Background: The Spinal Muscular Atrophy Independence Scale (SMAIS) assesses the amount of assistance required to perform daily activities for individuals with Type 2 and non-ambulant Type 3 SMA. The SMAIS was designed to be completed by individuals aged ≥12 years and by caregivers of individuals aged ≥2 years. The first version consisted of 29 items, scored on a 5-point scale. Validation analyses indicated that the five response options were poorly discriminated and that the SMAIS was not assessing a single construct. This led to the creation of the SMAIS Upper Limb Module (SMAIS-ULM), a 22-item upper limb total score (removing seven items measuring mobility/chores), scored on a 3-point scale. Higher scores indicate greater independence.
Objectives: This research aimed to evaluate item- and scale-level psychometric performance of the SMAIS-ULM, using cross-sectional patient-reported SMAIS data collected as part of the Cure SMA 2020 Community Update Survey. Classical test theory and polytomous Rasch measurement theory methods were used for the analyses.
Results: The analysis sample included Type 2 (N=173) and non-ambulant Type 3 (N=66) individuals, aged between 18–75 years (mean=36 years). Item means ranged from 0.14–1.71; mean of the total score was 18.3 (range=0–44). Item option distributions demonstrated minimal floor and ceiling effects. Item locations ranged from –5.45 to 4.93 logits, indicating a broad range of difficulty. Items displayed ordered response thresholds for the 3-point scale, and good scale-to-sample targeting supported the conceptual relevance of the item set. Internal consistency was high (Cronbach’s α=0.96) and the items showed acceptable overall item fit (OUTFIT mean square (MSQ)=0.84, INFIT MSQ=0.88). Confirmatory factor analysis supported unidimensionality of the SMAIS-ULM. Evidence of known-groups validity was supported by a statistically significant difference (P<0.001) between the SMAIS-ULM for Types 2 and 3 SMA (mean=14.8 vs. 27.5). Convergent/divergent validity was assessed via Spearman rank order correlations with the EuroQol-5D-3L. Results followed expected patterns, with the highest correlation observed with self-care (r=–0.60) and the weakest correlation observed with anxiety/depression (r=–0.03).
Conclusions: The SMAIS-ULM demonstrated acceptable item- and scale-level cross-sectional measurement properties. Future research will involve the re-examination of the seven mobility/chores items not included in the SMAIS-ULM.