This study aimed to identify the essential parts of the Quality of Life (QOL) Appraisal Profile that capture the most important differences in the ways that people with multiple sclerosis respond to patient-reported outcome (PRO) measures. This process will enable the eventual development of a more practical, less resource-intensive version of the QOL Appraisal Profile to facilitate its use in clinical research and practice.
This is a secondary analysis of longitudinal PRO data (n = 859) of participants in the North American Research Committee on Multiple Sclerosis registry. Following the Rapkin and Schwartz (Health Qual Life Outcomes 2(1):14, 2004) model, we computed a “standard QOL model,” and then multivariate analysis of variance (MANOVA) and discriminant function (DF) analysis to identify patterns of appraisal measures associated with group differences in response to each QOL outcome.
The “standard QOL model” explains a moderate amount of variance (i.e., 15–17 %) in physical functioning and disease-specific disability, and very little variance in mental health functioning. The MANOVAs identified the appraisal variables that mattered by PRO, and the DF analysis included 10–16 of the 83 potential appraisal variables in two DFs per outcome that distinguished groups with better, average, and worse expected scores, as well as groups with better-than-expected, as-expected, and worse-than-expected scores. The dominant appraisal parameters were more similar between the generic and disease-specific measure of physical functioning and disability, respectively, than between the mental health measure and the former two measures.
The practical implications of this work all revolve around a fundamental recommendation: Whenever one measures QOL, one should measure appraisal.