Letters Section Editor: Robert M. Golub, MD, Senior Editor.
To the Editor: In their Commentary, Dr Kraemer and colleagues1 draw attention to the importance of moderators of treatment outcomes, the often unsatisfactory way these are dealt with (if at all) in randomized clinical trials (RCTs), and the potentially serious consequences of this neglect on patient care. They propose that every RCT should include a search for putative moderators of treatment outcome. Although acknowledging that positive results of such analyses rarely provide proof of the existence of these moderators, they stress that the importance resides in the hypotheses that the analyses generate. These hypotheses can then be tested in subsequent adequately powered and populated RCTs. However, it may take years to design, acquire adequate funding for, and execute such an RCT.
Deinum J, van der Wilt GJ. A Word From Our Moderator. JAMA. 2007;297(2):156-157. doi:10.1001/jama.297.2.156-a