Construct validity, responsiveness, and interpretability of the Spinal Cord Injury Functional Ambulation Inventory (SCI-FAI)

We investigated the construct validity, responsiveness, and interpretability of the Spinal Cord Injury Functional Ambulation Inventory (SCI-FAI) to determine its usefulness in measuring the functional level of gait. This was a prospective observational study following the checklist of the Consensus-Based Standards for Selecting Health Measurement Instruments. The SCI-FAI consists of three items: Gait Parameter, Assistive Devices, and Temporal. We recruited 55 patients diagnosed with a spinal cord injury (SCI) in the subacute or chronic phase from Ibaraki Prefectural University of Health Sciences Hospital. Construct validity was clarified by hypothesis testing and the rate of hypothesis verification. We divided responsiveness into subacute and chronic groups and determined the correlation between the changes in the two timepoints of the SCI-FAI and 6-minute walk test. Interpretability involved the calculation of cutoff values for indoor and outdoor walking. Construct validity was confirmed for 12 of the 15 hypotheses, indicating high construct validity. In the subacute group, Gait Parameter and Assistive Devices showed moderate responsiveness. Interpretability showed that the Gait Parameter was perfect in 19 of the 20 subjects who did not need a walking aid when walking. The cutoff value for Gait Parameter for indoor walking was 17.5 points (AUC 0.91) and that for Assistive Devices was 9.5 points (AUC 0.88). The cutoff values for outdoor walking were Gait Parameter 18.5 points (AUC 0.96) and Assistive Devices 10.5 points (AUC 0.94). Our results demonstrated that the SCI-FAI has adequate construct validity, moderate responsiveness in SCI patients in the subacute phase, and interpretability in the gait assessment of individuals with SCIs. Gait Parameter is likely to show a ceiling effect for people with SCIs who can walk without using an upper-limb walking aid.