Comments on: Evidence based medicine and shared decision making http://www.kevinmd.com/blog/2008/08/evidence-based-medicine-and-shared.html medical blog Sun, 22 Nov 2009 07:50:13 -0500 http://wordpress.org/?v=2.8.4 hourly 1 By: Indiancowboy http://www.kevinmd.com/blog/2008/08/evidence-based-medicine-and-shared.html/comment-page-1#comment-87008 Indiancowboy Fri, 08 Aug 2008 17:17:00 +0000 http://clients.emmense.com/kevinmd/2008/08/evidence-based-medicine-and-shared-decision-making.html#comment-87008 Which leaves out the whole problem of just how flawed many of these studies are. Cox-2 inhibitor trials not controlling for the lack of cox-1 inhibition as a possible mechanism of increased heart attack risk. Glucosamine/chondroitin/MSM studies done on people with bone on bone contact. Etc.<br/><br/>Using numbers like ARR, RRR, NNT, NNH, and all that is especially problematic when we try to pretend that the numbers themselves have any sort of validity.<br/><br/>In something as complex as the human body, you can try to take an overall snapshot with a study but it's beyond our power to say 'this many will be saved and this many people will be hurt'.<br/><br/>I am not talking about statistical power but limits of study design. Which leaves out the whole problem of just how flawed many of these studies are. Cox-2 inhibitor trials not controlling for the lack of cox-1 inhibition as a possible mechanism of increased heart attack risk. Glucosamine/chondroitin/MSM studies done on people with bone on bone contact. Etc.

Using numbers like ARR, RRR, NNT, NNH, and all that is especially problematic when we try to pretend that the numbers themselves have any sort of validity.

In something as complex as the human body, you can try to take an overall snapshot with a study but it’s beyond our power to say ‘this many will be saved and this many people will be hurt’.

I am not talking about statistical power but limits of study design.

]]>