Take two tic tacs and call me in the morning

The other day, Kim asked me if I’d ever consider prescribing a placebo.  I replied that I would not, as I would consider doing so unethical.

In a lovely piece of symmetry, I came across an article on that very subject in The New Yorker a few days ago, probably from the issue that was current at the time of Kim’s question.  Titled “The Power of Nothing.” Michael Specter’s piece focuses on Ted Kaptchuk, a professor at Harvard Medical School who has spent years studying the placebo effect, and who clearly believes that there is value in exploring the use of placebos in medical practice.

Unfortunately, the article has already been archived in the online version of the magazine, and their archiving system makes it nigh impossible to cut and paste quotes.  For those of you who are interested enough in the article to read it, it’s their “Annals of Science” article from the December 12, 2011 issue, and starts on page 30 in the paper version.

As far as I am aware, there is really no established ethic for prescribing a placebo.  To the contrary, doing so conflicts with several of the ethical principles I was taught in medical school.  Prescribing a placebo generally indicates a certain degree of deception, giving the patient the impression that you believe the medication to have a pharmaceutical effect that it does not.  This is pretty obviously paternalistic.  While a certain degree of paternalism in medicine is probably unavoidable given the information and power differential within the physician-patient relationship, and one could argue that certain circumstances warrant a limited amount of benign paternalism, in general paternalism is something to be avoided or minimized.  Misrepresenting the benefits of a prescription also makes genuine patient autonomy impossible, as the patient is making a decision to take the medication without correct information about what it will (or won’t) do for them.

This isn’t to say that I don’t sometimes prescribe medications that I don’t think are necessary.  A common example is prescribing antibiotic drops or ointment for conjunctivitis (“pinkeye”), an illness that is overwhelming due to viral infections and is therefore self-limiting.  I’ve heard from ophthalmologists that the number of cases of bacterial conjunctivitis (the kind that would respond to and genuinely warrant an antibiotic) that they’ve seen in their careers number on one hand.  That said, pediatricians in general are pretty liberal with the eye drops.  This is, in part, because daycare centers treat conjunctivitis like ebola for some reason, and won’t let kids back in without a prescription.  Many providers justify medicating by arguing that putting kids on an antibiotic lessens the risk of bacterial infection, which is at least plausible.

What keep these drops (or, if you prefer, a prescription for an oral antibiotic that I have been mau-maued into writing) from being an outright placebo are two things.  First, they have an actual pharmaceutical effect, which may confer come benefit, even if I don’t really think it will be significant.  It’s possible that I’m simply wrong about the diagnosis, and am covering my bases (certainly some patients would likely argue so).  But moreover, whenever I write a prescription whose value I question, I say so explicitly.  I tell parents the drops are almost certainly more of a hassle than they’re worth (you try getting a drop onto the surface of a struggling toddler’s eyeball), though otherwise benign, and they don’t have to fill the prescription if they’re disinclined to do so.    It’s conceptually quite different from gussying up some saline solution and telling parents it will be curative in a few days’ time.

An interesting aspect of the placebo effect seems to be that some patients experience it even when they know they’re taking a placebo.  Some part of the therapeutic process helps alleviate illness, including in some cases taking a medication known to be physiologically inert.  If there were to be a developed ethic of placebo use, I cannot see how to make it square without honesty.  Arguing from a quality of life perspective still assumes too much paternalistic privilege for me to feel comfortable about deluding a patient.  If a body of evidence develops to support the use of an openly-prescribed placebo for relief of certain symptoms, I would at least admit the possibility of prescribing them in select circumstances.

Russell Saunders

Russell Saunders is the ridiculously flimsy pseudonym of a pediatrician in New England. He has a husband, three sons, daughter, cat and dog, though not in that order. He enjoys reading, running and cooking. He can be contacted at blindeddoc using his Gmail account. Twitter types can follow him @russellsaunder1.


  1. An interesting aspect of the placebo effect seems to be that some patients experience it even when they know they’re taking a placebo. Some part of the therapeutic process helps alleviate illness, including in some cases taking a medication known to be physiologically inert.

    This has always fascinated me. You could write a scrip that says “Placebo 25 mg administer orally PRN” and the pharmacist would fill it with Rolaids, charge the copay for them, and your objectively-healthy but subjectively-in-pain patient would feel better.

    I bet that you could even give the patient the scrip saying, “I’m prescribing you a which will have no medical effect on you whatsoever,” and the pharmacist could take the Rolaids out of the $2.50 over-the-counter container in front of the patient and rebottle them in a the “medication” bottle with the adult-proof lid in plain view, and charge $20 for them.

    I’d go so far as to bet that in a statistically significant number of cases, the patient wouldn’t even have to actually ingest the medication in order to percieve a palliative effect.

    What seems to matter is one or both of a) the process of seeing a doctor and filling the prescription at the pharmacy, and b) paying money. This defies all logic and rationality, yet I have absolute faith that there are substantial numbers of patients who would get this treatment and report that they felt better afterwards.

    • I agree with this. There is likely something in the experience of being cared for, in taking concrete steps to address a problem that lessens its severity.

      • Trying not to be a hypochondriac with the internet around is difficult… but at least now I know what “that’s emergency room material” is…
        There’s certainly some comfort in “someone listened” and some additional comfort in “I’m not missing something crazy bad for me, just cause I hurt…”

    • What’s funny is that I first read this paragraph as “…some parents experience it even when they know they’re taking a placebo…” and didn’t realize my mistake until I read it again in Burt’s quote. It made me think of how often parents (and I include myself) rush their sick children off to the pediatrician ultimately seeking treatment for something which will pass on it’s own in time, all really just to make themselves feel better.

    • I’ve only looked at one study where the placebo effect worked on patients aware they were receiving placebos, but it did involving ensuring that the patients were aware of the placebo effect, and thus presumably expected the pills to work for that reason.

  2. To dumb down the discussion for me, I’d be curious as to why a placebo is necessarily unethical. I can see why in a lot of cases it would be, but the example that comes to my mind is during winter months, when doctors perpetually cave and OK anti-biotics to people suffering from viral infections, for no other reason than the patients won’t let up about it.

    In these cases, would tic tags not be more ethical than giving them a prescription drug they do not need? (Obviously, the best alternative is to send them home and tell them to rest and drink fluids, but I seriously see a lot of antibiotics for colds being given out there.)

    • It depends whether you think what is ethical should be weighed in terms of consequences (benefits v. harms) or in terms of respect for persons. A majority of ethicists, but certainly not a huge majority, think it’s respect for persons. (So even though killing you and harvesting your organs to save 5 lives of people waiting for organs may make more people happier and healthier, we don’t do that because it doesn’t respect you as a person).

      So respect for persons (as opposed to just judging the best outcome) demands that you get their informed consent for medical procedures. And that that is more important than just creating the happiest, healthiest outcome.

      • I maintain that it’s perfectly legit to get someone’s consent for a battery of things, such that they might consent to being given a placebo. (just the same as someone might consent to “not being told” about death, or “being told in a roundabout way that’s socially acceptable to me”)

    • Essentially it boils down to this — placebos are a lie in pill form. The lie may be told with the best of intent, but it remains a lie. The argument would be that it is better to have an unhappy patient who is fully informed as to your best medical judgment than one you’ve patted on the head and said “take these and you’ll feel better” knowing all the while that they’re Necco wafers.

      The various nettlesome commenters I get who find me condescending or patronizing would doubtless stridently object to my deigning to write a prescription merely for the sake of placating my patients.

      • I read the New Yorker article last week. The whole question about the whether the FDA could ethically license a placebo seemed a bit forced.

        It would be really shortsighted of doctors to start prescribing FDA-certified placebos. Since the dawn of scientific medicine, physicians have been trying to dig themselves out of the PR hole dug for them by patent medicine salesmen and other quacks. Even today, a lot of patients already suspect, rightly or wrongly, that real drugs are just gussied up placebos.

        The article didn’t show that you get more placebo effect out of a fake pill than you do out of selling the therapeutic merits of something genuinely but mildly beneficial. In fact, if the placebo effect really is an outgrowth of the therapeutic ritual, it seems like any kind of medical attention could work.

        Doctors should do a better job of selling sound self-care advice as real medicine, as opposed to a consolation prize when there’s no drug for what you’ve got. So, instead of just suggesting offhandedly that a patient with low back pain might feel better after a hot bath, a doctor might try to harness an extra placebo effect by writing out the advice as a “doctor’s order” on an official looking prescription pad and present it with some gravity.

        We’d need trials to see if a little extra white coat theater boosts the effectiveness of run-of-the-mill self-care advice, but it seems like it might have all the benefits of a placebo with none of the ethical pitfalls.

  3. Actually, I think you as a pediatrician are in a better position to dispense placebos ethically than just about any kind of doctor. The whole concept of autonomy is predicated on respect for rationality, usually characterized as the ability to use reasons to set and meet ends. Children’s rationality is not fully developed. They don’t have a developed ability to weigh reasons, to introspect their own reasons, and their ends are rather famously oriented toward short-term rewards. They cannot act in their own interest. Autonomy simply cannot be the primary ethical consideration in dealing with children – otherwise, would any children get vaccinated or take their antibiotics or eat a not-primarily-cookie diet? They do not give their consent, and if we were respecting autonomy we would not press the issue. Of course, as children get older one should respect their autonomy to the degree possible. It’s a sliding scale. But if one were to actually treat children as fully autonomous (i.e., nothing can be done to them without their consent) one would actually be doing wrong by them. Informed consent is difficult with adults and impossible with children. In the case of children, especially very young ones, benificence should probably be ranked as a more important ethical consideration than autonomy.

    It seems ethical to me to give children something and say “Take this and it will probably make you feel better.” This is not a lie in the case of a placebo. (One should respect the parents’ autonomy and get their permission of course).

    I also wonder if phrasing a placebo effect in terms of effects on the brain (as opposed to talking about mental effects or effects on the mind) might not make it more effective. I mean, unless one is a dualist, one believes that a placebo does its work by affecting the brain. I’d be curious to see if informing a patient that the medicine will affect the way their brain processes information about their illness (as opposed to using the word “mind”) would be a way of being truthful and respectful of autonomy and perhaps yet more effective.

    • I have much less of a qualm about prescribing a placebo to a child out of concern for them as autonomous actors (for the very reasons you lay out). My reluctance to do so is that I’m not aware of good data that demonstrate the degree to which a placebo effect works on children.

  4. http://www.webmd.com/pain-management/news/20081023/50percent-of-doctors-give-fake-prescriptions
    … a lot of doctors feel otherwise, apparently.

    Personally, I think a person should be able to choose a paternalistic doctor, if that’s the best health care treatment for them (and for extreme hypochondriacs, it may very well be!).

    I prefer a doctor like Dr. S, because I like to know what I’m getting into. He’s more likely to tell me, “I don’t know, but here’s the odds:” and he’s more likely to offer me multiple treatment options, along with attendant risks.

Comments are closed.