Improving single-subject change assessment: deriving the minimal detectable change of questionnaires' ordinal scores from Rasch analysis measures.

Abstract

The minimal detectable change (MDC) of questionnaire ordinal scores (MDC_ord) is the smallest score difference exceeding the measurement error. Being ordinal, it suffers from flaws, successfully addressed by the Rasch analysis (RA) and its robust interval measures. However, RA measures struggle to become established, likely because scores are more straightforward. This study aims to derive the MDC_ord of two upper limb measures, the Fugl-Meyer Assessment-Upper Limb (FMA-UL) and the Functional Assessment Test for Upper Limb (FAST-UL), from the MDC of their RA interval measures (MDC_int). Two methodologies are applied. The first, based on a sensitivity and specificity analysis, defines the MDC_ord as the score difference with the highest accuracy in identifying a change per the MDC_int. The second derives the MDC_ord from RA strata, another MDC_int formulation. Various computations are tested, possibly resulting in slightly different MDC_ord. The MDC_ord of the FMA-UL from sensitivity and specificity analysis was 8 and 4-5 for the FAST-UL. Using RA strata, the FMA-UL MDC_ord was 8-10, and that of the FAST-UL was 4-5. An easy-to-use MDC_ord has been provided for the FMA-UL and the FAST-UL, which, anchored to the RA MDC_int, benefits from its robust measurement properties. NA.