Psychometric Properties of Functional Mobility Outcome Measures in Children with Arthrogryposis Multiplex Congenita

AIM:To establish the construct validity, agreement, and minimal important difference (MID) of widely used mobility measures in arthrogryposis multiplex congenita (AMC). METHOD:Participants (n = 248, 126 males, mean age 10 years 10 months, standard deviation 3 years 11 months) with AMC were assessed using the Functional Mobility Scale (FMS), Gillette Functional Assessment Questionnaire (FAQ), Functional Independence Measure for Children (WeeFIM), and Patient-Reported Outcomes Measurement Information System (PROMIS). Convergent and discriminant validity were evaluated using Spearman's rank correlations, while known-groups validity was examined using analysis of variance. Cohen's kappa and distribution-based methods were used to estimate agreement and MIDs respectively. RESULTS:Robust convergent (ρ = 0.66-0.82, 95% confidence interval [CI] 0.54-0.86) and discriminant (ρ = 0.06-0.31, 95% CI -0.11 to 0.43) validity were found for all four mobility measures. Known-groups validity was supported by significant mean differences across AMC subtypes (amyoplasia, distal arthrogryposis, central nervous system/syndromic; p < 0.001). The measures also showed weak to good agreement in classifying mobility. A difference of one and two levels on the FMS and FAQ respectively, was found to be minimally important. For the PROMIS and WeeFIM, estimated MID values were 3.19 to 4.34 and 14.24 respectively. INTERPRETATION:The robust construct validity, agreement, and MIDs provide clinicians and researchers with evidence-based benchmarks for assessing mobility in children with AMC.