Pharmaceutical companies are constantly pushing the envelope on drug development and testing, because each new product proven safe and effective means more cash flow. So is it really surprising to anyone that industry-funded, published studies are 55% more likely to provide positive results than non-industry funded ones?
There are two things at work here. First, the pharmaceutical companies are less likely to want to publish unfavorable results. They're scientists with an agenda and that agenda is selling more product. Publishing studies that showed deleterious side effects or no therapeutic effect at all aren't worth the time to put the study to print.
Second, pharmaceutical-funded studies are also half as likely to use a control group to test their drugs. This is a major science no-no. A control group is necessary to ensure that your intervention - drug or otherwise - is not simply inducing the placebo effect. This still-not-understood effect often shows healing results in people who merely think they are receiving therapy, even if they just ingest sugar pills. Two-thirds of the drug companies' studies eschew the control group - and any change of ruling out the placebo effect.
What does this mean? Next time you see a commercial for Havidol touting how "studies show miraculous effects," just remember who paid the scientist.
Note: I also recall a fascinating tidbit about statistical significance in scientific studies. There are a disproportionate number of studies published with findings that are just over the threshold of statistical significance (here's a link to an analysis of political science studies). The study of peer-reviewed journals revealed a slight bias in scientific publications toward publishing findings that could "prove" something, although a little bit of mathematical tinkering is often the only barrier between "just significant" and "insignificant." This may be why 54% of all pharmaceutical studies found positive results instead of just half.
No comments:
Post a Comment