Validating PROMIS Instruments in Depression

In order to validate the item banks and to examine their utility as computerized adaptive tests (CATs), the PROMIS network has designed a series of studies that will allow us to examine the attributes of the measures in "real-world" clinical environments. This protocol is aimed at comparing the psychometric properties of the PROMIS item banks with non-PROMIS "gold standard" instruments currently used in our respective fields (pain and mental health). In this context, note that the proposed study is not intended to evaluate treatment effectiveness, and no control group has been included. The main consideration has been to design a study involving ecologically valid treatments with established efficacy that can be administered and evaluated over the short term (i.e., 3 months). Regardless of their impact in the aggregate, such treatments will generate considerable variability in individual outcomes, and this heterogeneity is optimal for examining relevant psychometric issues. The psychometric issue of greatest concern is the validity of the PROMIS item banks as evidenced in convergent and discriminant validity and responsiveness to change. We also will make initial estimates of clinically significant change as reflected in our PROMIS measures. By combining efforts of the two sites that led in the development of the item banks for emotional distress (University of Pittsburgh) and pain (University of Washington), the study will maximize total sample size and provide a fertile ground for analyses of psychometric functioning of the PROMIS banks. In addition to psychometric questions, we will also address clinically meaningful questions related to pain, depression, and the relationship between the two. The complex relationship between pain and depression has been observed for years. Both syndromes are mutually exacerbating-pain worsens depression and depression worsens the experience of pain. The domain-related issues of greatest interest focus on the interaction between depression and pain and its impact on treatment outcome (including changes in symptoms of both depression and pain, in acute clinical status, and in social functioning).