Table 1 Aspects of analytical validation for SV95C for Duchenne muscular dystrophy

Accuracy

Compute mean difference between the digital and traditional outcome measures and its standard deviation.

Distance walked in 6 min measured by the wearable device was compared to the distance measured by physiotherapists. The difference was 0.75 m ± 8.9. Stride speed measured by the wearable was compared to that measured by a motion capture room (wearable device detected 98.7% of the strides).

Repeatability

Calculate test-retest reliability using the intra-cluster correlation (ICC), which measures relatedness of repeated responses.

ICC between individuals’ first 15 days and second 15 days within a 1-month baseline recording were computed. SV95C had a high ICC of 0.937.

Robustness

Plot variability of digital outcome measure vs length of measurement period and identify when stability is observed. Compare variability for environmental/contextual factors.

Low variability of 4.41% was observed for SV95C after 180 h of wear time. No differences were found between morning and afternoon, but significant differences were found between weekdays and weekends.