Construct validity of real-world digital mobility outcomes in patients after proximal femoral fracture: a cross-sectional observational study

Abstract

Abstract Digital mobility outcomes (DMOs) offer unique insights into recovery of real-world mobility after proximal femoral fracture (PFF), but their clinical validity remains to be established. This study assessed construct validity (convergent, divergent, and known-groups) of 24 DMOs measuring walking activity (amount, pattern) and gait (pace, rhythm, bout-to-bout variability) in patients within one year after PFF. Patients were recruited from inpatient and outpatient lists at five European sites, resulting in 505 included participants (66% female), with mean age of 77.6 ± 9.4 years and supervised gait speed of 0.7 ± 0.4 m/s. Mobility was monitored over seven days using a single wearable device on the lower back. Convergent and divergent validity analyses were stratified by two groups: acute (≤ 14 days since surgery) and non-acute (≥ 15 days since surgery). Correlations between DMOs and related (clinical- and patient-reported mobility outcomes) and unrelated constructs (hearing impairment and systolic blood pressure) were compared to a priori expected correlations. Known-groups validity was assessed across four recovery phases. The results were evaluated individually by experts and in a subsequent consensus meeting, with 17 of 24 DMOs showing evidence of construct validity in non-acute PFF patients. These findings represent an initial step in a larger process towards regulatory endorsement.