Musculoskeletal Science & Practice · Nov 10, 2026
The Self-reported Performance Of Lower Limb Scale (Self-POLLS): development and Rasch analysis of a subjective measure to assess unilateral lower limb physical function
By Pierobon A, Taylor W, Siegert R et al.
Background Currently available patient reported outcome measures (PROMs) for people with knee osteoarthritis have key limitations that impair assessment of interventions aimed at improving physical function in research and clinical practice. Objectives To develop a new, comprehensive, limb-specific PROM to assess lower limb physical function in people with knee osteoarthritis and evaluate its initial measurement properties. Design Development and measurement properties assessment study. Methods A novel 33-item preliminary Self-reported Performance Of Lower Limb Scale (Self-POLLS) was developed following COSMIN guidelines. The preliminary Self-POLLS was tested in a cross-sectional study including people with diverse lower limb physical function. Rasch analysis was used to reduce items and assess structural validity (model fit and dimensionality), internal consistency, item fit and item difficulty, and differential item functioning (DIF). A-priori hypotheses assessed construct validity. The final Self-POLLS was assessed for floor and ceiling effects. Results A total of 127 participants were recruited. Rasch modelling took ten steps: the removal of twelve items and creation of subtests resulted in good model fit and unidimensionality. Internal consistency was excellent (0.87). Construct validity was supported by confirmation of all hypotheses. No floor or ceiling effects were detected. The final Self-POLLS comprise 21 items. The questionnaire was rated as ‘very easy’ and ‘easy’ to understand by the majority of the respondents. Conclusion The Self-POLLS shows preliminary evidence of validity and internal consistency as a PROM to assess lower limb physical function in people with knee osteoarthritis. The Self-POLLS was found easy to understand. Further research should explore longitudinal measurement properties.
doi.org/10.1016/j.msksp.2026.103689Open accessPubMed 43025357Biomarkers