Validation of PROMIS emotional distress short form scales for cervical cancer trials.

Abstract

6633 Background: PROMIS is an NIH-sponsored effort to standardize the measurement of common symptoms and functional concerns of people with chronic conditions. The purpose of this study is to validate the PROMIS Emotional Distress (Depression and Anxiety) Short Forms (SFs) for use with cervical cancer patients. Methods: In a randomized psychosocial telephone counseling (PTC) trial, PROMIS measures of depression and anxiety were used alongside more traditional “legacy” measures of the same concepts. Questionnaires were administered to patients prior to randomization, four months post enrollment, and nine months post enrollment. The depression (8-item) and anxiety (7-item) SFs were evaluated in patients participating in this study, for internal consistency, concurrent validity, and responsiveness to change. SF performance results were compared to the Brief Symptom Inventory (BSI-18) subscales. Data were collapsed between PTC and usual care treatment arms for these analyses. Results: Patients showed significant improvement in depression from baseline to 4 months post study enrollment as measured by the Depression SF (p=0.003; effect size = 0.24), and BSI-Dep (p=0.041; effect size = 0.25). Internal consistency of the two PROMIS SFs were ≥0.95 at baseline (N=203), 4 months (N=165), and 9 months (N=151). The average inter-item correlation was 0.71 at the 4 month assessment for both depression and anxiety. The Depression SF T-score correlated well with the BSI-Dep (0.74), and was moderately correlated with the FACT Emotional Well-Being (EWB) subscale (-0.54), and the Perceived Stress Scale (PSS)(0.63). The Anxiety SF T-score was moderately correlated with BSI-Anx (0.60), EWB (-0.53), and PSS (0.60). Conclusions: The PROMIS Depression and Anxiety SFs reliably and validly assess cervical cancer-specific emotional distress, they capture salient features of distress in this population, and they perform as well or better than legacy measures in this gynecologic cancer population. Clinical trial information: NCT00496106.