The AJP Must Immediately Address Suicide Study's Flaws

Over the weekend, I became even more troubled by the conclusion reached in a recently-published study which asserted that 2004's black box warning on anti-depressants had led to a decrease in prescriptions for those drugs and had also led to an uptick in suicides among youth. The study was published in the American Journal of Psychiatry, but since its appearance two weeks ago its central assertions have been challenged.

Last week, the New York Times reported that, in fact, prescription rates had not changed much in 2004 (that came the next year), so that calls into question this conclusion by the study's authors:

"SSRI prescriptions for children and adolescents decreased after U.S. and European regulatory agencies issued warnings about a possible suicide risk with antidepressant use in pediatric patients, and these decreases were associated with increases in suicide rates in children and adolescents."

What's more, the black box warning in the US was not applied to anti-depressants until Oct. 15, 2004, so it is difficult to believe that the warning could have affected suicide data for the entire year.

Further doubt is cast on the study's findings because preliminary suicide data for 2005 released last week show a 3 percent drop in suicides for that year--the first full year of the black box warning and a year in which anti-depressant prescriptions actually did drop. This would seem to be the opposite of what the study's authors argued.

This whole affair is an embarrassment for the AJP. Central flaws in a significant study were not caught by peer reviewers and the journal's editors. A flawed and fundamentally incorrect paper was published to wide media attention. Americans--and especially the parents of America's youth--are now in essence being told that down is up and up is down by the world's leading psychiatric journal.

This situation must be remedied by the journal and promptly. The politics and emotions around the SSRI-suicide debate are too intense and divisive to permit this paper to just sit out in the public realm and be regarded by some forces, especially policy making ones, as "evidence." It doesn't make sense for the journal to let whatever debate will erupt over this paper to take place six months from now in the letters section of the journal.

The journal either needs to ensure that a correction is promptly issued for this paper--and done so very publicly--or that the paper itself is retracted. I don't think that these kinds of discrepancies can be cleaned up in any other way. I have no idea what the journal is doing to address this situation, but I will try to find out.

Readers from the mental health world might want to make a point of writing to the journal directly.

In addition, I have to wonder why Robert Gibbons, the study's lead author and a very experienced statistician, could screw things up as badly as has been reported. It kind of makes you wonder how reliable his testimony as an expert witness for Wyeth, makers of Effexor, could have been.

For background, I have written about this matter here and here and have also written about the CDC and media's odd handling of the 2004 suicide data.