It seems that the lead author is off the hook:
A Toronto doctor does not appear to be involved in deleting heart attack data about Vioxx from a study that helped put the arthritis drug on the market, says the editor of the New England Journal of Medicine.
Dr. Gregory Curfman, who on Thursday published an “expression of concern” about inaccuracies in the study, said yesterday that two of the 13 authors were Merck & Co. employees and “it appeared a Merck editor did the deletions.”
Chris Rangel:
“The three cases of heart attack (which have been known since 2001) increase the statistical power of the VIGOR results showing an increased risk of heart attack in the Vioxx taking patients but they do not at all change the conclusions. Why the VIGOR authors would withhold data on three cases from the Journal it would not have made a difference is beyond me. This is like accusing the government of trying to cover up the fact that 9/11 even happened by showing that they tried to cover up the evidence of United flight 93 which crashed in Pennsylvania.
The real questions that the Journal editors should be asking is why after these three cases became known to them in 2001 they didn’t call for a study looking specifically at the cardiovascular risks of Vioxx or of COX2 inhibitors or NSAIDs in general (to date no such study has been done)? Isn’t it a little too late for the Journal to be climbing on the “Vioxx kills people and we were lied to” bandwagon? These three cases apparently didn’t make much difference to the Journal in 2001 and they shouldn’t now. Monday morning quarterbacks unite!”
retired doc:
“This raises the lack of credibility to previously unreached levels as regards clinical trials funded by drug companies. Merck has issued a statement in reply to the NEJM editorial statement that in part states the three deaths occurred after the agreed upon cut-off date for the end of the trial and those deaths were reported to the FDA. We have not heard the last of this. We teach medical students that the reasons for an association or lack of one in a clinical study are: chance, causality,bias and confounding. The fifth factor, fraud, was not typically emphasized in that regard, now it maybe it should be.”
