Longitudinal metric properties of disability rating scales for Parkinson's disease.

OBJECTIVES This study analyzes the longitudinal metric attributes of three Parkinson's disease (PD) disability scales, taking Hoehn and Yahr (HY) staging as the reference measure of PD progression. METHODS A sample of 87 PD patients was assessed during regular medical visits, using the HY, the Unified Parkinson's Disease Rating Scale--Activities of Daily Living Section (UPDRS-ADL), the Schwab and England Scale (SES), and the Intermediate Scale for Assessment of PD (ISAPD), across a follow-up period of 2.6 +/- 1.0 years. RESULTS The following cross-sectional attributes were analyzed, at baseline and again on conclusion of the study: floor and ceiling effects, convergent validity, reliability, and standard error of measurement, all of which were found to be adequate. Longitudinal reproducibility values (intraclass correlation coefficient) were 0.81 (ISAPD) to 0.84 (UPDRS-ADL). Insofar as longitudinal validity was concerned, the change scores of the three disability scales correlated significantly with the HY change score, absolute value r = 0.33 to 0.45, P < 0.003. Slightly lower values were found when taking the annual rate of change, absolute value r = 0.20 to 0.36. The three scales were acceptable, even though there were small differences among them. The "minimal clinically important difference" proposed for these scales is: SES, -6; UPDRS-ADL, +2; ISAPD, +1.5 points. CONCLUSIONS The three scales proved adequate for longitudinal assessment of PD disability. UPDRS-ADL was more precise and ISAPD more consistent. Magnitude of change and correlation with change in HY were slightly higher with the ISAPD. Effect size and standardized response mean for the minimal change in HY were higher for the UPDRS-ADL.