Are more responsive depression scales really superior depression scales?

In contemporary clinical practice and research, many different instruments are used to measure depression severity. Santor et al. [1] identified no fewer than 280 scales developed in the last century, many of which are still in use. The responsiveness of these various depression scales to treatmentdand the comparison of responsiveness among scalesdhas become an important research topic. Responsiveness is defined as a scale’s ability to detect clinical change and to measure the true effect of treatment on depression severity [2]. There are numerous ways to assess and compare responsiveness. In this issue, Kounali et al. [3] introduce a meta-analytic method that allows to compare the responsiveness of different scales by constructing a network of randomized trials (i.e., scales can be compared across different data sets). In their proof of principle analysis, the authors examine instruments including the Patient Health Questionnaire (PHQ-9) [4] and the Beck Depression Inventory (BDI) [5], concluding that some scales perform better than others. This is consistent with the literature: responsiveness is often treated as a psychometric property [3,6], and scalesmore responsive to treatment have been advanced as superior scales that should be used in clinical trials [3,7,8]. Bech [8], for instance, has called for the use of the six-item subscale of the Hamilton Rating Scale for Depression (HRSD-6) instead of the 17and 21-item versions because, among other reasons, ‘‘fewer patients are then needed to identify antidepressant effect in controlled trials’’ (p. 310) (see also [9]). Similarly,Kounali et al. [3] advance that ‘‘responsiveness [.] canhelp guide the choice of outcome measures in clinical trials’’. This line of reasoning, that has crucial implications for future research, relies on a strong assumption: that the most responsive depression scales provide valid measures of