To the Editor: We read with interest the article by Shelley F. McMain, Ph.D., et al. (1), published in the December 2009 issue of the Journal. It is my opinion that additional information would have clarified the primary study findings. The authors reported that there was no difference between treatment groups in primary outcomes of frequency and maximum severity of nonsuicidal self-injurious behavior, presenting a group contrast odds ratio and coefficient, respectively, without confidence intervals. When no difference is found between two potentially active treatments, confidence intervals inform the question of whether a study has either demonstrated that the treatments are equivalent or failed to demonstrate that one treatment is superior (2). In the former case, confidence intervals for a group contrast are narrow enough to exclude clinically meaningful differences (defined by the authors as at least a 20% difference in the rate of self-injurious behavior), while in the latter case confidence intervals include both clinically meaningful and nonmeaningful differences. Confidence intervals are therefore important in understanding the authors' conclusion that study results "legitimize the recommendations of the APA practice guideline on the management of borderline personality disorder" (1, p. 1372) using general psychiatric management. This conclusion would be supported most strongly by the finding that general psychiatric management is either equivalent or not inferior to dialectical behavior therapy, a treatment with demonstrated efficacy (3, 4).