moldybluecheesecurds 2

Friday, December 11, 2009

Can we believe the published results of studies sponsored by drug companies?

MinnPost: In a meta-study of drug company studies on non-steroidal anti-inflammatory drugs (NSAIDs), this was what an independent researcher found:
"[The analysis] found all the studies that had ever been published where one NSAID was compared to another. In every single trial, the sponsoring company’s drug was either equivalent to, or better than, the drug it was compared to: All the drugs were better than all the other drugs. Such a result is plainly impossible." [emphasis mine]
O, ye of little faith. Drug companies clearly hail from Lake Wobegon, where all the drugs are above average.

1 comment:

rick said...

There are two issues here.

(1) Is the study designed properly?

(2) Are the researchers interested in the scientific method?

Regarding (1), if the study isn't designed properly, we can throw it out. An expert in the field should be able to review the study and determine if it is well-designed. So this issue should be manageable.

Regarding (2), the scientific method requires experimentation *and* replication.

If drug companies only publish the favorable studies, ignoring all the unfavorable studies, it is hard to know if study results are typical or a random fluke.

This issue is harder to resolve, because in order to draw meaningful conclusions regarding industry-funded studies, we have to trust that all industry-funded studies will be published, even if the results are unfavorable to the sponsoring organization.

The results of this survey suggest that such trust would be misplaced.