RT Journal
A1 Spielmans GI, Jureidini J, Healy D, Purssey R
T1 INappropriate data and measures lead to questionable conclusions
JF JAMA Psychiatry
JO JAMA Psychiatry
YR 2013
FD January 1
VO 70
IS 1
SP 121
OP 123
DO 10.1001/jamapsychiatry.2013.747
UL http://dx.doi.org/10.1001/jamapsychiatry.2013.747
AB 
Though their article purported to include data regarding patients with major depressive disorder, not all trials included patients with a primary diagnosis of depression. For example, in the LYAQ fluoxetine trial, participants either (1) met DSM-IV criteria for attention-deficit/hyperactivity disorder and had concurrent anxious/depressive symptoms or (2) met criteria for an anxiety or depressive disorder and had concurrent attention-deficit/hyperactivity disorder symptoms. Further, all participants taking fluoxetine were also taking atomoxetine hydrochloride, rendering the trial incapable of assessing potential fluoxetine-related suicidality; thus, this study does not provide evidence regarding either the efficacy or safety of fluoxetine relative to placebo.3 We would have gladly combed through descriptions of the remaining studies to assess for similar problems, but most of them are difficult or impossible to obtain. Readers should not have to attempt a search of manufacturer clinical trial databases, which are far from inclusive, to find basic descriptions of included studies; at the very least, for each included trial, a clearer description of participants' diagnoses, concurrent treatment, reports of suicide-related events, and change on depression rating scales is needed to better understand the nature of the data underlying the claims of Gibbons et al.