34

There has recently been a row about the marketing of pain killers. The makers of Nurofen have been accused of selling the same formulation at different prices for different conditions and claiming in advertising that they are "targeted" at particular sorts of pain. Australia has banned the ads.

In commenting on the row London Evening Standard columnist Richard Godwin said the following (my highlights):

What you’re paying for is the placebo effect of the marketing — and this is not as negligible as it might appear.

In the US, where drugs companies are allowed to market to consumers much more aggressively than here, researchers have noted a 30 per cent increase in placebo effectiveness from 1990-2013. In Europe there was no change.

This was news to me. But is he right? Has placebo effectiveness increased in the USA?

Community
  • 1
matt_black
  • 56,186
  • 16
  • 175
  • 373
  • I suspect "Bad Medicine" and "Bad Pharma" author Ben Goldacre has made the same claim. – Andrew Grimm Dec 16 '15 at 20:55
  • 2
    @nomenagentis But in most trials placebos **do** have measurable effects on the outcomes, though perhaps not clinically significant effects. So the question is answerable. – matt_black Dec 16 '15 at 20:58
  • 1
    @matt_black: Most studies cannot be used to demonstrate that. An effect in the placebo/control group cannot be attributed to the placebo itself. – Oddthinking Dec 16 '15 at 23:11
  • Among those people who believe the placebo effect is getting stronger, it *is* getting stronger, because they believe it! – AdamV Dec 17 '15 at 13:05
  • Indirectly related: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1505530/ considers the impact of branding and marketing on analgesic effects both of active formulations and of inactive substitutes – Henry Dec 18 '15 at 00:02
  • @Oddthinking The reason why placebos are used in randomised trials is that they do make a difference compared to *no treatment*. If the control group is no treatment, the investigators and patients know whether they are being treated or not and this may bias the results for all sorts of reasons. Placebos minimise any such effects by making everyone think they are treated and leaving only the actual effect as the difference between the control and actual groups. The effect of placebos is often large compared to no treatment. – matt_black Dec 20 '15 at 13:38
  • @matt_black: Yes, placebo groups get different results to no treatment groups. That eliminates the problems of "reversion to the mean" and the natural progress of disease that confounds such discussions. Still does not necessarily imply that there are physiological effects due to the placebo. I acknowledge I am being a pedant, but there is a bit too much magical thinking about placebos. Shall we take it to chat? – Oddthinking Dec 21 '15 at 01:55

1 Answers1

36

A recent study has suggested that effectiveness of placebo treatments increased over the years in the US, in patients suffering from neuropathic pain.

The study itself uses data from multiple previous studies performed over the years, in which real pharmacological therapy is compared in effect to placebo therapy.

We find that placebo responses have increased considerably over this period [1990-2013], but drug responses have remained stable, leading to diminished treatment advantage. This trend has been driven by studies conducted in the United States.

The term placebo response is not intended to suggest that the placebo itself is causing the response.

Note that in this article, our use of the term “placebo response” follows colloquial usage and encompasses all factors related to (apparent) analgesia in the placebo arm of a clinical trial (ie, control group), including the placebo effect itself. Nonspecific effects may include social support, attention, and education associated with clinical trials, spontaneous resolution of symptoms (ie, disease natural history), and regression to the mean.

Unfortunately, there is no comparison to cases where no therapy is administered. I don't know of, and couldn't find, any similar published study performed in Europe, although this study included US, European and Asian studies.

A number of possible causes are discussed, including changing trial size and length, publication bias, and advertising for recruitment for trials.

The study does not consider the idea that marketing of drugs has affected the placebo response.

Oddthinking
  • 140,378
  • 46
  • 548
  • 638
Egil
  • 526
  • 3
  • 7
  • [Welcome to Skeptics](http://meta.skeptics.stackexchange.com/questions/1505/welcome-to-new-users), Egil. I made a rather large edit to this answer - I considered posting it as a competing answer instead, but you did all the research. Please check it over and ensure I am not misrepresenting your position. – Oddthinking Dec 17 '15 at 00:26
  • I wonder if this is ironically due to better access to and awareness of alternative therapies in the US. I also wonder how the level of the effect (not the amount it's changed) differs between the US and Europe. – shadowtalker Dec 17 '15 at 01:21
  • Is it possible to update this answer with the actual increase in placebo responses? E.g. x% increase from y% to z%. – RockPaperLz- Mask it or Casket Dec 17 '15 at 17:07
  • 1
    I recall seeing an article talking about how the "placebo effect" was now working in people who were aware they were only taking a placebo. I do not recall the thoroughness of its research. – Trevel Dec 18 '15 at 15:51
  • 1
    @Trevel: http://skeptics.stackexchange.com/questions/8739/is-the-placebo-effect-dependent-on-the-amount-of-belief-in-it – Oddthinking Dec 21 '15 at 01:56