Letters Section Editor: Stephen J. Lurie,
MD, PhD, Senior Editor.
In Reply: We concur with the number needed
to treat of 10 that Dr Mathews and colleagues computed; this same number appears
in the original article on page 1037. The exclusion of zero in the 95% CIs
calculated by Mathews et al could be viewed as a validation of our statistically
Mathews et al and and Dr Price are concerned about our ITT analysis.
Furthermore, Price and Mr Spielmans are concerned about the clinical relevance
of our end point. To answer these concerns, we have reanalyzed the data with
respect to the CDRS-R response and have included all randomized patients as
well as various levels of response criteria (Table 1). These results show that inclusion of all patients in the
ITT analysis, in fact, favor sertraline and that the significance of the responder
analyses was not confined to the 40% criterion. We acknowledged modifications
to the ITT population in our article and we continue to feel that our analysis
population is defensible and was the most conservative option. In response
to concerns about clinical vs statistical significance, we feel that the 10%
difference in response between treatment groups, in addition to being statistically
significant, is clinically meaningful. The 65% rate in our study is a high
medication response rate for children and adolescents with MDD.
Wagner KD, Wohlberg CJ, Yang R. Efficacy of Sertraline in the Treatment of Children and Adolescents With Major Depressive Disorder—Reply. JAMA. 2004;291(1):40. doi:10.1001/jama.291.1.42-a