Integrating Gait Analysis and AI Into Knee Osteoarthritis Care: Protocol for a 3-Phase Participatory Qualitative Study

Clinical Snapshot

Key Findings

• Effect Size: Not applicable - protocol study

• Primary Outcome: Protocol paper - no results yet available

• Nnt Or Sensitivity: Not applicable - qualitative protocol

Clinical Application

Moderate - requires institutional support and clinician engagement, dependent on technology infrastructure Framework potentially applicable to Australian orthopedic services, though would need adaptation for different healthcare system structures and TGA regulatory requirements for AI-based clinical tools Orthopedic clinicians managing knee osteoarthritis patients, particularly in Canadian healthcare settings

Abstract

BACKGROUND: Knee osteoarthritis (OA) leads to pain, disability, and reduced quality of life. For advanced stages, knee arthroplasty surgery is the standard treatment, yet dissatisfaction and persistent mobility deficits remain common. Current surgical decision-making processes seldom incorporate objective predictors of outcomes, such as biomechanical data. Advances in computer vision, wearable sensors, and artificial intelligence now enable efficient capture and interpretation of clinically prognostic gait features in real-world settings. However, the clinical adoption of such innovations remains limited, hindered by usability challenges, misalignment with stakeholder needs, and system-level barriers. OBJECTIVE: This study aims to inform the development of a knee OA clinical decision support (CDS) tool that integrates gait analysis and artificial intelligence with clinical decision-making. More broadly, it seeks to advance a framework for participatory digital health CDS tool implementation by identifying and addressing suboptimal workflows, stakeholder priorities, and organizational challenges. METHODS: This study uses a 3-phase participatory design process, with in-depth interviews conducted in each phase to generate insights that inform subsequent stages. Phase 1 investigates perceived barriers to and facilitators of the implementation of a digital CDS tool and examines current workflow processes in knee OA management. Phase 2 focuses on identifying and defining user requirements to guide solution ideation for translational digital decision support. Phase 3 assesses user acceptance of prototyped solutions. Field observations in phase 1 and usability testing in phase 3 will supplement interview findings with real-world evidence. RESULTS: The study was funded in November 2024. Ethics approval was granted by the Nova Scotia Health Research Ethics Board on July 28, 2025. Data collection began in September 2025. As of February 2026, a total of 6 clinicians have been interviewed, with all 6 interviews transcribed and 4 coded as part of phase 1 analysis. Study completion is anticipated by August 2026. Findings will be published by January 2027. CONCLUSIONS: Our stakeholder-driven approach prioritizes both technological rigor and practical usability, addressing key human factors for successful implementation. While initially focused on Nova Scotia's orthopedic setting, this research provides a scalable framework for integrating digital decision support tools into broader clinical environments, advancing innovation in knee OA care and beyond.