Clinical Utility of Wearable Sensors After Hip and Knee Arthroplasty: A Scoping Review

Background: Wearable sensors can measure recovery after hip and knee arthroplasty, but clinical utility depends on whether their information changes care. Methods: This scoping review searched PubMed and Europe PMC (September 2026), OpenAlex (October 2026) and supplementary sources. AI-assisted screening, charting and comparative risk-of-bias assessment were followed by a targeted human approval of additional AI source checks. Results: Of 3084 unique records, 273 reports from 243 studies were included; 624 of 980 reports sought could not be assessed in full. Measurement or monitoring featured in 256 reports, association or prediction in 123, feedback or management in 71, and comparative sensor-supported care in 43 (37 studies). None of the 43 AI-assessed comparative results was rated at low risk of bias. Two human verifiers approved 322 additional AI review rows on 8 October 2026, covering comparative findings, candidate eligibility, bias-domain triage and cohort links. This was an unblinded approval exercise, not independent duplicate screening or a new formal risk-of-bias reassessment. Sixteen recovered candidates remained outside the map pending complete charting and reconciliation. Source checks identified denominator and numerical inconsistencies requiring qualified interpretation. Conclusions: Wearables support measurement of recovery; clinical utility remains specific to the action and care programme evaluated. Activity benefits do not establish patient-important benefit, and access limitations, unresolved source discrepancies and provisional AI-derived assessments constrain causal conclusions.