To the Editor.
—The study on depression and MI contains important new data.1 However, I have two concerns.First of all, by choosing to dichotomize variables, the authors lost important information. Ejection fraction should have been considered as a continuous variable, and more levels of smoking, Killip class, and education should have been used. The authors' argument would be strengthened by repeating their analysis using ejection fraction as a continuous variable.The treating physicians were not made aware of the diagnosis of depression. The "Methods" section states that the data were not analyzed until 3 months after the last patient was admitted to the study. "Thus, blinding researchers and clinicians to the DIS [National Institute of Mental Health Diagnostic Interview Schedule] results did not constitute an ethical problem." Failure to analyze the data is insufficient justification for withholding potentially important information (as the authors point out, depression is frequently
Thompson WG. Depression and Survival Following Myocardial Infarction. JAMA. 1994;271(14):1080-1081. doi:10.1001/jama.1994.03510380036028