Lies, Damned Lies, and Medical Science by David Freedman
We are taught that modern medicine is “evidence based medicine”. Evidence based medicine means rigorous scientific studies underlie every aspect of medical practice. Particularly drugs are supposed to be grounded in randomized, double-blind, placebo controlled trials. This means that the drug is tested against a placebo, something that looks, smells, feels and tastes just like the drug but without the drug. It means the participants are randomized to receive either drug or placebo. And it means that neither the participants nor the data collector know whether a participant is receiving drug or placebo.
This article reveals that there is great incentive to bias a study, either knowingly or unknowingly, that studies are frequently biased, and that most of medicine is pretty far from being evidence based. Including, and especially, the drugs. This quote is particularly shocking: “[John Ioannides] charges that as much as 90 percent of the published medical information that doctors rely on is flawed.”
One incentive is that it is simply nearly impossible to publish negative results. “X doesn’t work” isn’t nearly as interesting as “X cures cancer”. It’s difficult to publish confirmatory results. “Breaking news–X cures cancer” is more interesting than “Dr. Lab confirms Dr. Superstar’s findings last year that X cures cancer”.
So how unbiased is science? I’m sure no one is really surprised to find out that although we are taught the scientific method, that actual science is pretty biased. It’s all a big religion. I don’t have anything against religion, but science claims to be different than religion.
Public Library of Science (PLoS) has seven journals which publish studies with sound study design, including negative results. Articles are peer reviewed and published online. That’s a step in the right direction.