HOME > STATS > ARTICLES

The line between the New York Times editorial page, which has been highly critical of the FDA’s handling of the potential risk from the diabetes drug Avandia, and the health news pages appears to have disappeared in the Times’ coverage of the forthcoming FDA hearings on the drug. The Times claimed that diabetics

“face an increased risk of heart attacks while those who take Actos, a similar drug, do not, federal drug reviewers conclude in a new report.”

But that “conclusion” is only one of many points raised pro and contra in the briefing papers published in advance of the FDA meeting on July 30, and the introduction to the FDA’s briefing paper does not endorse any interpretation or conclusion of the data contained within. After all, the entire point of the meeting is to figure out what all this data really means or doesn’t mean.

In fact, the actual review contains so many references to inconsistencies in the data and different ways of interpreting the findings that the Times’ bold declaration is simplistic and misleading – especially when compared with the results of outcomes-based randomized clinical trials conducted on Avandia, which, so far, have not demonstrated an increased risk of heart attack. Although there are limitations with the data here too, a well-conducted randomized control trial is always superior to meta-analysis in determining risk.

One of the few clear findings in the FDA review is that Avandia does increase the risk of heart attack in insulin trials compared to a placebo – but evidence from clinical practice suggests that doctors are not prescribing Avandia for this subgroup anyway.

The Times also failed to note that a study by drug maker Glaxo Smith Kline of insurance records of 400,000 diabetes sufferers did not find Avandia was more risky than other oral diabetes medications.

But the leading evidence to suggest that the Times has abandoned objectivity in this story is the way it characterizes the debate.

“The report largely confirms concerns first raised in May by a study written by Dr. Steven Nissen, a renowned cardiologist from the Cleveland Clinic, and published in The New England Journal of Medicine. Dr. Nissen and the Journal were criticized by Republicans on Capitol Hill and others for rushing to print a study they said was flawed.”

This makes criticism of Nissen’s study seem like sour partisan grapes – and incorrect. What the Times ignored in the category of the “others” include highly critical editorials in The Lancet, and Nature Clinical Practice, an article in the spring issue of Reviews in Cardiovascular Medicine, which called for "higher quality methods" in conducting meta-analysis on the basis of the flaws in the Nissen study - flaws that "worked to find statistical significance when in fact it was not there," and a submission by Dr. Brian Strom, Chair and Professor of Biostatistics and Epidemiology and Professor of Medicine and Pharmacology at the University of Pennsylvania School of Medicine, to the House Oversight Committee, which detailed the problems of Nissen’s study in considerable detail.

All this criticism was consistent: the Nissen study was objectively flawed.

The New York Times also turned to Dr. Bruce Psaty, a drug safety expert, for expert commentary, and he told the paper that there was no reason to prescribe Avandia. But the Times didn’t note that Psaty co-wrote (along with Dr. Curt Furburg) the accompanying editorial to the Nissen study in the New England Journal of Medicine that was highly critical of Avandia. The editorial was highly controversial in that it appeared to reject the idea of glycemic control as a justification for prescribing Avandia, a contention that runs counter to the prevailing consensus within endocrinology.

View the Technorati Link Cosmos for this entry