Comments on: How Well Do Drugs Work? Hidden Research Sometimes Makes It Hard to Tell

By: derp gellington

derp gellington — Wed, 11 Jan 2012 23:26:35 +0000

“Latest results on a few studies show our drug isn’t actually very effective.”
“Slip some cash discretely to the lead scientists not to publish and no one will ever know. But make sure any favourable studies do get published. If the general public and doctors aren’t told this drug barely works we’ll make billions. So crush as many negative studies as you can. Who’s gonna stop us? The FDA? Ah hahahaha”

Just another day in the office with our lovely, bloated, and corrupt big pharma. Money has always been more important then human lives to these people.

By: TouchTheRiot

TouchTheRiot — Sun, 08 Jan 2012 17:53:48 +0000

We need to make it illegal to not publish the results within one year of the completion of the study and make them ineligible to even request more tax payer money with any outstanding study results unpublished. If tax money is what makes a finding possible, then the tax payers should be the owners of the IP, but the researcher can receive a sizable royalty. We need overwatch and accountability, geez.

By: Iain

Iain — Sun, 08 Jan 2012 03:22:20 +0000

All investors are entitled to see what their money has done. If taxpayers are funding a study, it needs to be published in a public accessible data base within a reasonable time, say 12 months after study complete or refund monies received. Also the database needs to be advertised so people can easily find and access it. A single government spot with sub-directories should suffice.

By: dave chamberlin

dave chamberlin — Sun, 08 Jan 2012 02:33:35 +0000

you mean big pharma has clout in obscuring what they don’t want known? naaaaa. They are in it just to help people.

By: Bob Snyder

Bob Snyder — Sat, 07 Jan 2012 21:32:41 +0000

@drkala, My understanding of the article was that the issue is in regards to people making decisions with incomplete information and taxpayer dollars going into clinical trials which take years to get published after they are completed(because there is a lack of control), not whether or not “publication bias” exists.

By: drkala

drkala — Sat, 07 Jan 2012 15:50:20 +0000

This seems like a trivial finding. If you increase the size of the sample in meta analysis, you increase the power of the finding. Therefore, you would expect to see stronger results in both directions–work better, work worse. That is a consequence of increasing the number of underlying data points, by adding more studies. What would be problematic would be if the unpublished findings were skewed one way, say not publishing ones that show they are working worse. Since the changes are equally distributed, this actually validates the publcation process and confirms that there has not been a publication bias.
Kala Ladenheim, PhD, MSPH