Well, the placebo effect could be regression to the mean in situations where patients are systematically reporting to doctors and entering trials on their worse days. But this does not seem like a universal explanation for all placebo effects. The problem with the recent article "proving" that the placebo effect is a "myth" is that it lumps all sorts of trials together. I don't think anyone doubted that in some situations there is no placebo effect, and in other situations there is. If you load up your analysis without alot of studies for situations of serious physical problems with hard outcomes, it will swamp out the real placebo effects in situations of milder problems more susceptible to psychosomatic phenomena, as well as studies with soft outcomes susceptible to subjective interpretations. Another related issue is whether all trials need a no-treatment control group. Again there can be no universal answer to this. Ideally all trials would have three arms, including a no-treatment (or placebo) group, to allow assessment of whether the active control worked properly. In laboratory research it is standard procedure to run both negative and positive controls. But in medical research this strategy is limited by ethical concerns. In the many situations where the active control has established efficacy, and no-treatment would be an undue burden on patients, then the active control should be checked against historical results. It is good that this study has challenged the use of placebos and no treatment arms when they are not necessary and ethical. But the bottom line for all these conroversies is that medical research is complicated. There are few truly universal and simple answers. People who try to live by simplistic truths will be right sometimes, but wrong alot. The solution is to think each situation through carefully, and avoid blindly applying the same research design to every situation. David L. Doggett, Ph.D. Senior Medical Research Analyst Health Technology Assessment and Information Services ECRI, a non-profit health services research organization 5200 Butler Pike Plymouth Meeting, Pennsylvania 19462, U.S.A. Phone: (610) 825-6000 x5509 FAX: (610) 834-1275 http://www.ecri.org e-mail: [log in to unmask] -----Original Message----- From: Ted Harding [mailto:[log in to unmask]] Sent: Thursday, August 16, 2001 3:59 AM To: [log in to unmask] Subject: Re: FW: Re: Placebo effect Hi Folks, On 14-Aug-01 Ted Harding wrote: > The following has just been posted to the 'allstat' list. > ... > In passing, I don't know what "regression to the mean" > can have to do with it, at any rate in the usual, statistical > sense of the term! > ... > -----FW: <[log in to unmask]>----- > > From: Troels Ring <[log in to unmask]> > To: [log in to unmask] > Subject: Re: Placebo effect > > Could it be > Is the Placebo Powerless? An Analysis of Clinical Trials Comparing > Placebo with No Treatment > Hrobjartsson A., Gotzsche P. C. > N Engl J Med 2001; 344:1594-1602, May 24, 2001. > > Best wishes > Troels Ring, Aalborg, Denmark > > At 00:33 8/13/01, you wrote: >>I heard something in passing lately to the effect that someone has >>proven that the placebo effect is actually regression to the mean, >>and otherwise doesn't exist. > ... > --------------End of forwarded message------------------------- A follow-up on allstat has confirmed the NEJM reference, and also cites a story of May 24 in New York Times, of which the following are excerpts: Kolata, Gina. "Placebo Effect Is More Myth Than Science, Study Says," New York Times, 24 May 2001, p. A20. In a new report that is being met with a mixture of astonishment and sometimes disbelief, two Danish researchers say the placebo effect is a myth. ... Instead, the researchers theorize, patients seem to improve after taking placebos because most diseases have uneven courses in which their severity waxes and wanes. ... Dr. Donald Berry, for example, a statistician at the M. D. Anderson Cancer Center in Houston, said: "I believe it. In fact, I have long believed that the placebo effect is nothing more than a regression effect," referring to a well-known statistical observation that a patient who feels particularly terrible one day will almost invariably feel better the next day, no matter what is done for him. On which I can only comment that this should work the other way round as well: If one feels particularly well one day, then one will almost invariably feel worse the next day, no matter what is done (at least, that's my experience, especially if the euphoria occured in the later hours of the day before). So it should average out -- unless of course the placebo patients are particularly sad people ... [And I'm still not sure that I rank this as "regression to the mean" in the usual sense] Best wishes to all, Ted. -------------------------------------------------------------------- E-Mail: (Ted Harding) <[log in to unmask]> Fax-to-email: +44 (0)870 167 1972 Date: 16-Aug-01 Time: 08:59:10 ------------------------------ XFMail ------------------------------