Jeremy

That has come up in in Lustig’s book (Metabolical) that Mitch and I are reading. The meta-analyses are now so polluted with the industry studies that these studies neutralize findings from the academic studies. This has had a disturbing dilutional effect. Unless you remove the industrial studies, your data comes out pretty much mixed. Those are attempts to broaden randomized controlled trials. However, it is still the same mental model regarding what proves efficacy.

Jack

Jeremy, there is also the purchase of researchers.  We had the case in Colorado of a very prominent health-related researcher who was well-known nationally for doing health research.  He ended up board-disgraced because he took a whole lot of money specifically to study the safety of high fructose corn syrup.  It was really sad because he was a prominent proponent of health, but he ended up taking money from essentially the equivalent to the R. J. Reynolds of corn syrup.

Bill G

We talked a little bit about that before—particularly about the meta-analysis. I served on the pharmaceutical and therapeutics committee of a major association for decades. It came out that pharmaceutical companies who sponsor research can actually suppress negative results. So, an awful lot of the prominent meta-analyses are distorted, as Jeremey said, because the negative results or the non-superiority results were eliminated from the studies that were included.  Actually, we had some brilliant clinical pharmacists on those committees. You could actually get all of the pharmaceutical sponsored studies reconsidered. The meta-analyses were redone. Of course, things that were touted to be strongly in favor of position A or position B turned out basically to be noninferiority studies when you included all of the studies including those with negative results.  It has been well published that even the prominent academic journals, like the New England Journal of Medicine tend not to publish studies that show no significant differences. The editors of these journals are interested in studies that show positive benefits. While they should publish an equal number of studies that show no benefits, these studies tend not to be published in these journals.  So, the corruption exists at multiple layers of the system.  It is not just a function of obtaining evidence, it is the function of an ability to manipulate evidence that show outcomes and the ability to suppress those studies that counter the message being delivered by the pharmaceutical and agri-business industries.

Bill B

This is an overarching mental model that operates in may disciplines–including my own discipline of psychology.  You cannot get your study published when you have no significant results. Much of our understanding of psychotherapeutic outcomes is profoundly distorted because studies that show that there is no positive outcome don’t get published.  It is only the ones where there are positive outcomes that get published. It helps if there is money behind study—such as money that is provided by the institute where you get trained to do this particular therapeutic procedure. As a result, there is lost trust not only the field of medicine but also in the field of psychotherapy. Mitch, I think it is a broader mental model that allows only positive feedback (negative feedback is ignored).  The espoused theory is that we are interested in all information.  We are open. We want to be informed. Scientists learn more from non-results and negative results (failures) than from positive results (successes).  However, as Bill G. has noted, our theory-in-use is if you don’t obtain positive result at the .05 level or .01 level of significance then you don’t get published.