Self-assessed Dynamic Gait Index correlates with physical therapist assessed Dynamic Gait Index in people with multiple sclerosis

Background: The Dynamic Gait Index (DGI) is a clinical measure of walking ability comprised of eight walking tasks. In people with multiple sclerosis (PwMS) the DGI has demonstrated validity, reliability, and ability to identify fallers. A self-assessed version of the DGI that demonstrates concurrent validity with the original DGI in people with MS would be valuable for remote assessment of walking ability. We therefore developed a questionnaire-based self-assessed version of the DGI (sDGI) that asks participants to self-rate their predicted ability to perform the eight DGI walking tasks. The purpose of this study was to determine the validity and internal consistency of the sDGI in people with MS who had self-reported gait impairment. Methods: We enrolled 53 ambulatory people with MS with self-reported gait impairment. Participants completed the sDGI, the Multiple Sclerosis Walking Scale-12 (MSWS-12), the Telephone Interview for Cognitive Status (TICS), and self-reported their number of falls in the past 3 months. Then, up to 4 weeks later, they completed the DGI by in-person evaluation by a physical therapist (PT). We calculated the internal consistency of the sDGI and concordance between the sDGI and DGI. To determine if cognition impacted concordance, we calculated concordance in the cognitively impaired and non-impaired groups and also tested the difference between groups. We also calculated correlation between the sDGI and the DGI and the MSWS-12. The ability of the sDGI and the DGI to identify fallers was evaluated using receiver operating characteristic curves. Results: 51 participants completed the study. They had a mean age of 60.9 (SD 11.5) years, median PDDS of 4.0 (interquartile range 3, 5), 32 % used walking aids, 25 % were cognitively impaired, and 62 % were female. The sDGI was internally consistent (Cronbach's alpha 0.85, 95 % CI 0.76, 0.90) but was not concordant (CCC = 0.45, 95 % CI: 0.28–0.60) with the DGI. Concordance between the sDGI and DGI was not different for cognitively intact versus cognitively impaired individuals. The sDGI was moderately correlated with the DGI (R = 0.64, p < 0.01) and strongly correlated with the MSWS-12 (R = -0.71, p < 0.01). Neither the sDGI nor the DGI identified fallers in this sample. Conclusion: The moderate to strong correlation among the sDGI, DGI and MSWS-12 supports the validity of the sDGI as a measure of walking ability in people with MS. However, because there is poor concordance between the sDGI and the DGI performed in clinic by a PT in both cognitively intact and cognitively impaired participants, the sDGI should not be used as a substitute for the DGI. The sDGI could be used as a screening tool because most participants underestimated their walking performance compared to assessment by a PT so the sDGI would be unlikely to miss poor walking ability. As neither the sDGI nor the DGI identified fallers in this sample, our results do not support using these tools to identify fallers in people with MS and self-reported gait impairment.