The Mysterious Healing Power Of Placebos Placebos, or sugar pills, are one of the building blocks of good science. For decades they've played a major role in medical experiments and drug trials. In some cases, placebos work as well as or even better than the real treatments -- for things like pain, nausea, depression and Parkinson's.
NPR logo

The Mysterious Healing Power Of Placebos

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
The Mysterious Healing Power Of Placebos

The Mysterious Healing Power Of Placebos

The Mysterious Healing Power Of Placebos

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

Placebos, or sugar pills, are one of the building blocks of good science. For decades they've played a major role in medical experiments and drug trials. In some cases, placebos work as well as or even better than the real treatments — for things like pain, nausea, depression and Parkinson's.


This is TALK OF THE NATION. Im Jennifer Ludden in Washington.

You know that feeling. You feel bad. You don't know why. You hope the doctor will do something, anything, to make you feel better. You get your prescription, you fill it, and amazingly, you do feel better.

But what if this were all in your head? What if the doctor had given you a sugar pill, a placebo? It turns out the sugar pill can be pretty powerful. Some patients do experience an improvement in health when they are on a placebo, despite the fact that what they're receiving is essentially fake medicine.

It's called the placebo effect, and some of these sham treatments produce such good results, some doctors are wondering if the placebo itself should be a treatment in its own right.

We'll talk to doctors today about their experience with placebos and how the placebo effect has been changing over time. We want to hear from you, as well. If you're a medical professional, tell us whether you've had experience with the placebo effect. Do you wish you could prescribe placebos? To potential patients out there, would you want one?

Our number here in Washington is 800-989-8255. Email us at Or you can join the conversation on our website. Go to, and click on TALK OF THE NATION.

Later in the hour, Afghan President Hamid Karzai gets the star treatment in Washington this week. NPR's Jackie Northam will join us to talk about the change in tone.

But first, can placebos cure? Dr. Joshua Ammerman is a neurosurgeon here in D.C. at Washington Neurological Associates, and joins us on the phone from his office today. Welcome back to the program.

Dr. JOSHUA AMMERMAN (Neurosurgeon, Washington Neurological Associates): Thank you.

LUDDEN: First, Dr. Ammerman, tell us: What is your experience with placebos?

Dr. AMMERMAN: Well, as you've very, very well illustrated, placebos or the placebo effect is commonly associated with the use of medications. As a neurosurgeon, I don't spend a lot of time in drug trials, but I commonly see the placebo-type effect in two groups of patients, the first off being a patient who comes to see me with complaints, symptoms and no diagnosis.

They may come to my office with headaches or back pain, leg or arm pain, and they don't know why. And so we begin to evaluate that patient. We examine them. We talk to them. We order diagnostic studies, and ultimately we establish a diagnosis.

And it is remarkable to me how often simply giving a patient a diagnosis makes them feel better. Their headaches go away. Their arm pain goes away. Sometimes they even get stronger. I've done nothing to intervene in the underlying problem, but that psychological benefit leads to a physiologic or medical benefit in that patient that we can see.

LUDDEN: That's funny, because you'd think if you're told you have something, it might actually make you feel worse.

Dr. AMMERMAN: You'd be surprised how many folks feel just so much better. The anxiety of not knowing what's going on leads to a significant benefit for that group of patients.

The other place I commonly see it is a patient who does have a diagnosis and comes to me for treatment, and I'm going to use the example that many of your listeners will be familiar with, and that's a herniated disc of the spine. A herniated disc, as many folks know, is a piece of a cushion of the spine that chips off and can put pressure against the nerve.


Dr. AMMERMAN: Yes, it certainly can be, and it puts pressure against the nerve in the neck. Symptoms may develop in an arm, if it happens in the low back symptoms in the leg, and that's a frequent scenario that we deal with.

We also know that if you take 10 folks, and you treat them all conservatively, meaning without surgery, about seven of them are going to get better on their own.

So we prefer that route if possible, and one of the mainstays of conservative treatment is physical therapy, and it's remarkable. I'll get phone calls from patients who have been to PT, physical therapy, for one or two sessions, and they already feel markedly better. And I know a lot of good physical therapists, but nobody works that fast, and it's quite clear that the patient is deriving significant psychological and therefore physical benefit from now being engaged in treatment. I'm doing something about my problem.

LUDDEN: So is this are you saying it's in our brain? Our brain can heal the body? Is that is it as simple as that?

Dr. AMMERMAN: Well, I don't know if it's as simple as that, but there is a clear physical benefit to settling down one's psychological stresses, whether it is not knowing what's going on or not having anything to do about it that folks across the board benefit from. We see this time and time again.

LUDDEN: Now, do you think, then, that this placebo, the power of placebos, should be turned into a treatment in and of itself?

Dr. AMMERMAN: Well, we commonly already do that. And in the scenario I just described to you, I know that if I take those seven patients who are going to get better with conservative management and did nothing with many of them, sort of sent them home and said, you know, basically behave for a while, a bunch of them would get better on their own.

But we as humans find it very psychologically difficult to do nothing about a problem, and a lot of conservative treatment, while it certainly makes folks feel better, it's not what makes the disc fragment go away from the nerve. That's with the passage of time. But folks derive a whole bunch of benefit from being actively engaged in treating their problem.

LUDDEN: More so than if they just did, if they chilled out and went home and did nothing?

Dr. AMMERMAN: Very few people do well in that scenario. If you try that with most patients, they're pretty unhappy. They want to be actively fixing their problem. It's like you go home, you see a hole in your wall. Most people won't just let it sit there. They want to do something about it, and your body's no different.

LUDDEN: So, I mean, you're maybe you're not in the business of prescribing medicine. You're more neurological issues. I don't know. Do you think, though, there should be, you know, prescriptions for sugar pills?

Dr. AMMERMAN: Sure. I think the answer is probably yes in a limited role. I would be uncomfortable, for example, if a patient came to me, and among their problems was high blood pressure - I would be uncomfortable providing them a medication that I did not, could not guarantee a benefit for that type of a problem.

However, other conditions, for example dealing with pain. Okay, you know untreated elevated blood pressure can be a killer in folks. For dealing with chronic pain, there may be a role for that. For dealing with some psychiatric illnesses, there certainly may be a role for that. So I think if used appropriately in the narrow area, it would be a very reasonable approach.

LUDDEN: All right, Dr. Ammerman, we know you've got to go very soon. We're going to get one quick call in for you before we say goodbye.


LUDDEN: Christopher(ph) is in San Antonio, Texas. Hi there.

CHRISTOPHER (Caller): Hi, thanks for having me. I'd just like to comment. I've always I've never been prescribed a placebo. However, I'm very into holistic healing. And I don't know if it actually has any credence, but in my experience, when I, you know, have a crystal in my pocket that I'm told has this certain property, I feel like it really helps me, and I'd just be interested to hear your thoughts on that in being an active placebo of sorts with the crystals.

LUDDEN: Okay, Dr. Ammerman, crystals?

Dr. AMMERMAN: Whether it's crystals, or whether it's medication or whether it's a test, I think it all goes to the same point - that we as human beings like to be actively engaged in doing something about a problem, and if that works for you, clearly I shouldn't say clearly I don't know that there is a physical benefit that having that crystal in your pocket causes for you, but clearly the psychological benefit leads to a physical change for you. And if that works, we should be doing that, and we should be encouraging that.

LUDDEN: But if you know it's a placebo, does it still work, or do you have to think it's real?

Dr. AMMERMAN: Well, by definition for it to be a placebo, you can't know that it's a placebo. In other words, placebo implies that you, as the one receiving the treatment, doesn't know that this is not supposed to cause a benefit. It requires what we call blinding.

So it's no longer a placebo. So he believes, our caller believes, that he derives benefit from this, and therefore he does. If he doesn't believe he's going to derive any benefit from this crystal, and I have no medical data to support that it would be helpful, then it's no longer a placebo issue.

ROBERTS: All right, well, Dr. Josh Ammerman, thank you so much for your time.

Dr. AMMERMAN: It's been a pleasure.

ROBERTS: Dr. Ammerman is a neurosurgeon at Washington Neurological Associate and joined us by phone from his office in Washington, D.C.

Now we're joined by Dr. Arthur Barsky. He is director of psychiatric research at Brigham and Women's Hospital in Boston, and a professor of psychiatry at Harvard Medical School, and he joins us from a studio at the medical school. Welcome to you.

Dr. ARTHUR BARSKY (Director of Psychiatric Research, Brigham and Women's Hospital; Professor of Psychiatry, Harvard Medical School): Well, thanks very much.

LUDDEN: Can you give us a little placebo history here? How were they first used?

Dr. BARSKY: Well, it's quite interesting, because if you think about it, until relatively recently, the vast majority of what doctors did was offer placebos. That is, most of what was done in the past, in centuries past, actually wasn't specifically curative.

LUDDEN: Because they had no idea what would actually cure, is that what you're saying?

Dr. BARSKY: Yeah, you know, they didn't exactly. They didn't know what the disease process was. They didn't have specific antibiotics or specific surgical procedures, and the things that patients were told to do actually didn't have a specific curative effect, and yet a lot of people got better for a long over many centuries.

LUDDEN: But did the doctors believe what they were doing would make the person feel better?

Dr. BARSKY: Yeah, that's the key point. And that was made earlier by Dr. Ammerman, that part of the placebo phenomenon is that the patient and the physician believe that, in fact, this is going to be helpful. This idea, the positive expectation that you're going to benefit from this interaction is a very important part of it, and so...

LUDDEN: And so they've actually go ahead.

Dr. BARSKY: No, no, and I was just going to say, so in the past, centuries past, when we did this, yeah, the physicians or healers and patients had very powerful beliefs that it was going to work.

LUDDEN: So placebos have actually come to be a very common they're used as kind of a neutral placeholder now. I mean, research on a drug isn't thought to be valid unless you've got a placebo in there. Is that right?

Dr. BARSKY: Yeah, because when you give an active pill - what we believe to be an active pill to someone, you're really getting two benefits. One is a biochemical benefit - that is, the composition of the pill is actually acting against the pathology in your body.

But the other part of it is this placebo phenomenon, the kind of non-specific benefits that come with that: encouragement, support, positive expectation that you're going to get better, the sense that someone is doing something to help you. That's all bound up in the intervention.

So there are really two elements to the prescription. One is the actually chemical composition of the pill, and the other is everything that goes along with it.

LUDDEN: And can you just fill us in briefly. There are studies showing that the effect of this placebo is actually increasing the time. Is that right?

Dr. BARSKY: Yeah, it's very interesting phenomenon. What appears to be happening is that the placebo effect, at least with certain medication, compared to certain active medicines, has been getting has been improving over time, particularly over the last 20, 25 years. People are reporting more benefit from placebo than they used to.

LUDDEN: We're talking with Dr. Arthur Barsky at Harvard Medical School about the placebo effect and whether or not it can be used as real medicine. If you're a medical professional, tell us whether you've had experience with the placebo effect, and potential patients, would you want one? 800-989-8255 is our phone number. The email is I'm Jennifer Ludden. It's TALK OF THE NATION from NPR News.

(Soundbite of music)

LUDDEN: This is TALK OF THE NATION. Im Jennifer Ludden in Washington.

We know what a placebo is, a sugar pill or some other fake treatment. Today, we're talking about what happens when those phone treatments work. In some cases, placebos work as well as or even better than the real thing, and for things like pain, nausea, even depression and Parkinson's disease.

Dr. Arthur Barsky is with us. He's professor of psychiatry at Harvard Medical School and director of psychiatric research at Brigham and Women's Hospital in Boston.

We also want to hear from you. If you're a medical professional, tell us whether you've had experience with the placebo effect. Do you wish you could prescribe placebos? Our number here in Washington is 800-989-8255. Our email address,, or you can join the conversation at our website. Go to, and click on TALK OF THE NATION.

Dr. Barsky, we're going to take a call now from Jenny(ph) in Roseburg, Oregon. Hi, Jenny.

JENNY (Caller): Hi. So speaking to how doctors can ethnically prescribe an earlier caller spoke about alternative treatments. I read an article, I think it was in Scientific American a while ago, it was a doctor who, seeing that maybe 30 percent of patients were overcoming depression with placebo in studies, started to tell his patients, for example, you know - I could prescribe you these drugs, but maybe before we go in that direction, there are studies that show that about 30 percent or maybe the doctor can fill in the percentage have had very good results for their depression from taking St. John's Wort tea in the morning and valerian in the evening. And I'd like you to try that for a month, and then we'll get together and see how it's going for you.

So the patient's not being lied to. The doctor's speaking from real studies, where the placebo effect is effective and giving the patient something to do and having the follow-up and so that all the pieces are in place.

And I just always want to mention in London, there is a homeopathic hospital run by the National Health Service, and there's been discussion about whether it should be funded, but again it's, like, maybe it works, and maybe it doesn't, but it's being you know, the care that's delivered there is delivered there with every belief that it works, and the patients get good effect at lower cost.

LUDDEN: All right, Jenny, thank you, thank you so much for calling. Dr. Barsky, homeopathy as placebo, what about that?

Dr. BARSKY: Well, I think Jenny has put her finger on the really critical question, which is: How can we harness the placebo effect in clinical practice? And no one really knows. What happens if you were to inform the patient that we're going to start out with an inert pill? And that helps a lot of people. Let's see if it helps you.

The question is: If the patient was actually told it was a placebo, would that rob the effect, or would that still work? And the whole question of how to try to integrate placebo prescribing into medical practice is really a very interesting one.

People have suggested it's possible that we could begin treating somebody with active drug and then gradually, over time, let them know that we may begin to increase the amount of placebo in the active pill. So essentially, we're weaning them off the active medicine and substituting an increasing amount of placebo.

And there are some experimental data that we would still get the same effect in that way.

LUDDEN: But they wouldn't know when that was happening.

Dr. BARSKY: That's correct, and ultimately, then, that would allow us to prescribe less active medicine over the whole course of the illness without really deceiving the patient in any way.

LUDDEN: We have an email here from a doctor in Oregon. He says: I've given placebos in my nursing practice and observe that the majority of patients receiving them do get the response they're seeking. Indeed, they sometimes rely on it and make favorable comments on the help it gives them.

None I know of were aware of this. But what happens in the case of those few patients who have reactions or adverse effects from the pill? I've heard of this occurring, though I have no personal experience of it. What about that?

Dr. BARSKY: That's another fascinating aspect of this whole problem, which is that about 25 percent of people will develop side effect excuse me will develop side effects on placebos. They will report headaches and dizziness, trouble concentrating...

LUDDEN: From a sugar pill? Really, from...?

Dr. BARSKY: From a sugar pill.

LUDDEN: I mean, is that psychological?

Dr. BARSKY: Well, the way we understand that, is that these are rather common symptoms in daily life and that you may often have them periodically but really not notice it. And once you're put on a medication, you then say oh, that must be due to the pill. That's a secondary effect of the medicine. You re-attribute a long-standing symptom to the medicine that you've just been placed on.

LUDDEN: So what about, though, the ethics is there any regulation of this? You know, what ethics are at play here?

Dr. BARSKY: Well, the ethics that we're most concerned about is this idea that in some way, we're deceiving our patients, that it's a betrayal of the relationship with the patient to prescribe something that you, that the patient believes is an active drug that in fact is not.

And that's really, then, a real issue that we've not solved yet in American medicine. My understanding is there's a little less concern about that in some European countries, where they're a little less reluctant to prescribe placebos initially and not and for patients to believe that in fact they're getting the active drug.

But here, that's been a real obstacle. There is this real sense that it's somehow profoundly unethical to write the prescription.

LUDDEN: And here's my question also: Don't you have to pay for it? If it's a sham, and you're trying to convince them it's real medicine, don't you have to send them down to the drugstore, and they have to pay for something that's not really medicine? How does it work?

Dr. BARSKY: Good question. Yup, that's right. You know, typically, people have used things like vitamin pills as placebos, but then obviously the patient knows.

In studies, which we referred to earlier, we do what's called blinding. That is, the placebo pill looks exactly like the active medicine, and the doctor who's prescribing it doesn't know which whether he's giving a placebo or an active medicine, and the patient doesn't know, either, but that's in a study situation. That's not in clinical practice.

LUDDEN: All right. Let's take another call. Melissa is in Charlotte, North Carolina. Hi, Melissa, how are you?

MELISSA (Caller): Hi, I'm good. I was just a little concerned with your earlier caller. I work in long-term health care, a lot of psychiatric stuff in my building, and he did say placebos could be used to treat psychiatric disorders, but I just wanted your guest to be a little more clear and say they can't be used for things like schizophrenia and bipolar disorder.

LUDDEN: Good point. Good point. Dr. Barsky, can you give a sense, what types of conditions do placebos, just would not work for, and what are some conditions they would be most likely to be helpful in?

Dr. BARSKY: Well, there's an important distinction we want to make, and that is whether the placebo is active against symptoms, against sensations that people are reporting and whether they're active against the actual, underlying disease process, whether they actually change the body's physiology or the body's anatomy.

And clearly, they are more effective in altering symptoms. So chronic pain symptoms, things like that seem to really respond more. But there are, in fact, psychological, real if you will, diseases like asthma and high blood pressure and Parkinson's that do respond to placebo, as well. So you have to really distinguish whether you're talking about response in symptom improvement or response in the actual, underlying disease process.

LUDDEN: But are there some diseases where this is just not there's no question of using a placebo?

Dr. BARSKY: Well, I mean, if you think about the most extreme examples, I haven't heard about a placebo response to cancer. I'm not sure I'd want to treat a lung cancer with a placebo.

LUDDEN: Right.

Dr. BARSKY: So there are clearly some really very serious, severe, major medical illnesses where we have very little reason to believe that the placebo would be helpful and where there are already established treatments, it's probably not ethical to prescribe placebo.

LUDDEN: All right, let's have another call. Roger(ph) is in Cincinnati, Ohio. Hi, Roger.

ROGER (Caller): Hi. Yes, I'm not a medical professional, but I just thought that I should share this. My grandfather, who passed away many years ago, attributed well, let me first say my grandfather was not literate. He knew nothing about science, couldn't read and write, and he attributed the effectiveness of medication to the color. So if you gave him aspirin, that was white in color. He would look at it to be something that is not effective, but if you gave the same aspirin, same dosage, red color or green or any other color, he thought that one was more potent, more powerful and more effective.

LUDDEN: And do you know what gave him that sense?

ROGER: I don't know why. I thought I think I'm not very sure why it was that way, but if he did if you gave him any color medication, he would say he felt better with that, and that, you know, if you brought him the white one, he didn't think it would help him.

LUDDEN: All right, Roger, thanks for calling. Dr. Barsky, something as subtle as color, can that make a difference for someone?

Dr. BARSKY: Yeah, I mean, Roger's pointing to a really fascinating aspect of this whole thing, which is the nature of the placebo itself does seem to have an effect, and one of the things that's pretty clear from a lot of drug company studies, because they do a lot of testing of what color the pill ought to be, is for example, blue and green pills tend to be perceived more as tranquilizing and reducing of anxiety, and red, orange pills seem to be thought of more as altering, activating.

So the color of the pill does seem to make a difference and then of course the nature of the intervention. Prescribing pills more frequently seems to have a more powerful placebo effect than infrequently. Injections or procedures, more major procedures, seem to have more placebo effect than simply taking a pill. So the nature of the intervention, whether it's a medicine or a procedure, and the nature of the medicine itself, actually do seem to affect the placebo response.

LUDDEN: Hmm. We have an email here I'd like to read to you from someone who's concerned about this trend. The person writes: As a patient who's been recently diagnosed with Lyme disease and Coeliac disease, I would hope that if the placebo effect were used, it would only be permitted with mild conditions that have already been diagnosed. I went to more than 25 doctors before it was finally discovered that I had Lyme and Coeliac. Many of these doctors tried pushing anti-anxiety drugs on me as part of treating what they called fibromyalgia. I'm not sure I'm saying that right. These doctors never took a substantive look at my symptoms or tried to get to the bottom of things.

I'm worried if this is allowed to go into standard practice that it will advance this trend among doctors, that rather than actually getting to the bottom of complex problems, doctors will just issue a patient a sugar pill assuming they're a hypochondriac.

I can sympathize with that. But what do you think, Dr. Barsky?

Dr. BARSKY: Well, I think you've got to bear in mind two really important distinctions, which I think most physicians, most responsible physicians would really make, and that is the diagnostic process is divorced from the therapeutic process. You first have to establish the diagnosis, what it is you're treating.

LUDDEN: But you did say that a lot of the people who respond to placebos come in and they have a condition that has not been diagnosed. Or maybe that was our previous guest said that.

Dr. BARSKY: Yeah. I think so. But the - you know, there - if in fact we can't find out what's wrong, and there are lots of situations like that - chronic fatigue and dizziness and insomnia - there are a lot conditions where we really can't establish a medical diagnosis. Now, that's a different issue.

But we're talking about, first - and those are by exclusion, what we would say is diagnosis by exclusion - first you have to look carefully for all the known medical explanations. And if there is an accepted treatment that has demonstrable benefit, then obviously that's what we go for.

LUDDEN: I'm curious. What are the implications of this for the drug industry?

Dr. BARSKY: Well, it's really - it cuts two ways, actually. It's a two-edged sword for the drug industry because on the one hand it's a big problem for them. They are trying to sort out how much of their medication is actually helpful, specifically, and how much of it is this kind of non-specific effect of support and encouragement and positive expectation. So it's actually rather difficult for them.

And in conditions where there's a high rate of response to placebo, for example, anxiety, depression, it really becomes difficult to show that your active medicine is better. Since people are going - a high proportion of patients are going to respond to the placebo, you really have to have a very powerful medicine for it to have a statistically significant benefit that's over and above the benefit that you're getting from the placebo.

LUDDEN: We're talking with Dr. Arthur Barsky of Harvard Medical School about the placebo effect.

You're listening to TALK OF THE NATION from NPR News.

Let's take a call now from Selena(ph) who is in Kansas, Missouri. Hi, Selena.

SELENA (Caller): Hi. It's Kansas City. Kansas City, Missouri.

LUDDEN: Sorry about that. Welcome.

SELENA: Thanks. I just am concerned - I'm all for natural healing and the power of the body and the mind. But I'm concerned about these types of studies advancing faster than support - than helping people who are being told they should pull themselves up but their bootstraps, they should work a little harder with their mind, that that trend might actually outpace some of the true studies of whether these placebos are really effective or not - making people think that they, you know, kind of that, you know, continuing the myths that people should work a little bit harder from the inside out or use their mind, use their brains and their body symmetry and all that to get better.

LUDDEN: All right. Well, thanks for calling. Dr. Barsky?

Dr. BARSKY: Well, I think what you may be getting at is a real problem, and that is that somehow we kind of feel as if, if your symptom gets better with a sugar pill, then it wasn't real, that you were somehow imagining it, it was all in your head. And that's not true. The symptoms are real. These patients are not lying. They're not making it up. And the fact that it responds to a placebo does not in any way mean that the symptoms aren't real.

LUDDEN: All right. We have time for one more call. Charles is in Virginia Beach, Virginia. Welcome, Charles.

CHARLES (Caller): Hi, thank you for taking my call. I was just curious - as we begin broadening public awareness, placebos and the effect of placebos, could there possibly be an unintended consequence of breeding distrust of medications that we're taking and then lowering the overall impact of all medications? And what would the result of that be?

LUDDEN: All right. Thanks for calling. Dr. Barsky?

Dr. BARSKY: Great question. And it comes back to the issue we spoke about earlier, which is how much does knowing that the placebo phenomenon is occurring, how much does that really erode it? How much does that impair it? How much does the respond, the positive response to a placebo depend upon your belief that it's not a placebo, that you believe that it's an active medicine? Would the placebo phenomenon occur if you knew it was a placebo? That's kind of the question you're asking at a population level, as a country as more and more people are aware of this phenomenon. Will it begin to erode the magnitude of the phenomenon? It's a fascinating question.

LUDDEN: Well, it also - I think he asked about, you know, how we look at other medication, right? If a placebo can work for this, couldn't it work for that?

Dr. BARSKY: Yeah. You know, as we were saying earlier, it really - much of this is unknown. We really don't know clearly exactly which conditions are going to respond best to placebo and which conditions don't respond very well. There's a whole lot about the phenomenon we don't know. We don't even know whether the same person, for example, will have the same placebo response over time. We don't know how long it lasts. The placebo helps your headache tonight. Will it help it every night for the next month or will the phenomenon peter out over time?

LUDDEN: And we should note there have been some critics who say that kind of the effect of it has really been overblown, that some of the studies may be -isn't there a bit of controversy about the fact that - the thought that there's a growing impact of placebos?

Dr. BARSKY: Well, I think the big controversy is people have said you don't really know what the natural history of the condition is that you're treating. So for example, to take that example of a headache that responds to placebo, we don't know what would have happened to that headache if we didn't do anything at all. We're assuming that the headache would have not have gone away if you hadnt taken the placebo. But to really determine what's going on, you really need three groups. You need people who get an active analgesic for headaches, you need people who get a placebo for headaches, and then you get people - you need people who get nothing at all. And then we could compare the response to placebo to what we would call the natural history of the condition, whether it gets better on its own with no intervention whatsoever.

LUDDEN: All right. Well, next time I get sick, I think I'm going to try doing nothing but tell myself I took a pill.

Dr. Arthur Barsky is the director of psychiatric research at Brigham and Women's Hospital in Boston. He's a professor of psychiatry at Harvard Medical School, and he joined us from Boston. Thank you so much.

Dr. BARSKY: Thank you very much.

Copyright © 2010 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.