"Our mistake then -- and we have yet to put it right -- was that we did not demand better proof before we embarked on mass medication of a large group of patients who looked to us for advice and treatment."
In a commentary, which was also published with the Cohen article, Nick Freemantle of the University of Birmingham in England, argued that much of the controversy surrounding the thiazolidinediones (the drug class to which Avandia and another diabetes drug, Actos, belong) can be laid at the feet of poor trial design -- so poor that critical questions were not answered.
For example, regulators have been willing to permit high rates of loss to follow-up, especially in trials that rely on surrogate endpoints, a position that he contended "makes no sense."
And "even when trials examine serious morbidity and mortality, loss to follow-up remains a problem. The RECORD trial, conducted to examine the safety of rosiglitazone and sponsored by GlaxoSmithKline in the U.K., failed to follow-up survival in 127 participants (2.9 percent)," he wrote. The loss to follow-up in RECORD meant that one could conclude that rosiglitazone "was associated with either an increase or a decrease in mortality given different assumptions on the fate of those lost to follow-up."