Equivalence based tests of clinical significance: Assessing treatments for depression

Treatment efficacy is largely determined by statistical significance testing, and clinical significance testing is often used to quantify or qualify the efficacy of a treatment at the individual or group level. This study applies the equivalence based clinical significance model proposed by Kendall, Marrs-Garcia, Nath and Sheldrick (1999), and a revised model proposed by Cribbie and Arpin-Cribbie (2009), to the assessment of treatments for depression. Using several studies that investigated treatments for depression, we tested whether the post-treatment means were equivalent to the means for a similar normal comparison group. All of the studies had significant improvement from pretest to posttest, although for many of the studies the treated group was not equivalent to a normal comparison group at posttest. Further, there are important differences between the conclusions drawn from the Kendall et al. and Cribbie and Arpin-Cribbie methods for assessing equivalence based clinical significance.