Vioxx study withheld data: medical journal

The New England Journal of Medicine says that researchers withheld data from a 2000 study on Vioxx (rofecoxib): three additional heart attacks among subjects taking Vioxx were not reported; if they had, the already significant gap in heart attack risk between Vioxx and naproxen would have been even greater (5× vs. 4¼×). Here's a PDF of the NEJM editorial. Merck denies the allegation, saying that the infarctions occurred after the study's cutoff date. Read the news coverage from CBC News, CP via the Globe and Mail, and Reuters via CNN.

A note about the comments

This blog isn't exactly a high-traffic site, but we've begun to get some comment spam anyway. As a result, I've turned on Blogger's comment moderation feature: comments on posts now have to be manually approved before they appear. Concomitantly, I'm allowing anonymous comments (since I have to review them anyway, it can't hurt). Hopefully the net effect of these changes will be positive.