Over at Gooznews, Marilyn Mann explains very-clearly, and efficiently, what was not found in the post-hoc analysis of SANDS. In this case, The Wall Street Journal largely missed the actual story.
As ever, do go read it all, but here is a particularly tasty-snippet:
. . . .The [not-so] new study is a post-hoc analysis, a term which refers to testing data for patterns that you had notplanned to look for when the study was designed. Sometimes called "data dredging," this technique has been compared to shooting an arrow into a target and then drawing a bull's-eye around it.
In this case, the researchers compared patients who achieved low cholesterol levels with a combination of a statin and ezetimibe with patients who achieved similar levels with a statin alone, and found no difference in the effect on atherosclerosis. However, only 69 patients received ezetimibe and the two groups were not determined through randomization. . . .
. . . .Although SANDS was funded by the National Institutes of Health, several of the researchers have close ties to Merck and Schering-Plough. . . .
Gee -- I am shocked -- simply shocked, to learn that the researchers have significant ties to Schering and/or Merck. This is an excellent, sober look at last week's hype on SANDS -- out of Kenilworth. As ever, spot-on, and well done, Marilyn!
No comments:
Post a Comment