Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: In designing our analytic strategy, we had considered accounting for the clustering associated with group randomization by employing mixed effects linear models.1 As also noted by Dr Schlenger and colleagues, we had recognized that the traditional multivariable regression models that we used might inflate type I error rates.2 However, in light of this yielding a conservative test of the safety of youth suicide screening programs (maximizing the statistical power to detect a difference between the experimental and control groups), we considered it prudent to use this analytic strategy.
Gould MS, Kleinman M, Davies M, Marrocco FA. Iatrogenic Risk of Screening for Youth Suicide—Reply. JAMA. 2005;294(20):2578-2580. doi:10.1001/jama.294.20.2579