Letters Section Editor: Jody W. Zylke, MD, Senior Editor.
Author Affiliations: INSERM et Université Pierre et Marie Curie, Paris, France (Dr Costagliola; email@example.com); and Departments of Epidemiology and Biostatistics, Harvard School of Public Health, Boston, Massachusetts (Dr Hernán).
To the Editor: Observational studies are increasingly being used for comparative effectiveness research. To be relevant for clinical practice, observational studies need to mimic RCTs as closely as possible. The study by Dr Bangalore and colleagues1 illustrates several ways in which an observational study may unnecessarily differ from an RCT. The authors wanted to estimate the effect of β-blocker use on cardiovascular disease among patients with prior MI, those with prior CAD but without MI, and those with CAD risk factors.
Costagliola D, Hernán MA. β-Blocker Use for Patients With or at Risk for Coronary Artery Disease. JAMA. 2013;309(5):438. doi:10.1001/jama.2012.128862