OBJECTIVES: Lower Extremity Functional Scale (LEFS) is a widely used 20-item patient-reported outcome instrument with five response categories. The purpose of this study was to investigate the measurement properties of the Finnish version of the LEFS among foot and ankle patients and refine the scale to measure more accurately that what matters to the patient.

METHODS: Data were obtained from an earlier cross-sectional validation study of the Finnish version of the LEFS. Altogether 185 patients who had undergone foot and ankle surgery for various reasons were included. Rasch measurement theory was used to analyse construct validity, model and individual item fit, and reliability.

RESULTS: Misfit of thresholds was noted in 13 of the 20 original LEFS items and thus scale unidimensionality was not supported. After collapsing the response categories 1 (“Quite a bit of difficulty”) and 2 (“Moderate difficulty”) together, two items (item 10 “Getting into or out of a car” and item 11 “Walking 2 blocks”) showed threshold misfit. As unidimensionality was not supported by deleting items 10 and 11, three more items were omitted. This new 15-item scale with four response categories supported unidimensionality (proportion of significant t-tests, 1.4%). All items had ordered thresholds and good item fit (fit residuals inside -2.5 and 2.5). Reliability of the revised scale was high as the Person Separation Index and Cronhach’s alpha were 0.85 and 0.95, respectively.

CONCLUSIONS: The new 15-item LEFS scale with four response categories has a unidimensional construct. The refined LEFS seems to be more valid than the original 20-item LEFS in assessing foot and ankle function. This scale should be further tested for its reliability, validity and responsiveness.