学术文献库

Background: Scale equating is a statistical technique used to establish equivalence relations between different scales. Its use is quite popular in educational evaluation, however, unusual in the health area, where scales of measures are tools that integrate clinical practice. With the use of different scales, there is a difficulty in comparing scientific results, such as NESSCA and SARA scales, tools for assessing the commitment to Machado-Joseph disease (SCA3/MJD). Objective: Explore the method of scale equating and demonstrate its application through NESSCA and SARA scales, using the Item Response Theory (IRT) approach in assessing SCA3/MJD commitment. Methods: Data came from 227 patients from the Hospital de Clínicas de Porto Alegre with SCA3/MJD who have complete measures for NESSCA and/or SARA scales. The equating design used is that of non-equivalent groups with common items, with separate calibration. The IRT model used in the estimation of the parameters was the generalized partial credit, for NESSCA and SARA. The linear transformation was performed using the Mean/Mean, Mean/Sigma, Haebara and StokingLord methods and the equation of the true score was applied to obtain an estimated relationship between the scores of the scales. Results: Difference between NESSCA score estimated by SARA and observed NESSCA score has shown median of 0.82 points, by Mean/Sigma method. This was the best method of linear transformation among the tested. Conclusions: This study extended the use of scale equating under IRT approach to health outcomes and established an equivalence relationship between NESSCA and SARA scores, making the comparison between patients and scientific results feasible.