Translation, cross-cultural adaptation and analysis of the psychometric properties of the lower extremity functional scale (LEFS): LEFS-Brazil

Abstract

Background: There is a lack of questionnaires in Brazilian Portuguese to evaluate patient-reported lower limb function. Objective: To translate, cross-culturally adapt to the Brazilian population, and evaluate the psychometric properties of the Lower Extremity Functional Scale (LEFS). Method: The LEFS was translated by two independent assessors and back-translated to English. Then, the LEFS-Brazil was tested on 20 patients who answered the questionnaire in the cross-cultural adaptation phase. For the evaluation of the psychometric properties, 100 patients answered the questionnaire. The reliability was tested by two independent assessors. The Medical Outcomes Study 36-item from Health Survey (SF-36) was used as the criterion method for construct validity. The sensitivity to change was tested for four consecutive weeks. Results: The internal consistency was α=0.96. The intra-observer reliability was CCI (intraclass correlation coefficient)=0.96 and CCI interobserver=0.98; the Bland and Altman mean difference (d) intra-observer=-1.52 and d interobserver=0.46. The correlation between the LEFS and SF-36 in the first week was the following: physical function r=0.82, physical role r=0.57, emotional role r=0.43 and mental health r=0.33. The LEFS was responsive when comparing the mean of the first week to the second, third and fourth weeks and comparing the second to the fourth week. The cut-off point was 11, and the area under the receiving operator curve was 0.96 95% CI [0.88;0.99], with sensitivity=0.96, 1-specificity=0 and standard error=0.02. Conclusion: The LEFS-Brazil is reliable, valid and responsive.