The validity, responsiveness, and score interpretation of the PROMISnq Physical Function – Multiple Sclerosis 15a short form in multiple sclerosis
Cook, Karon F.
Alschuler, Kevin N.
von Geldern, Gloria
Background A valid, sensitive patient-reported outcome (PRO) measure of physical function (PF) for people with multiple sclerosis (MS) would have substantial value in routine care and clinical research. We now describe development of the PROMISnq Short Form v2.0 PF – Multiple Sclerosis 15a [PROMISnq PF(MS)15a] for assessing PF in relapsing and progressive MS. Also, the validity, reliability, and responsiveness of the PROMISnq PF(MS)15a is evaluated, minimal important difference (MID) thresholds for score change estimated and a score interpretation guide developed. Methods A mixed-methods sequential design was employed. Relevant PF concepts were elicited through semi-structured interviews with people with relapsing MS, and then mapped to the PROMIS PF item bank. Measurement experts integrated results from interviews with people with MS and input from a panel of neurologists to generate a draft short form. Relevance and comprehensiveness of the draft short form were assessed in cognitive debriefing interviews with people with relapsing or progressive MS. Subsequently, item reduction and evaluation of psychometric properties were performed in two observational studies: a cross-sectional study in the US (n = 296), and a 96-week longitudinal study in the UK MS Register cohort (n = 558). The main outcomes and measures are estimates of: known-groups validity, convergent validity, reliability, responsiveness; MID for worsening. Results Factor analyses supported the unidimensionality of the newly derived 15-item short form. Cronbach's alpha (≥ 0.97) and intraclass correlation coefficient (≥ 0.97) of test-retest scores (5–27 days) indicated strong reliability. Convergent validity was demonstrated by moderate-to-strong correlations with scores on related PRO measures. Scores discriminated among patient groups classified by levels of physical health and other criteria. Score changes of 2.3–2.7 points are proposed as MID criteria for minimal worsening in PF. Conclusion PROMISnq PF(MS)15a demonstrated reliability, validity and sensitivity to change. Input from patients and clinicians ensured the content is comprehensive and relevant for people with MS.