Science based Medicine: Placebo Myths Debunked

The main thing to know is that the idea of the powerful placebo was based on an article published in 1955 which jumped to a conclusion involving the healing power of suggestion instead of carefully considering other explanations for why patients often improved even with a sham treatment.

Some reasons an illness can improve or falsely appear to improve even without effective treatment:

- Many illness get better without treatment.
- A change in how patients perceive, think or report their symptoms, such as answers of politeness, hope, undergoing therapy that aims to change perception of the illness, etc.
- Patients are receiving some other treatment at the same time.
- Fluctuations in illness severity, occurring by chance at a time that creates the appearance of a sham treatment working.

There seems to be a great deal of confirmation bias in the literature promoting the idea of the powerful placebo.

 

It is interesting to note that there has been a study where ME/CFS patients were treated with homeopathy. It is difficult to compare the results directly with the PACE study, as Weatherley-Jones et al used the multidimensional fatigue inventory (MFI) for outcomes. However, if the improvement is normalized by its standard deviation (SD), the fatigue for patients that underwent homeopathy improved by 0.7, the fatigue in the CBT group in PACE improved by 1.0, and the fatigue in the GET group in PACE improved by 1.2.* Note that the homeopathy study was blinded, and that the placebo response probably would have been even larger in a non-blinded trial.

This comparison suggests that the order of magnitude of the subjective improvements in PACE is consistent with a placebo response. If Peter White and colleagues believe that subjective outcomes are appropriate because “that is how the illness is defined”, then they should also promote homeopathy as a useful treatment.

Weatherley-Jones E, Nicholl JP, Thomas KJ, et al. A randomised, controlled, triple-blind trial of the efficacy of homeopathic treatment for chronic fatigue syndrome. J Psychosom Res. 2004;56(2):189–97.

* As the SD of the outcomes were specified at baseline and end of treatment in PACE, I derived the SD for the difference under the assumption that the baseline values and the improvements were independent.

 

Can we include this article in our science library please @Cheshire, because I find it very educational.

 

How familiar is that! The BPS crew's reaction to their unwarranted reliance on subjective outcome measures, is to downplay the importance of objectivity, and instead start getting creative why subjective outcomes are adequate.

 

I'm sure they've posted a nearly identical article before!

Still, it seems that it needs repeating.

 

Damn, I guess this means that we can't take the word of royal German veterinarians as proof that homeopathy works :

I don't completely agree with this regarding the size of the placebo effect. If it's deliberately exploited as part of the treatment process, it makes sense that the placebo effect will be bigger, such as with deliberate brain-washing regarding symptom denial:

 

done. see https://www.s4me.info/index.php?thr...l-of-that-shiny-new-study-a-bibliography.212/ I've included the following summary:

Yay, first time we've cross-referenced an article in the library with its corresponding S4ME discussion!

Note in that section of the library, there are loads of articles on the placebo effect, expectation effects and other artefacts that affect clinical trials. There are also some that assess how big these effects might be.

 

The Lilienfeld article in the science library talks more comprehensively about spurious factors affecting psychotherapy trials in particular:

https://www.s4me.info/index.php?thr...l-of-that-shiny-new-study-a-bibliography.212/

Lilienfeld et al refer to this effect as "Response shift bias". (you're taught to evaluate your symptoms differently, and this has a direct effect on self reported symptoms, in the absence of genuine change).

 

So people would especially self report differently if they are convinced they have a real underlying illness, but are reprogrammed to then believe they are fundamentally out of condition, and can can exercise their way back to health again. Presumably there must be a timing issue here also, because the self reporting will be affected by whether the subject is on the crest of a euphoric wave of optimism following their reprogramming, or later after the bubble has burst and the truth dawns once more.

 

Totally. Hence PACE's CBT and GET "worked" for a while, and then stopped working (no benefit at long-term follow-up).

 

So wouldn't it be much more ethical to simply inject people with morphine and then get them to fill in a questionnaire saying how they feel - and then produce a paper or 27 that says morphine relieves some symptoms and improves the sense of well being in those with chronic illness?

Coz apart from the fact that it would be more reproducible, and "easier" on people's health, I can't see much of a difference - well and cost - it doesn't cost £5 million to supply a few hundred people for a few months.

....and whilst it is unlikely to improve anyones long term health, it's also unlikely to damage them either, the same cannot be said for GET/CBT.

I'm not suggesting this as a "treatment", I'm simply using it as an example to point out the absurdity.

Hopefully.

 

Yeah, but let them BPS people have their share of the cake...it's quite unfair to give everything to pharma, isn't it?

 

I can absolutely agree with that, at least in my case. When I realized I'm actually doing that I fixed a certain date to fill out the forms. I also can agree to the fact that I wanted to please the clinician. I am very strict, though, and honest in that aspect, so I disciplined myself.

Interesting topic, and thanks for the information!