therapeutics initiative letters

[Guest guest]

I've been at the Therapeutics Initiative site trying to understand their reasoning behind why they would still favor statin therapy and came across this comment from a reader about Randomized Controlled Trials (RCT) which I think we might appreciate.

http://ti.ubc.ca/en/node/159

As Always, I thoroughly enjoy Therapeutics Letter as it is the only publication that captures the true essence of a RCT and its clinical implications. I would like to add a couple more possible biases that can significantly alter putative benefits. First, when multiple endpoints are combined to get a composite endpoint, the standard p of less than 0.05% may be inadequate to prove benefit as there remains a probability that each of the components of the composite may have reached statistical significance by chance alone. Second, it is now well accepted that RCT's sponsored by Industry are up to 4 times more likely to show benefit of a drug intervention. This factor alone can make a RCT that reached statistical significance invalid. Third, if benefits identified in a well controlled RCT are just statistically significant, then such benefits are unlikely to be realized in clinical practice where many patients who do not fit the inclusion criteria of the study are prescribed as an extrapolation of the study findings. Therefore, it is my humble opinion that unless benefits in a RCT are substantial, then it is unlikely that such benefits are going to be realized in a clinical setting thereby wasting our very limited financial resources. I would appreciate a comment from one of your esteemed experts. See you in Vancouver, my annual pilgrimage for medical enlightenment! Keep up the fantastic work and thanks for the email reminders.