn’t apply if patients can exhibit a belief-independent (perhaps conditioned) response to a placebo.

    Agreed. However, there is no way to induce a placebo response without at least some level of belief dependence via deception. Benedetti has shown that indeed you can condition the response by using real drugs and then switching to placebo. However, that would rely on lying at some point. If the person agreed to beforehand, I suppose that would probably circumvent the ethical dilemma – at least superficially. The problem then is actually doing it. Conditioning the response takes time and resources. And if the patient is aware of the fact that (s)he signed a consent for it, then (I reckon) they would be more vigilant and there would be less likelihood of success. In other words, a lot of effort for not much benefit, with a significant chance of failure, plus the ever looming possibility of it dissipating for myriad reasons. As I said – I just don’t see a potential clinical application here.

    Such consent is considered ethical in the context of a clinical trial. Does it become unethical if done in the context of clinical practice?

    Yes. It does actually. The ethics of human subjects research is such that it is always considered unethical – even in a research project or clinical trial. There are exceptions when certain criteria is met – and it is extensive. However, it is an exception for reasons of increasing the overall good while absolutely minimizing the individual harm. It is never actually considered “ethical.”

  33. nybgrus says:

    The other is the slippery slope of cynicism. In the discussions on this blog I have seen that anything that speaks to the patient’s narrative, rather than objectively documented disease manifestations, is dismissed as leading to quackery, because quacks take advantage of that first slippery slope. This means telling patients who benefit psychologically from some “treatment” – no matter how benign (meditation?) – that they aren’t really getting any benefit. Seems kind of heartless to me.

    Nobody here is discounting the patient narrative. I personally, and I am certain all of the authorship here agree, feel that the patient narrative is extremely important. However, we also feel that you cannot divorce the two concepts. There is the old adage of “treat the patient, not the number.” That statement is wrong. You must treat the patient and the number.

    And there is indeed plenty of mention on this blog of the importance of the patient experience in healthcare.

  34. pmoran says:

    Thankyou for bringing this up, Harriet.

    As some will realize from my clumsy contributions over the past few years, I have been hoping that SBM would look seriously at the implications of modern placebo research.

    It actually has ramifications for all aspects of medicine, even basic concepts, such as what it means to say a treatment “works”, something that EBM and SBM (mostly) have regarded as a simple and well-set matter.

    If there is some potential for limited benefits from almost any “treatment” our patients might choose, we also have to ask to what extent we are justified as doctors systematically working against what patients choose to do at their own expense and risk, and especially when we have nothing superior to offer. Perhpas this is what the so-called shruggies sense.

    That is part of the reason for my suggestion that our “healthfraud” position (as doctors) should emphasize safety concerns. Of course safety includes not being made to overlook more effective treatments when available , but the shift in emphasis would entail a slightly more laid-back attitude when methods are reasonably safe . There should still be monitoring of the claims arising from within CAM and attempts at regulating any that could lead to serious harm.

    If a different kind of skeptic wants to put more emphasis on the fraud content of CAM, or the pseudoscience, they are welcome to do so. However THIS doctor feels that WE doctors should as far as humanly possible be seen to be on our patient’s side against the common enemy of illness. That is what the public wants of us and it looks to me to be a rough and ready guide for all our dealings with CAM.

  35. Jimmylegs says:

    The slippery slopes Gregory stated are a real concern. What is to stop anyone, now and in the future, to come up with a crazy idea that is impossible to work on the pathophysiology (like waving a dowel rod over ones head 3 times) but it creates a placebo effect (patient feels better). This is in every way, shape, and form unethical.

    @qetzal:

    “Besides, suppose a patient needs to reduce stress. The doc tells them to take a sugar pill and imagine that the simple act of taking will help them relax. IF there were data showing that worked, why would that be unethical? Especially since it wouldn’t preclude any other stress-reducing intervention that might be implemented.”

    If the doctor says it’s a sugar pill or not, I still feel that it is unethical. I feel that way because you are trying to get some benefit for a treatment that is known to do nothing. I would inform the patient to try to identify their stressors and act accordingly (e.g. driving to work by taking the main stress has a lot of traffic which stresses you, try taking the highway or another route even if it takes more time).

    Lifestyle changes can help people, I wish there was a magic pill that makes all our problems go away but I live in reality (this is in no way a negative comment/tone to you, just a statement of fact).

  36. qetzal says:

    Jimmylegs

    One of my conditionals was that clinical testing had already shown a benefit of giving an open placebo using that approach. So no, you wouldn’t be trying to get a benefit from something that was “known to do nothing.” True, the sugar in the pill would not be doing anything. But the act of taking the pill, in full knowledge that it was only sugar, would still be providing a benefit.

    Of course, I’m presupposing you could demonstrate that benefit in a clinical trial first. If you couldn’t, then I’d agree with you that it would be unethical because you couldn’t honestly promise an expected benefit.

    Does that make sense? If so, does it change your opinion?

  37. PJLandis says:

    ” Sometimes, companies will begin a trial with a “placebo run-in” in which everybody gets a placebo, and dismiss from the trial anybody who reports an effect, in hopes of enriching the population for placebo resistant subjects. However, the evidence does not support the idea that this reduces the magnitude of the placebo “response.””

    I’m interested in reading more about that. I’ve never heard of weeding out placebo-responders before and it sounds intriguing.

  38. Jimmylegs says:

    @qetzal:

    I honestly would still be apposed to it. Reason being your not treating anything, your just making the patient “feel better.” If the patient had a terminal illness that was going to kill them in the next few days, I’m sure there are drugs out there right now that will make them feel better without attempting a placebo effect.

    It all goes back to what Gregory said, if you can show a placebo has some benefit (regardless of the patient knowing or not) than all of CAM can be viewed as ethical. However, the CAM proponents would have to admit it’s purely placebo.

    My view point may change in the future, but as of now I cannot say I would get a placebo to a patient, with both parties, knowing full well that it is inert. There must be something I could do to help like lifestyle or diet, anything.

  39. nybgrus says:

    @qetzal:

    Does that make sense? If so, does it change your opinion?

    The issue I see is that in order to ethically make that claim a true test of placebo has to be done – one in which there is no question that patients were not mislead in any way. In other words a test framed as:

    “Here is a sugar pill. It is completely inert and should not have any effect on you whatsoever.”

    If people still experienced a placebo effect based improvement in symptoms only then could we say, ethically, in the next trial:

    “Here is a sugar pill. It is completely inert and should not have an effect, but some people still experience an improvement in their symptoms anyways.”

    I don’t think such a study has been done and I think the evidence points to concluding that such a study would show no reliable effect.

    So really it is a catch-22. You have to mislead or lie at least a little and then use that to further mislead subsequently in order to create an effect. No matter how you slice it, it all boils down to necessitating some sort of mis-direction that is unavoidable. Otherwise we’d all just be taking spoonfulls of sugar and skipping the medicine going down (and all those woo-ey people who think that they can just think themselves into feeling better and curing their ailments would have succeeded).

  40. fledarmus1 says:

    I think I am confusing placebo effect and outcomes studies in some of these discussions. For those of you stating that the placebo effects are ephemeral, are you saying that over the long term, the people observed to improve in the placebo arm of a trial will eventually revert to be indistinguishable again from the general population, while those observed to improve in the treatment arm of a trial do not? In that case, it would seem that if you just ran the trial longer, the placebo effect would disappear?

    I think the ethical way to distribute a placebo would be with the simple facts and a choice. “Of those treated with the placebo, x% showed this amount of improvement over this time, while y% showed these side effects. Of those treated with the drug, a% showed improvement over this time, while b% showed these side effects. Which would you like?”

    The reason this doesn’t work with msot of the woo (especially “natural” and “supplement”) is that neither x% or y% are ever measured in any significant way.

    For those that have tried the drug and were not part of the a%, then trying the placebo and hoping to be part of the x% may be a valid option. For that matter, those terrified of the b% might think the difference between x% and a% isn’t so bad. And for actual effective treatment, being able to say that 80% of the people taking the drug showed long-term improvement, while only 10% of those taking the placebo showed any improvement at all and it was strictly short-term, would be a hard argument for any placebo-seller to overcome, unless the b to y contrast was simply horrendous.

  41. mousethatroared says:

    Harriet Hall “But now we are learning that while the placebo itself is inert, the act of giving a placebo is not: it can produce actual physiological effects through suggestion and expectation.”

    A number of people seem to be stuck on the placebo’s don’t do anything” mode. In my mind, Harriet Hall’s statement should be the basis of people’s judgement. The placebo/brain interaction does do something, although there are questions of reliability and comparison to other methods.

    If a doctor hands a patient a pill and says “Our research has shown this can help you symptoms.”(And the research actually shows that a placebo can help the patients symptoms,) then that is no more dishonest than the typical, using the typical euphemism “uncomfortable” for a procedure that many non-medical people would honestly report “hurt”.

    ——-

    Also, I think it’s less productive to discuss the use of placebo’s in generalities, because people are imagining particular scenarios that, possibly confirm their bias that placebo is bad or good.

    Hypothetical situations might be useful. Here’s a couple I thought of, although the medical people could probably come up with more accurate ones.

    What if a doctor is giving a patient with Parkinson’s a hypothetical medication. The medication is helping symptoms, but may have a negative effect on kidney or liver function. The doctor tried to cut back on dosage, but the patient has more symptoms. Supported by (hypothetical) research into placebos in this disease’ the doctor gives the patient a calender pack of medication at the effective dosage, but every 5th pill is a sugar pill that looks like the others.* The Doctor schedules a two week follow-up to confirm that the placebo is working (based on conditioning) and the patients response to medication is still positive.

    Is that unethical?

    Another hypothetical situation. A patient visits a doctor with “Morgellon’s”. They tell the doctor that antibiotics have worked in the past, but his doctor won’t prescibe them anymore. The docotr suspects that a placebo would “work” as well as the antibiotics, but they do a work up to check for other causes. Nothing else turns up. He give the patient a placebo saying This is a medication I’ve recently tried for this skin conditions similar to yours. It think it can relieve your symptoms.” And schedules a follow-up in two weeks.

    Is that unethical?

    *Benedettii discusses conditioned response to pain medication in his interview on Brain Science (I think). This is based on that.

  42. qetzal says:

    Jimmylegs,

    If CAM were truly honest about being placebo effect, I think it would be more ethical. Suppose someone fully understands that homeopathy is just water, that there is no reasonable evidence of efficacy, and that it violates fundamental science. Suppose the person selling it is completely honest about all that, saying nothing more than that they personally believe in it anyway. Is it still unethical to sell it to a fully informed and willing buyer?

    I don’t think so. I think the buyer would be acting foolishly, but I think barring the transaction would be more unethical because it unduly violates the buyer’s autonomy. If they want to believe in and pay for something ridiculous like homeopathy, and as long as the seller isn’t misleading them, that should be their right.

    nybgrus,

    I’m not convinced you would need to lie to elicit a placebo effect. I think something like my original scenario could well elicit subjective, symptomatic improvement in trial subjects with no deception at all. But if you’re right and I’m wrong, I agree it would be unethical. And even if I’m right, I suspect there would be more effective ways to ethically elicit the same benefit than with a sugar pill.

  43. daedalus2u says:

    fledarmus1, the problem is that the term “placebo effect” is not well defined and different people here are using different definitions. Some of the definitions are mutually exclusive. If you define placebo as “a treatment that does nothing”, then any treatment that has a positive effect is by definition not a placebo, and any treatment that fails to work (i.e. does nothing) is by definition a placebo.

    Placebo effects are not well studied, and part of that is because different researchers are using different definitions and many of these definitions get lumped together and used imprecisely.

    The definition I like is “a treatment with positive physiological effects where those effects are not mediated through pharmacology, surgery, or direct physical effects.”

    If a treatment has no effects, then there are no effects to consider about it and no physiology by which those non-effects are mediated. If a treatment has effects, then those effects can only be mediated through physiology. We may not understand the physiology (this true for most drugs), but if there are effects, those effects can only be mediated through physiology.

    There are many assertions about placebos which are made in the absence of data, but they directly follow from the definitions that some are using. If a placebo is a treatment that has no effects (by definition), then of course it can’t have positive long term effects (by definition). If it can’t have any long term effects, then reports of long term effects are in error, either due to wishful thinking, lies, or delusion on the part of the individuals who believe they have experienced long term positive effects, or error in labeling the treatment that had long term positive effects a “placebo” using the definition of “a treatment that does nothing”.

    A problem is that the quacks and scam artists have co-opted much of the language surrounding how the brain regulates physiology in terms of brain-mind duality so it is difficult to talk about how the brain regulates physiology without falling into mind-body woo woo speak.

    Many aspects of physiology must be regulated globally. The liver supplies glucose to the whole body, including the brain (which is a major glucose consuming tissue compartment). There must be global control systems that link glucose consumption by the brain with glucose production by the liver. How exactly that coupling is regulated is not understood. That there is coupling is indisputable. What would happen if that coupling got dysregulated? Either the brain wouldn’t get enough glucose and would fail, or the liver would try to produce more glucose than it has the capacity to do so and would fail. What would not enough glucose to the brain look like? It would look exactly like neurodegeneration. What would the liver trying to produce more glucose than it has the capacity to do so look like? I think it would look like chronic liver failure.

    The regulation of liver gluconeogenesis capacity is not under conscious control, but there have to be control signals from the glucose consuming tissue compartments (including the brain) that do control it. When a fight-or-flight state is entered, the liver ramps up its glucose release rate. To some extent, that increased release is not sustainable indefinitely. The liver has to “stand down” from its emergency state to restore its glycogen reserves, to repair whatever damage has occurred, and to grow itself to better match a higher demand in the future.

    A fight-or-flight state can be entered through non-pharmacological, non-surgical and non-direct-physical means. The image of a bear on the retina can trigger a fight-or-flight state. Is having a bear chase someone a placebo? It certainly does have physiological effects, short term and long term (if only an increased fear of being chased by bears and an aversion to medical professionals who use such treatments ;) ).

    If a “placebo” (an image of a bear on a retina) can trigger as severe a physiological response as fight-or-flight, why should we be surprised if other “placebos” can trigger less severe responses?

  44. Jimmylegs says:

    I accept that a placebo can have an effect on a patient, but I still don’t feel right giving a patient one. I do not know what the reliability on it would be, perhaps some one else knows better?

    @mouse:

    Scenario 1: I don’t really have an answer for that, I don’t think I would be ok with it.

    Scenario 2: Aside from Morgellon’s disease (or similar disease) being interesting, I would be apposed to this scenario. Reason being, this is a psychological problem and needs to be treated as such. I find this to be akin to giving sugar pills to a patient with depression or anxiety, it may help but it is still deceit (the wording does not disclose it being a placebo).

    Again I am aware and accept that a placebo can have an effect on a patient. But I feel even when it is used as a last treatment it is still unethical. Instead of telling the patient “I’m sorry but there is nothing left I can do for you” it seems like (to me) you may be saying “here try this, this might help you” and you’re hoping for a positive response to make them feel better.

    I have not taken any kind of class or lectures on medical ethics (or ethics in general), so my position may change in the future. If the general consensus among doctors (non-quacks) that a placebo is fine to give to a patient under certain circumstances I would probably change my stance.

    I truly do enjoy intellectual discussions (something I rarely get among peers) so I thank you all very much.

    1. Harriet Hall says:

      I don’t condone either the Parkinson’s or the Morgellon’s examples. Both are dishonest. The pill pack is given to the Parkinson’s patient in a way that makes him believe they are all active pills: that’s lying. The Morgellon patient’s delusion is reinforced.

      Instead of telling a patient “I’m sorry but there is nothing left I can do for you” or offering a placebo, there is another option. We can look for ways to help the patient cope with his symptoms and live with them while improving his quality of life in ways others than taking pills. In pain clinics they are often unable to eliminate the pain entirely, but they focus on lifestyle and attitude changes that allow the patient to function more normally in spite of the pain.

  45. Gregory Goldmacher says:

    Harriet,

    The approach you describe is exactly what I mean when I talk about physicians taking narrative into account.

    ~G

  46. mousethatroared says:

    Harriet Hall, If I understood the Benedetti interview properly, you could tell the Parkinson’s patient that some of the pills were sugar pills. Since the effect is conditioned response (Dog, bell, saliva) to the color, shape, process of taking the pill, the intellectual knowledge that some are sugar pills, wouldn’t prevent the physiological response.

    repeat – if I understood it correctly – still seems kinda wacky to me.

    The one problem with telling the patient that they might be looking for differences caused by the sugar pills, which could give unpredictable result…I don’t know. I’m completely speculating.

    If you told the patient what you were doing, would it be unethical?

  47. mousethatroared says:

    would it STILL be unethical, that is…

    not would telling the patient be unethical, like it read.

  48. Harriet Hall says:

    @mouse,

    If a patient told a doctor that he would pay him in cash but that sometimes the cash would be counterfeit, would that make it ethically acceptable to pay with counterfeit bills? (Apart from the legal implications)

    A friend of mine once asked his doctor if he could sign a blanket permission for the doctor to sometimes use placebos if he thought it was in his best interests (i.e. a non-serious condition where a real medication might cause adverse effects). The doctor refused, and I would too. Because it still involves deliberate deception and paternalism.

  49. Scott says:

    @mouse:

    Absent clear and convincing evidence that the same effect cannot be attained without the sugar pill (e.g. by improvement in the doctor/patient relationship), IMO it would still be unethical.

    As I see it, the professional trust patients place in their doctors means that there is always an implicit understanding that whatever’s suggested by the doctor is in fact efficacious (to be clear, I mean a specific effect not a placebo effect). Hence the absence of an explicit claim of efficacy, or even an express disclaimer of efficacy, is largely unavailing. The mere fact that the person in the white coat suggested it will lead the patient to infer such efficacy.

  50. mousethatroared says:

    @Harriet Hall & Scott – Just to be clear. I’m not advocating either scenario, but I don’t understand.

    How is the sugar pill every five days (when the patient knows they’re there) different than the dummy pills in a birth control pack or the different doses in steriod taper pack?

  51. Harriet Hall says:

    @mouse,

    “How is the sugar pill every five days (when the patient knows they’re there) different than the dummy pills in a birth control pack or the different doses in steriod taper pack?”

    BIG difference. The patient knows which pills contain what.

  52. mousethatroared says:

    But if it’s important to the patient to know, they can say, “No, I’m not comfortable taking this when I don’t know which pills are which.” And you do have the follow up, if things aren’t working.

  53. nybgrus says:

    @qetzal:

    But you can’t say that either. There is, by definition, deception somewhere in that chain no matter how hard you try and distance yourself from it. We can only say that placebo has effect in X% of cases because at some point people were actively lied to in order to observe that outcome. Either they were unintentionally lied to by a genuinely believing CAM practitioner or they were intentionally lied to in the course of a clinical trial. Either way, there exists nobody who has ever picked up a sugar pill knowing it was nothing but that and expecting it to be inert and it ameliorating something. Otherwise Britain would be the healthiest place on earth with all those lumps of sugar they put in their tea ;-)

    All you are doing by telling your patient “studies have shown…” is displacing the lie from yourself directly to the study indirectly. It is still deception and still unethical.

    I will also end with a similar caveat as you – if a study is done that really and truly had no deception whatsoever and yet still a measurable placebo effect then I will amend my view. To my knowledge no such study exists and all the evidence to date leads me to believe such a study would not pan out.

  54. Harriet Hall says:

    @mouse,

    “they can say, “No, I’m not comfortable taking this when I don’t know which pills are which.”

    If the patient refuses your offer of placebo deception, that doesn’t negate the intent to deceive. Still unethical, IMHO.

  55. nybgrus says:

    @mouse:

    If a doctor hands a patient a pill and says “Our research has shown this can help you symptoms.”(And the research actually shows that a placebo can help the patients symptoms,) then that is no more dishonest than the typical, using the typical euphemism “uncomfortable” for a procedure that many non-medical people would honestly report “hurt”.

    See my above comment to JimmyLegs wherein I posit that the research does not show that placebo can actually help patients – it shows that lying to them and suggesting help does. Therefore the claim “Our research has shown this can help your symptoms” is, in my opinion, a direct lie but at least an indirect misleading based on a faulty interpretation of the data.

    To contrast that with the euphemism scenario it is quite different since indeed many people would genuinely find procedure [X] “uncomfortable” rather than “hurt.” To put it conversely, there are procedures that the vast majority of people find innoccuous or only very mildly distressing but that a select few find extremely unbearable (like venipuncture for example. Most people aren’t bothered too much by it, but my girlfriend has panic attacks and passes out). It would not be reasonable to tell everyone getting venipuncture that it will be “extremely unbearable.”

    In cases where the most common experience is actually painful rather than merely “uncomfortable” I agree that it would be unethical not to say so. And in my practice of medicine to date I have done exactly that – I tell people that this will hurt and that there is no way around that, but that I have found people tend to find it less distressing than they typically imagine. I also add that it is not my goal to hurt them and I will do everything I can to minimize that aspect of the otherwise unnecessary procedure. I have found without exception that this tack yields excellent results.

    re: your scenarios:

    1) That is a very good scenario. I do think, however, that it would also be unethical and not likely clinically useful. Dr. Hall has addressed the basic ethics involved. The clinical aspect is interesting as well. Parkinsons is a waxing and waning but inevitably downhill process. Managing it in the longer term would prove difficult if the physician didn’t know what the actual dose of medication was, especially since we don’t have a way to measure the placebo induced basal output of dopamine. Changes in symptom status could be due to a worsening of the disease process, a decrease in the placebo induced dopamine release, or any other factors. Changing dosages to accomodate this would be quite difficult. This relates back to my earlier post where I discuss the difficulty in using placebo clinically since titrated doses and knowing the “pharmacokinetics” is essentially impossible.

    2) This one is readily addressed – it is considered unethical because it is poor treatment to acknowledge a patient’s delusions. Psychiatrists consider the acceptance of a patient’s delusion as not only not actually treating the pathology but also likely to worsen the condition. You wouldn’t treat a paranoid delusion by hiring the person a body guard.

  56. nybgrus says:

    sorry – I should have said qetzal when I said JimmyLegs in the above post

  57. nybgrus says:

    @mouse:

    they can say, “No, I’m not comfortable taking this when I don’t know which pills are which.

    Either way it reeks of paternalism in that you are atempting to decide which scenario is best for the patient. If they reject your offer then I would agree little harm is done, though doctors shouldn’t be in the practice of even attempting to deceive their patients.

    But if they accept, you are still deceiving your patient – knowing or not. In this case the principle of the matter is actually important (I find the ends rarely justify the means and if they do it is only because you’ve managed to “get away with it”) but also there are potential consequences. What if there is an adverse outcome? Or what if the patient’s condition deteriorates rapidly? Who or what gets the blame then? Can you prove that the patient would have deteriotated just as fast without the placebo deception even if that really is likely? Do you think the patient will actually take comfort in assuming that to be the case?

    Being ethical and following efficacious standard of care not only protects the patient but it also protects the physician (and yes, we do deserve protection too!). Perhaps there will come a time when Benedetti’s research demonstrates reliable ways for us to monitor, titrate, and anticipate the effects of placebo and then it will become more viable. That certainly does not exist now and my opinion is that such a scenario will never arise because the only way to reliably control physiological responses is with known active molecules. Perhaps I will be surprised and change my tune in the future.

  58. nybgrus says:

    I just realized that I should also add that physicians have an added onus of ethical responsibility – we have much authority and power in what we say merely because of our title and place in society. Offering someone a placebo and saying it will have some sort of effect (or even as in the case of your Parkinsons example) means there will be pressure on a patient to accept what we say – even if they are genuinely uncomfortable at the proposition. How we phrase patient options and our opinions on the matter is actually important and something that I am actually taught in the course of my schooling. We need to – and are expected to – hold ourselves to a higher level of ethics than would otherwise seem reasonable for exactly that reason.

  59. bgoudie says:

    My wife is a PSY.D and her comment on this discussion was as follows, “Deception is unethical and worse damaging to any therapy you may be trying. Best to be upfront about the real nature of anything you present to a paitient.”

  60. daedalus2u says:

    Yes, even it you tell the patient you are giving them a placebo it is unethical. The physician-patient relationship is one of asymmetric knowledge. The patient hires the physician for the physician’s superior knowledge and judgment. The patient is entitled to know what he/she is getting.

    The patient has a right to expect that all treatments given or suggested are “standard of care”.

    If giving placebos were the “standard of care”, then physicians would be expected to lie in the course of treating patients. There is no profession where lying to the client is considered acceptable, except maybe prostitution. Not surprisingly that is what people who are paid to lie are sometimes called (no disrespect meant to honest and professional sex workers).

  61. mousethatroared says:

    nybrgus – I don’t think I ever suggested that doctors shouldn’t be ethical. The question is what is ethical in different situations in light of the new evidence that Harriet Hall was talking about.

    Also, In a thought experiment you have to abide by the premises of the scenario, you don’t get to argu with the hypothetical evidence.

    The problem is multi-faceted. As I wrote it, continuing the drug at the current levels would hold a higher risk of kidney or liver damage than using a placebo response to cut the dosage. So if we go through the question and answer that HH and I did, you are giving your patient a higher dose of a drug, so that they will know the individual doses of their medication each day.

    You have to chose between the patient knowing the details of their daily dosage and lowering patient risk.

    That’s the idea behind the question. If placebos can elicit a physiological response that will allow reduction of risk in medication, how do you balance between patient informed consent on small details of their treatment and patient safety?

  62. nybgrus says:

    very well, I will stick strictly to your hypothetical.

    I can see your scenario being ethical and utilizable only if it become standard of care. In other words, just as some dosing regimens require once a day vs twice a day, some also require once every other day dosing. If it becomes assessed by the appropriate expert bodies that administering a packet of Parkinson’s meds with with a certain amount of them being placebo and everyone knew that as the most efficacious and safest standard of care, then it would be ethical and I would prescribe it myself. In other words if we changed the dosing from daily drug levels to weekly or monthly, and administered them in packets of [X]g/month such that everyone involved was aware that some of those would randomly be sugar pills sure and it proved to be just as effective (or even slightly less effective but significantly less harmful) then sure, I would call that ethical.

    Also, if unicorns were real my girlfriend would be the happiest person on earth.

  63. mousethatroared says:

    Cool, thanks All. Several issues that I had not thought of…

    I had meant to make more observations, but I’m exhausted tonight.

  64. nybgrus says:

    I’ll look forward to them tomorrow, if you are still so inclined. My schedule tomorrow is very light so I will have a bit more time to devote to a response :-D

    I appreciate you bringing up the questions – it does help push the envelope and make me think a bit more.

  65. Jimmylegs says:

    My point is what bgoudie stated, that I do not feel that it is right to deceive patient because there is always other options, like a different prescription or lifestyle changes (identifying stressors, symptoms, etc.).

    In my previous post I should have restated that I would try to get lifestyle changes before I would do placebo. I would also start with the most effective, reliable, and cost effective choices and work my way from there.

    Ethics is an interesting topic indeed.

  66. PJLandis says:

    Funny, but I meant to post that to a different thread.

  67. nybgrus says:

    @jimmylegs:

    I agree with you. However, I believe the scenario I described above would be a reasonable usage of placebo as a first line treatment. As I alluded, however, I think the likelihood of that coming to fruition is slim to none for myriad reasons.

  68. BillyJoe says:

    “Funny, but I meant to post that to a different thread.”

    Oh, I thought for a moment that you were suggesting that god is a placebo and therefore unethical.
    That couldn’t be right. Could it?

  69. trrll says:

    I apologize but I am still not following. If the priming was working, then the placebo response would be noticed. If the priming was not working, then no placebo. I am not sure what you mean by “false priming.”

    The subjects were told what was in the pills and were told what the investigators expected to be the outcome, so nothing there can reasonably be characterized as misleading. I would say that it would be a false priming if the patients were told that they were likely to exhibit a placebo response if the investigators knew or believed that in fact such priming was ineffective.

    They were correctly informed of the contents of the pill but incorrectly informed as to what to expect.

    And based on the results, that was indeed what they experienced. So how were they incorrectly informed?

    The way they were informed would be the equivalent of me giving my patient a dose of NSAID and saying I expect it to get them rip roaring high and loopy anyways.

    Assuming that was indeed what you expected, and that it was indeed the result, which part of that would be incorrect?

    Also, the quote in the study stated “have been shown in rigorous clinical testing to produce significant mind-body self-healing processes.” That is patently misleading and a lie. There are some clinical tests that have shown significance – but that is statistical significance, not clinical significance. A lay person would be mislead to think along the lines of the common – not scientific – usage of the word.

    Now you are resorting to what I referred to above as “quibbling” about the meaning of the words. Yes, “significant” has a broad meaning that can encompass a number of things. It can be significant to a statistician (“statistically” significant), significant to the doctor (“clinically” significant), or significant to the patient. Considering that the fraction of patients who reported adequate pain relief went from about 35% to about 60%, I’d say that it satisfied the first and third meanings, although physicians might debate the second.

    The second part is also false – there is no evidence, rigorous or otherwise – that shows the placebo effect can leading to “healing” processes.

    Interesting that whereas above you insisted on a lay meaning of “significant,” now you are demanding that “healing” must satisfy a very limiting technical definition. Most lay people would include feeling better under the definition of healing. And I note that you have omitted a key part of what the patients were told, which was that the placebo could produce “significant improvement in symptoms due to mind-body self-healing processes.”

    It is indeed still a lie to the patients. If it were worded more like “This is a sugar pill that is completely inert. However, studies have shown that if you really believe it is a drug it can still help you with your symptoms” that would be much less misleading and I wager have a much smaller effect size and significance.

    But given that was not what the investigators expected, and not what the patients experienced (a substantial increase in the number achieving adequate pain relief in spite of not believing that the placebo was a drug), would not that have been a false statement.

    You are correct that drug effects are transient – after all we don’t give morphine once for life.

    There are certainly people with chronic pain who take morphine or other opiate pain relievers for much of their life.

    The key word I used (and will use solely from now on) is ephemeral. We know the pharmacokinetics of drugs so we can reliably estimate the intensity and duration of effect for the population at large. For individual patients we can readily titrate the effect. We can also reverse many drugs but, and this is key, the patient typically cannot. We can give narcan to reverse an opioid but the patient cannot suddenly stop being influenced by the drug spontaneously on his/her own. Placebo effects are ephemeral in that we have no way to anticipate, regulate, or otherwise control the “pharmacokinetics” of it. Induction of effect is highly cultural and individually specific – to the point where it becomes nigh impossible to determine the best way to induce effect. It can also be completely nullified by factors outside of anyone’s control. A patient may be racist, for example, and a placebo effect able to be induced by a white doctor would not be able to be induced by a black doctor – by the same or any other method.

    In practice, trials on humans are required to estimate the intensity and duration of the therapeutic effect of pharmacological agents. Even then, individual patients may react very differently from what the doctor expects, or what the patient is told to expect. There is a lot of individual variation in receptors and also in the P450 enzymes that mediate biotransformation of drugs, and interactions with the patient’s physiological state and disease progression introduce additional variability. (the technical term is “idiosyncratic” reaction–medical-speak for “I gave the right drug–it’s the patient’s fault”) You have not offered any actual evidence that the placebo effect is any more variable or unpredictable than the effects of pharmacological agents.

    By the way, having a specific antidote to reverse the effects of a drug is very much the exception rather than the rule. In most cases, symptomatic treatment is all that is available if a patient has an adverse reaction to a drug.

    Additionally, the level of effect cannot be reliably titrated. A person in pain, for example, might not be able to have it adequately managed all the time. Lastly, it can disappear just as fast as it came about.

    This is a problem with pharmacological pain relief as well (e.g. “breakthrough” pain). Patients with painful conditions are cautioned that they cannot necessarily expect complete elimination of the pain. Sometimes the dose required for adequate pain relief causes unacceptable adverse reactions, such as respiratory depression. Again, you have not offered any evidence that placebos are worse than active drugs in this respect, or that there is not a subpopulation of patients who will exhibit greater, more reliable pain relief from a placebo than from the pain medications that the patient can tolerate. This would have to be addressed by further research.

    As I said above, if the patient discovers it really is a placebo, that their doctor has lied or mislead them (in any way), or even if they have just had a bad day and are not as suggestible, then the effect can disappear.

    However, we are discussing an example in which patients were frankly told that they were receiving a placebo. Again, you have not provided evidence that pain relief from placebo is any less reliable than pharmacological pain relief, and anybody who has experienced pharmacological treatment for pain will tell you that the relief can be quite variable.

    In fact, the nocebo can kick in as well.

    And patients often exhibit adverse effects from active drugs. Shouldn’t the answers to risk-benefit questions be addressed in clinical trials rather than merely assumed, as you seem do be doing?

    But that doesn’t apply if patients can exhibit a belief-independent (perhaps conditioned) response to a placebo. Agreed. However, there is no way to induce a placebo response without at least some level of belief dependence via deception.

    Again, you are assuming facts not in evidence. A conditioned response does not necessarily depend upon belief, or even conscious knowledge.

    Benedetti has shown that indeed you can condition the response by using real drugs and then switching to placebo. However, that would rely on lying at some point.

    Again, facts not in evidence. The IBS study seems to indicate the opposite.

    If the person agreed to beforehand, I suppose that would probably circumvent the ethical dilemma – at least superficially. The problem then is actually doing it. Conditioning the response takes time and resources.

    Most patients presumably will have been conditioned to expect benefit from treatment in the normal course of their medical experience, so this is a non-issue.

    And if the patient is aware of the fact that (s)he signed a consent for it, then (I reckon) they would be more vigilant and there would be less likelihood of success.

    I’m not aware that the magnitude of a conditioned response depends on “vigilance.” Is there evidence of this?

    In other words, a lot of effort for not much benefit, with a significant chance of failure, plus the ever looming possibility of it dissipating for myriad reasons.

    And what kind of significance are you talking about here? To the statistician, the doctor, or the patient who is not responding well to standard therapy?

    Yes. It does actually. The ethics of human subjects research is such that it is always considered unethical – even in a research project or clinical trial. There are exceptions when certain criteria is met – and it is extensive. However, it is an exception for reasons of increasing the overall good while absolutely minimizing the individual harm. It is never actually considered “ethical.”

    So you are arguing that placebo-controlled trials or always unethical? You’ll find a lot of them on clinicaltrials.gov, which have presumably passed IRB ethical review. I think you must be using some personal definition of the word “ethical” here.

  70. daedalus2u says:

    The prohibition of the use of placebos in clinical trials is discussed in the Declaration of Helsinki (paragraph 32).

    http://www.wma.net/en/30publications/10policies/b3/index.html

  71. Scott says:

    How is the sugar pill every five days (when the patient knows they’re there) different than the dummy pills in a birth control pack or the different doses in steriod taper pack?

    I’d say the critical difference is the claimed function. I don’t know how the steroid pack works, but for the birth control pills they’re simply a memory aid – to help the woman remember to keep taking it every day, as opposed to “drat, I was supposed to start up again yesterday.”

    And a functional memory aid, they certainly are.

  72. nybgrus says:

    @trrll:

    Your responses are very superficial and miss the point. In fact it seems clear you didn’t even really read what I wrote:

    I stated:You are correct that drug effects are transient – after all we don’t give morphine once for life.

    Your response:

    There are certainly people with chronic pain who take morphine or other opiate pain relievers for much of their life.

    Utterly missed the point of what I was writing entirely.

    The rest of your argument basically boils down to a post hoc justification of genuine belief. “The study investigators truly believed that would be the outcome. And it was, so clearly they were right in their belief and thus justified the statement before the fact.” That doesn’t cut it.

    The fact that an individual truly expects it does not mean that it is valid to say. If I truly believe that advil will get you high as a kite, that doesn’t give me carte blanche to actually say it and be considered justified and ethical in doing so. We just call that “ignorant.”

    But besides that one little passage there are significant other issues with the study. Things like selection bias, Hawthorne effect and observer effect plus reporting bias are all heavily in play. So besides the fact that using the results of a study to justify the premise of the study is utterly non-scientific, you can’t even do that based on the results.

    As for the initial claim that you are continuing to try and defend, Benedetti himself stated quite clearly that the outcomes of placebo arms in clinical trials cannot be used to draw inferences about the placebo effects in play. So whilst the investigators may have truly believed their statement to be true, they did not have a valid body of evidence to rely on in making the claim.

    As for your understanding of pharmacokinetics and individual variation – it is superficial at best and your argument misses the mark. I am very intimately aware of exactly the issues you speak of, and I would reckon much more so than you. Regardless, there is a plethora of data to guide us and straightforward ways to titrate dosages. The same simply does not exist for placebo – unless you can find me some studies that demonstrate dosages for placebo analgesia. The rest of your argument is a tu cocque and doesn’t address the point. Yes, I know about breakthrough pain and yes I know about idiosyncratic effects and unwanted side effects. The fact that they exist does not address the core point of the distinction I was trying to make.

    You say I have not offered any evidence that the placebo effect is more variable – Benedetti’s own research shows it neatly.

    This distinction is also potentially significant when one is trying to interpret the
    results of clinical trials, because it may partly explain why the numbers for socalled ‘placebo responders’ vary so widely.

    You then go on to say:

    Again, you are assuming facts not in evidence. A conditioned response does not necessarily depend upon belief, or even conscious knowledge.

    I concede that technically you are correct. There are conscious and unconscious responses. For pain, a conscious response is needed. As Benedetti says:

    …you have to trust your doctor, you have to believe in the therapy you are getting, you have to expect, consciously, something.

    But even for the unconcious placebo effects (which, according to Benedetti can only affect unconcious physiological processes) there must be an actual administration of an actual drug in the beginning. After a certain period of time, regardless of what you say or actually give, the result will be the same. However, the administration must appear the same each time. You can’t go from giving a blue pill to a red one since that will negate the effect.

    Either way, there is deception involved, albeit less with the second scenario.

    So in sum, there is actually no evidence to support the statement made by the researchers in question, and the results are highly questionable anyways because of myriad confounders.

    As for your last point:

    So you are arguing that placebo-controlled trials or always unethical? You’ll find a lot of them on clinicaltrials.gov, which have presumably passed IRB ethical review. I think you must be using some personal definition of the word “ethical” here.

    Yes that is exactly what I am saying. And I am also certified in human subjects research and currently doing a research project with human subects, so no, I am not using a personal definition of ethical. As D2U was kind enough to cite for me, the use of placebos is considered unethical. However, in the case of a clinical trial an IRB ethics committee is supposed to weigh the factors and decide if an exception can be made to justify the ethical transgression by balancing the good of the patient, the community, medical science, and the harms expected from said ethical transgression. If it can be determined that there is direct potential for significant good and that it is unlikely to significantly harm the trial participants then the trial can go on anyways.

  73. mousethatroared says:

    nybrgus, HH and other’s who responded to my scenarios- As I said before, such thoughtful answers, it’s hard to disagree with any of them. A couple things stuck in my mind.

    Harriet Hall – you felt that in the Parkinson’s scenario that is would still be deceptive because the patient didn’t know which pills were sugar pills. Personally, as a patient, I wouldn’t feel deceived by that detail. It would be similar to having surgery and trusting that the doctors, anesthetists, will use appropriate medication while I’m out. BUT, I’m only one person, you’ve had the opportunity to get to know more patients, So if you think that many patients would feel this is deceptive, then I would take your word for it.

    Nybgrus, if you haven’t already read it, you might be interested. here is the paper that Benedetti was discussing in the interview I heard. http://www.neuro.cjb.net/content/19/1/484.full
    Understand, Benedetti was sure to clarify that he is against using placebos as treatment. His goal is to understand the neuroscience behind placebo.

    I think the points folks made about of standard of care is very import to the discussion and it was something that didn’t even occur to me.

    Sorry, to be rude if I don’t reply to any additional comments. I’ve been having some annoying throat, face rash, fatigue, symtoms, that I thought were going to be cleared up easily, with Prilosec and a change in facial products, Now it turns out that I’m visiting specialists and going through ruling out scary and extremely unlikely conditions before we all decide that I.need to take Prilosec for longer or elevate my bed and change facial products, laundry detergent or wear a hat or sunglasses or something. But until I get all the extremely unlikely, scary things ruled out, it’s probably better that I focus on non-health related diversions like facebook, gardening and teaching our new dog silly tricks.

    @nybrgus , unicorns? hehe. I’m sure that your girlfriend is at least consoled by the existence of narwhals. I know I am. Strange you never see them pictured swimming through clouds and rainbows.

  74. nybgrus says:

    @michele:

    When you go in for surgery you are given consents which include many of the things you’d expect. Obviously all the nitty gritty is left out since you don’t need to know the level of detail we do. However, explicit in that is the declaration that all interventions would be standard of care. So even then, there isn’t deception, merely omission which can be looked up in detail at any point. That is why I said if the randomly mixed in placebo pills method and dosing regimen became standard of care then that would be fine. But first one would have to demonstrate equivalence and safety of outcomes as well as how to manage changes in patient status before it could become standard of care.

    As for narwhals, yes she does enjoy them. And yes, the do fly through air as well :-D

    She also enjoys this silly song about them.

    I am sorry to hear about your current medical troubles. Hopefully those two links will put a small smile on your face. Best of luck. I hear that eating organic, removing toxins, taking plenty of vitamic C and omega-3′s, and eating bee pollen harvested from Tibet will clear that right up for you though. ;-)

  75. pmoran says:

    What doctors “should” do? There is a conflict between three impulses/obligations:

    1. “first do no harm”, which might warrant trying harmless placebos in some non-critical clinical settings before resorting to medications with potentially serious risks e.g. NSAIDs, (also see Mouse’s hypothetical).

    2. the duty to do everything we can that might benefit the patient. How methods “work” is, strictly speaking, irrelevant to that ethical obligation,

    3. We nevertheless find ourselves backed into just that corner through having to adopt “working better than placebo” as the convention via which the “effectiveness” of medical treatments for subjective complaints are judged, which in turn sets off the ethical qualms about using methods “unproven” by that standard even though they seem to “work”when applying the exact same measurements in other types of comparison.

    The overall debate is not helped by the evil connotations that have come to be attached to the word “placebo” — since the very beginning, when it was considered somehow shameful to be a placebo responder. .

    Another vexatious aspect is that regardless of the policies of professional organizations, 50% of doctors (or strictly speaking those who respond to surveys) admit to using placebos, presumably always in the form of CAM or drugs not being used for proper purposes. Scientific medicine has taken away many once popular treatment options while offering fewer replacements for some conditions.

    We get wound up about this partly through horror at the thought that someone somewhere is going to use a placebo treatment instead of important active treatments. That is not likely outside of CAM, yet even CAM methods are nearly all used in a complementary to mainstream care, or where it has nothing definitive to offer, or for minor illnesses and remote health expectations. Also in any conceivable conceptual environment in which placebo medicine came to be more tolerated it would be understood that we are addressing the subjective side of medicine not the pathohysiology of underlying diseases processes, where, in general, placebo are inactive.

    One matter helps sway the issue in my mind. That is that doctors brought up on EBM or SBM would probably not use placebos well, and certainly not if they have to also be wishy washy in any explanations they give to the patients as to what they might expect. So that the use of placebo will probably not be a generally cost-effective option under the conditions that apply in most medical practices.

    That is a reasonable conjecture but one that should be tested out . If we found otherwise, that complex interventions like acupuncture can produce the scale of benefits that some studies suggest within a particular conventional medical setting then the ethical argument does start to look somewhat contrived. “It seems to work, and we are not sure why.” is surely close enough to the truth.

  76. nybgrus says:

    “first do no harm”

    Does that mean harm to the individual or harm to the population?

    Does it mean harm in the near term or long term?

    Does it only mean direct physical injury?

    Surgery is harm in the near term for long term benefit. Screening is minor individual harm for population benefits.

    In a single patient, using a placebo instead of antibiotic for a viral illness might reasonably be argued as being “less harmful.” Certainly less physically harmful. How about psychologically in the long term? What if your patient discovers you have been giving a placebo? Does psychological harm count?

    And what if your patient finds out and is totally fine with it? Do the ends justify the means?

    What if we had a policy of prescribing placebos in such scenarios? What would be the population level harm generated from mistrust of the medical profession (as linked by SkepticalHealth)? How much physical harm would that indirectly lead to?

    I would argue that condoning the practice of placebo medicine, regardless of the individual outcomes and whether it “works” or not, would do more harm than good. Citing studies where 50% of doctors use them from time to time is not an argument. Many people exceed the speed limit – doesn’t make it any less illegal.

    So the argument is only further supported by the erosion of SBM by CAM and those people who substitute CAM for legitimate treatment and suffer for it. The basis of the argument is a valid philosophical one not amenable to your contention about people using CAM in addition to actual medicine.

  77. mousethatroared says:

    nybrgus – thanks for the narwhals links. they cracked me up.

  78. Quill says:

    Great article and a wide-range of fascinating comments. One thing is clear to me: science has made great strides in understanding this thing called “mind” but has a long way to go. Given the intense interest in this placebo effect and what exactly it means in medicine I am hopeful that once more understanding of the workings of the “mind” are discovered that it will not be a matter of lying but simply of providing accurate information that produces palliative results.

  79. pmoran says:

    In a single patient, using a placebo instead of antibiotic for a viral illness might reasonably be argued as being “less harmful.” Certainly less physically harmful. How about psychologically in the long term? What if your patient discovers you have been giving a placebo? Does psychological harm count?

    You should never “use a placebo instead of an antibiotic” for a simple viral illness. We want the public not to be seeking antibiotics for mild viral illnesses at all, and if there is any reason for antibiotics the patient should have them.

    What doctors might try to do with less understanding or more demanding patients (where there are risks from NOT prescribing) is to divert the patient onto anything else that has some public credibility at the time — there is always something: vitamin C, zinc etc. “Some people think X helps. Try that!’

    Through being too obsessive about EBM within the hurly burly of everyday medical practice we do deny ourselves some potetnially useful options. Any good doctor will already know which of his patients might be offended by being suggested something that is not supported to FDA standards. Smart patients wil also appreciate that the doctor is trying to protect them from overtreatment and drug side effects.

  80. BillyJoe says:

    “Some people think X helps. Try that!”

    I will never understand why anyone would want to say something like that.

    Some people think vitamin C helps. Try that.
    Aren’t you just promoting a myth that you should be trying to correct?
    Some people think Noni juice helps. Try that.
    Why would you promote something someone has simply pulled out of their nether regions?

    In short, aren’t you just working against what SBM stands for?

  81. PJLandis says:

    This a interesting story…http://whitecoatunderground.com/

  82. daedalus2u says:

    The origin of the “first do no harm” admonition was in the pre-scientific era where there was very little understanding of how physiology or treatments given by health care practitioners worked. A very large risk to patients at the time was iatrogenic injury from health care practitioners who didn’t know what they were doing and injured the patient by doing too much. The compulsion to help, the compulsion to do something is what drove many people into medicine, it can drive those same compassionate people into giving bad care.

    The extended version of the “pledge”, also promises to never do surgery for a kidney stone. The reason was because surgery to remove kidney stones would not be successful 2 millennia ago. Even as recently as a couple of centuries ago, people with typhoid treated with homeopathy did better than what was the “standard of care” of the time. It is not that homeopathy provides any benefit, but that the treatments of the time were worse than doing nothing.

    That is what the admonition “first do no harm”, means, it means don’t do something that will make things worse than doing nothing.

    Physicians still do things that are worse than doing nothing. Giving an antibiotic for a viral infection doesn’t help, and does increase the population of antibiotic resistant bacteria in the patient. Giving a placebo instead of an antibiotic is less harmful, but it is not less harmful than doing nothing. Giving coffee enemas and toxin cleanses is worse than doing nothing.

    It may be easier to give a placebo or an antibiotic to a demanding patient, and also more lucrative. No doubt there were patients with kidney stones who demanded their removal. The “standard of care” has been for millennia, to “first do no harm”. To me, that precludes giving a placebo (unless you can invoke the placebo effect pharmacologically ;) ).

  83. nybgrus says:

    We want the public not to be seeking antibiotics for mild viral illnesses at all, and if there is any reason for antibiotics the patient should have them.

    But we do want the public seeking vitamin C and echinacea for cold and flu? Or zinc for sore throats?

    …is to divert the patient onto anything else that has some public credibility at the time…

    So if I could show you a good study/poll that demonstrated that antibiotic for viral illness is something that the public find credible and desirous, that would then make it OK do so?

    Any good doctor will already know which of his patients might be offended by being suggested something that is not supported to FDA standards

    Right. Because no doctor ever sees brand new patients or even regular patients just occasionally. Nor do patient views and thoughts change over time. So lets further focus in the meaningless response by only looking at those doctors who both have been practicing for some length of time and whose patient population is comprised of only regular patients whom (s)he knows well.

    Smart patients wil also appreciate that the doctor is trying to protect them from overtreatment and drug side effects.

    Because clearly they are smart enough to appreciate that, but only if you distract them with some other shiny bit of woo. Because they aren’t smart enough to appreciate the same thing by explaining to them why you recommend no treatment.

    Nice job not actually addressing the substance of my comment and further espousing this nebulous double standard that only you can suss out the particulars of.

    I can do it to:

    You said simple and mild viral illness. Does that mean it is OK for complex, moderate, or severe viral illness? Is that just prophylactically in case it might predispose to bacterial superinfection?

    See how easy it is to focus on one little bit instead of actually addressing the point of a post?

  84. Diane Jacobs says:

    Hi there, a bit late to the conversation.
    Benedetti defines placebo as follows:
    “Placebos are not inert substances, as thus far believed. They are made of words and rituals, symbols, and meanings, and all these elements are active in shaping the patient’s brain. Inert substances, such as saline solution, have long been used in clinical trials and double-blind randomized protocols to assess the efficacy of new therapies, for example, new pharmacological agents. Although inert substances are of great validity in the clinical trial setting, the clinical trialists have always drawn their attention on the inert substance itself, thus diverting it from the real meaning of placebo (Moerman, 2002). The placebo is not the inert substance alone, but rather its administration within a set of sensory and social stimuli that tell the patient that a beneficial treatment is being given. If drawing attention to the inert substance is correct in pragmatic clinical trials, in which the only purpose is to see whether drugs are better than placebos, this surely does not help understand what a placebo is (Benedetti, 2008a).” From “How Placebos Change the Patient’s Brain” – http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3055515/

    Diane Jacobs
    Human Primate Social Groomer and Physical Therapist, general all-round placeboic and descending pain modulation inducer, Communications Liaison for The Pain Science Division of the Canadian Physiotherapy Association.

Comments are closed.