The Future Is Mobile: Pilot Validation Study of Apple Health Metrics in Orthopaedic Trauma.

BACKGROUND

Surgeons often lack objective data on patient functional outcomes, particularly as compared with the patient's baseline. The present study aimed to determine whether gait parameters recorded on Apple iPhones provided longitudinal mobility data following lower-extremity fracture surgery that matched clinical expectations. We hypothesized that iPhones would detect the mobility changes of injury and early recovery, correlate with patient-reported outcome measures, and differentiate nonunion.

METHODS

This cross-sectional study included 107 adult patients with lower-extremity fractures who owned iPhones and had at least 6 months of follow-up. Participants shared Apple Health data and completed Patient Reported Outcomes Measurement Information System (PROMIS) surveys. The primary outcome was the daily step count. Four other gait-related parameters were analyzed: walking asymmetry, double support, walking speed, and step length. Mixed-effects models compared mobility parameters at pre-injury, immediate post-injury, and 6-months post-injury time points. Correlations between mobility parameters and PROMIS surveys were assessed. A mixed-effect model evaluated the relationship between step count recovery and surgery for nonunion.

RESULTS

There was a 93% reduction in daily step count from the pre-injury period to the immediate post-injury period (95% confidence interval [CI], -94% to -93%). Other gait parameters also showed increased impairment from pre-injury to post-injury. At 6 months, step count improved sixfold relative to the immediate post-injury period but remained 52% below baseline (95% CI, -55% to -49%). PROMIS Physical Function correlated moderately with step count (r = 0.42; 95% CI, 0.25 to 0.57) and weakly with other gait parameters. Patients with a known nonunion had a 55% slower recovery of step count than those without a nonunion (95% CI: 44% to 66%).

CONCLUSIONS

Apple Health mobility parameters captured changes in mobility following lower-extremity fracture and throughout the subsequent recovery period. These metrics distinguished between patients with and without nonunions, demonstrating their potential usefulness as objective, real-world functional outcome measures. These "digital biomarkers" may aid clinical decision-making and research and could be utilized for the early identification of patients at risk for poor outcomes.

LEVEL OF EVIDENCE

Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.