Placebos

[Sceptical Punter]

I am not convinced that the episode on placebos tells the whole story. I found an interesting link on YouTube where Richard Dawkins interviews Nicholas Humphrey (professor of psychology at the London School of Economics). You will have to google the link as I can't post it yet. He claims that the placebo effect is stronger than the Skeptoid episode suggests and can actually make physiological changes to the body, increasing it's power to self-heal. Specifically, he mentions placebo surgery that has the power to cure an injured knee (though I wasn't quite clear whether this referred to a physiological or a psychological 'cure'). I also notice today that USA Today has an article in a similar vein.

It does leave open the question of whether a doctor should prescribe placebos and if so whether the patient should be aware that they are placebos. An interesting thing from the interview was that placebos work even if the subject knows that they are placebos (though perhaps not as well). This may be because all that is necessary for them to work is to believe that they will work and, as we know, placebos do work - at least for some ailments.

Incidentally, Brian says a couple of times in some of the earlier episodes (I am just getting round to listening to them) that early medicine had no beneficial effect. This is not strictly true. They probably had a strong placebo effect. It is likely that bloodletting had a very definite placebo effect - provided the patient survived it. The more dramatic the procedure the greater the placebo effect.

[Professor Yaffle]

I assume he is talking about this study wrt the knee surgery
http://content.nejm.org/cgi/content/full/347/2/81 (http://content.nejm.org/cgi/content/full/347/2/81)
- which is really just showing that a surgery which was thought to have an effect does no better than placebo. What we don't know is what the results would have been like in the absence of any treatment (including placebo). And also the outcomes seem to be pain related, which is a very subjective symptom which placebo can affect greatly. I don't know that there is anything similar actually showing an improvment in an objectively measured parameter in sham surgery versus no treatment.

[fls]

I am not convinced that the episode on placebos tells the whole story. I found an interesting link on YouTube where Richard Dawkins interviews Nicholas Humphrey (professor of psychology at the London School of Economics). You will have to google the link as I can't post it yet. He claims that the placebo effect is stronger than the Skeptoid episode suggests and can actually make physiological changes to the body, increasing it's power to self-heal. Specifically, he mentions placebo surgery that has the power to cure an injured knee (though I wasn't quite clear whether this referred to a physiological or a psychological 'cure'). I also notice today that USA Today has an article in a similar vein.

I haven't listened to the podcast, but I did watch the Dawkins/Humphrey interview a while ago. My comments on it were posted here:

http://forums.randi.org/showthread.php?postid=4316200#post4316200

Basically Humphreys was sloppy in distinguishing between 'works' in the sense that 'subjects will report small improvements in subjective symptoms' and 'works' in the sense that 'subjects will show changes in objective measures of disease and disability or improvements that are clinically useful in subjective symptoms'. It's really the latter we are interested in achieving, yet the former is all that placebo has accomplished.

It does leave open the question of whether a doctor should prescribe placebos and if so whether the patient should be aware that they are placebos. An interesting thing from the interview was that placebos work even if the subject knows that they are placebos (though perhaps not as well). This may be because all that is necessary for them to work is to believe that they will work and, as we know, placebos do work - at least for some ailments.

While there may be benefits associated with mental state and expectation, there's no particular reason to think that deceit is a useful way to go about it.

Linda

[fls]

I assume he is talking about this study wrt the knee surgery
http://content.nejm.org/cgi/content/full/347/2/81 (http://content.nejm.org/cgi/content/full/347/2/81)
- which is really just showing that a surgery which was thought to have an effect does no better than placebo. What we don't know is what the results would have been like in the absence of any treatment (including placebo). And also the outcomes seem to be pain related, which is a very subjective symptom which placebo can affect greatly. I don't know that there is anything similar actually showing an improvment in an objectively measured parameter in sham surgery versus no treatment.



It should be noted that had this been a homeopathy study, it would be reported as a positive study, showing that homeopathy improved physical functioning at one year compared to placebo.

Linda

[Sceptical Punter]

What's your view on the USA Today article? There isn't enough in the article for me to draw many conclusions, but I wondered if you had more details on the research it refers to? I should say I have no real predisposition here. I genuinely don't know how effective placebos could be.

Would you be in favour of prescribing placebos and being honest about it - if it worked? I guess there is some middle ground where the doctor could say - "take these, they will help you to get better" which may be true, but doesn't reveal the whole truth. If a patient asked how it works, the doctor would tell the truth.

The other point to make is that having doctors diagnose illnesses and prescribe placebos only where appropriate is infinitely preferable to having proponents of pseudo-treatments do the diagnosis, which is the bigger danger.

[fls]

What's your view on the USA Today article?

What article are you talking about? This one?

http://www.usatoday.com/news/health/2010-02-20-placebo-effect_N.htm

Linda

[Sceptical Punter]

Yes that's the one. (I will be able to post links soon at this rate ;-) )

[fls]

What's your view on the USA Today article? There isn't enough in the article for me to draw many conclusions, but I wondered if you had more details on the research it refers to? I should say I have no real predisposition here. I genuinely don't know how effective placebos could be.

I haven't read the Lancet article yet, but in the USA Today article, the authors make reference to research using Parkinson's drugs as an example, and this is incorrect. That research is based on conditioning (patients are conditioned to release dopamine in response to a stimulus). It gets used as an example of the placebo effect, but the only similarity seems to be in the use of a pill as the stimulus for conditioning. Placebo use in clinical trials is set up in such a way as to remove the possibility of conditioning. It may have more relevance in real life, but it would be very limited and couldn't provide a novel effect. You already have to have a drug which causes the same effect in order to condition your body to do so, and the effect from conditioning is weaker and unsustained compared to simply using that drug.

Would you be in favour of prescribing placebos and being honest about it - if it worked? I guess there is some middle ground where the doctor could say - "take these, they will help you to get better" which may be true, but doesn't reveal the whole truth. If a patient asked how it works, the doctor would tell the truth.

We've had this discussion before here. The minor effects you can get from placebo do not seem worth the bother compared to the potential harm to the doctor-patient relationship, regardless of whether the doctor technically is telling the truth (like when I was putting in my first IV and answered, "I really couldn't say" when the patient asked, "how many of these have you done?").

The other point to make is that having doctors diagnose illnesses and prescribe placebos only where appropriate is infinitely preferable to having proponents of pseudo-treatments do the diagnosis, which is the bigger danger.

Well, this goes back to what I said earlier. There's no particular reason to think that deceit is the only way to engage the trust of those who would otherwise be inclined to go to SCAM proponents.

Linda

[Belgian thought]

My take on this is that placebos should only be given if the issuer is 100% certain that no detrimental effect will be experienced by the patient. I can only state this from experience, which even writing about it, saddens me.

I have been suffering from depression with varying degrees of episodes since my teens and many years ago (in the early 80s) during my first year at University I had a very severe one. I went to the campus doctor, and naively thought that this University, having one of the best teaching hospitals in the country, would treat and take my symptoms seriously.

I now realise that when the doctor went to his cabinet and gave me a bottle of red pills to take for 4 weeks and to come back for more, that he had simply given me placebos. I dutifully took them, getting no better, took them again until by the end of the first year I had spent over 3 months in bed, suicidal etc. Needless to say, I dropped out, only to go to another college with the depression still untreated. I had stopped taking the red pills since they had not worked and was conditioned into thinking that medication would not help.

It was only in my final year when another severe episode occurred whereby I became close to psychotic that I was finally prescribed a working medication, and that only at the behest of a very good friend who took me to a “proper” doctor.

Therefore, I went through a three year long nightmare resulting from some belief (and maybe experimentation, since it was teaching hospital) in the efficacy of placebos which naturally still rankles today. When I hear or read Dr. Ben Goldacre (a man who I greatly admire in his debunking of Homeopathy), I get the shivers when he states that research into placebos should be increased, since I worry about the boundaries that might need to be crossed, and the danger in doing so for such research.

[Professor Yaffle]

What makes you think the pills he gave you were placebos? Did they have a name on the label, dosage instructions etc?

[Belgian thought]

I cannot remember any name, or dosage, only the instructions i.e. take one a day. It was a long time ago to be perfectly honest. I should add I was studying Chemistry and Mathematics at the time, so the name should have registered, but your question is good - it did not register at all, and I know the names of what I have been given since that "proper" doctor.

[Professor Yaffle]

Its just, for me, its a bit of a leap from the doctor gave me some pills that didn't work to, he gave me a placebo. I have taken many different kinds of antidepressants myself, and most of them didn't work - and I can't remember the names of the first couple.

ETA: A doc would never be able to slip a placebo past me - I always look up the side effects etc of my medication and check out its mode of action.

[Belgian thought]

It might indeed be a leap, but what I do now consider, is the fact that the medication was not prescribed, as I stated, simply taken out of his cabinet. This was not a prescribed medication, plus, he was not a physiatrist.

[Professor Yaffle]

I dunno, that seems to be common practice in university health centres. I know when I went to a uni clinic for depression, I was also given a box of pills direct from the doctor. But it was definitely a real antidepressant - a tricyclic of one kind or another. And it wouldn't have been a shrink I saw either, just a general uni doctor.

I must say I would be very surprised indeed if you were given a placebo. Its possible, but I don't think likely.

[Belgian thought]

You maybe indeed be right. What remains is the need for the implementation of careful ethical protocols for any research into placebos. They might already exist, but not working in this field I could not comment but only express my, perhaps invalid, fears.

[maximara]

You maybe indeed be right. What remains is the need for the implementation of careful ethical protocols for any research into placebos. They might already exist, but not working in this field I could not comment but only express my, perhaps invalid, fears.

A tangential aspect of the placebo effect I was thinking of is the use of antibiotics for viral infections. Medically antibiotics can't really do anything (unless there is a bacterial and viral infection) and any doctor worth the name should know.

So why in the name of sanity do reports like "Antibiotic prescribing for adults with colds, upper respiratory tract infections, and bronchitis by ambulatory care physicians." out of the JAMA in 1997 show that over 20% of doctors proscribe something that their medical training tells them should not work?!