Premium
Psychometric Properties of the Walking Impact Scale (Walk‐12) in Persons with Late Effects of Polio
Author(s) -
Brogårdh Christina,
Lexell Jan,
Westergren Albert
Publication year - 2021
Publication title -
pmandr
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.617
H-Index - 66
eISSN - 1934-1563
pISSN - 1934-1482
DOI - 10.1002/pmrj.12403
Subject(s) - rasch model , differential item functioning , reliability (semiconductor) , rating scale , psychometrics , scale (ratio) , physical medicine and rehabilitation , psychology , activities of daily living , physical therapy , item response theory , medicine , clinical psychology , developmental psychology , power (physics) , physics , quantum mechanics
Background Many persons with late effects of polio (LEoP) perceive walking limitations in everyday life. A common rating scale to assess walking limitations is the Walking Impact Scale (Walk‐12). However, there is limited knowledge about its psychometric properties. Objective To investigate the psychometric properties of Walk‐12 in persons with LEoP. Design Rasch model analysis of cross‐sectional data. Setting University hospital. Participants A total of 325 persons with LEoP (175 women; mean age 70 ± 10 years). Main Outcome Measurement The Walk‐12, comprising 12 items with five response categories ranging from 1 (not at all) to 5 (extremely). Methods Data of Walk‐12 were collected by a postal survey. The Rasch model analysis was used to analyze unidimensionality of the scale, local dependency, targeting, hierarchical order of items, Differential Item Functioning (DIF), response category functioning, and reliability (Person Separation Index, PSI). Raw score transformation to interval measurements was also performed. Results The analysis revealed that Walk‐12 was multidimensional and suffered from some local dependency. Targeting was compromised among persons with less and more walking limitations. Hierarchically, the most difficult item to perform was “running” and the easiest was “walking indoors with support”. There was a minor DIF for gender in one item (“support when walking outdoors”). Reliability was high (PSI = 0.94). Disordered response category thresholds were found for three items; when merging the middle response categories for these items model fit slightly improved and unidimensionality was achieved. Conclusions The Walk‐12, in its current version, does not fully meet the rigorous psychometric Rasch measurement standards in persons with LEoP. Further development of the scale is warranted, including merging response categories and complementing Walk‐12 with objective measures of gait in order to improve targeting. As these limitations can be considered minor, the current version of Walk‐12 can still be useful for research and clinical practice.