[tokenadult]

Now why did I know that Ted Kaptchuk[1] would be the quoted "expert" (he is not a medical doctor) in this 23 December 2010 story even before I read it? Because he is always the guy pushing this line[2] in press releases[3] that get picked up by the popular media.

Meanwhile, the medical researchers who look at the issue with proper study designs and statistical controls know that placebos are essentially useless, as they at most have influence just on self-reported subjective symptoms, not on any sign that affects the progression of a disease or maintenance of good health.[4] Ladies and gentlemen, you know you aren't going to seek "placebo medicine" if you have cancer or congestive heart failure, and you know that no compassionate parent would seek "placebo medicine" for minor children who have a childhood disease. So why does this topic keep coming up over and over and over here on Hacker News, now most recently from a brand-new participant here? Take the time and effort to learn a bit more about the actual research base before assuming that this story is anything other than the outcome of carefully crafted press release.

Findings on placebo effects by researchers who have considered the issue carefully include

"Despite the spin of the authors – these results put placebo medicine into crystal clear perspective, and I think they are generalizable and consistent with other placebo studies. For objective physiological outcomes, there is no significant placebo effect. Placebos are no better than no treatment at all."[5]

"We did not find that placebo interventions have important clinical effects in general. However, in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting. The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important. Variations in the effect of placebo were partly explained by variations in how trials were conducted and how patients were informed."[6]

[1] http://harvardmagazine.com/2013/01/the-placebo-phenomenon

[2] http://www.sciencebasedmedicine.org/dummy-medicine-dummy-doc...

http://www.sciencebasedmedicine.org/dummy-medicine-dummy-doc...

http://www.sciencebasedmedicine.org/dummy-medicine-dummy-doc...

[3] http://www.phdcomics.com/comics/archive.php?comicid=1174

[4] http://www.sciencebasedmedicine.org/michael-specter-on-the-p...

http://www.sciencebasedmedicine.org/ted-kaptchuk-versus-plac...

[5] http://theness.com/neurologicablog/index.php/the-rise-and-fa...

[6] http://www.ncbi.nlm.nih.gov/pubmed/20091554

[disgruntledphd2]

I would say that you should perhaps not shoot the messenger in this case. While I am not a massive fan of Kaptchuk's experimental rigour (and I suspect most of what those articles say is not news to me), the study also had Irving Kirsch, who whatever else he may be, is a fine experimentalist. Kaptchuk's pretty good at getting funding though, so hence his appearance.

Also, if you think that placebo is nonsense, I would humbly suggest that you read Benedetti http://www.amazon.com/Placebo-Effects-Understanding-mechanis.... Its a very good summary of the state of the art in 2008, from someone (Benedetti) who runs extremely tight, well-designed experiments in reasonably valid conditions (typically post-surgery patients). Some of the findings are extremely interesting, and it is all well-referenced and supported.

[tokenadult]

Thanks for your comments on the quoted researchers. I have been trying to find the DIRECT link to a quotation from Fabrizio Benedetti, a co-author of one of the most cited papers who is also a medical doctor, in which he sums up his view this way: "I am a doctor, it is true, but I am mainly a neurophysiologist, so I use the placebo response as a model to understand how our brain works. I am not sure that in the future it will have a clinical application." (The stuff in the quotation marks appears online in articles on other websites, but I don't know specifically when Benedetti said that, except it was after his most famous paper, co-authored with Kaptchuk.)

The state of the art since 2008 has not been an advance in finding clinically useful placebo effects so much as it has been an advance in finding statistical flaws in previous studies of placebos. I really appreciated your comments in dialogue with another participant in this same thread about what the research shows, and indeed how one might define "placebo effect," and I'll have to digest that for the next time this issue comes up here on HN. Thanks.

[disgruntledphd2]

No worries, I've really enjoyed your postings on many topics, mostly around hiring techniques (and have always been a big fan of Hunter & Schmidt).

I think that I am somewhat biased, given that I started a PhD in the placebo effect around then, so I actually (sortof) know all of these people. I would argue that there are a few problems with placebo research as currently practiced.

1) clinical studies without no-treatment arms 2) Relatively small experimental studies with not completely explicit treatment protocols 3) A fascination with colourful brain images at the expense of good experimental design (though that is sadly not limited to placebo research).

Statistics is very, very difficult to get right (and I've often struggled) and the incentives are not lined up in the correct way. For instance, if I find a counter-intuitive results in an experiment, it does not benefit me to engage in rigorous fact-checking, I am more likely to benefit if I just publish it, given the demands of tenure-track. To be honest, its a wonder any science gets done at all.

And hence why I no longer work in science.

[gruseom]

That's fascinating. What did you leave science to do?

[disgruntledphd2]

I work for a tech company as an analyst. Better perks, less ethical committees, and (somewhat) more career stability.