Video-based physiotherapy assessment of knee osteoarthritis in patients with knee pain: a validity and reliability pilot study.

Abstract

Background: Digital health technologies are advancing rapidly, with an increasing number of physiotherapists favoring real-time, video-based platforms over telephone-based modalities. Osteoarthritis affects an estimated 595 million individuals worldwide, with approximately 62% of cases involving the knee joint. However, evidence regarding the validity and reliability of video-based assessments for knee osteoarthritis (KOA) remains scarce. The purpose of this pilot study was to explore the feasibility of video-based physiotherapy assessment of KOA in patients with knee pain. Additionally, we aimed to provide preliminary data on its concurrent validity and interrater reliability compared with conventional face-to-face assessment.

Methods: A cross-sectional validity and reliability pilot study was conducted in June 2024. Participants were recruited through public advertisements. Eligible individuals were aged 45 years or older and reported knee pain. Each participant underwent both a real-time video-based physiotherapy assessment and a conventional, face-to-face assessment. The video-based assessments were recorded for later analysis. Concurrent validity was examined by determining the exact or potential agreement between the video-based and face-to-face assessments. Interrater reliability was evaluated by comparing the live video-based assessments with those obtained from the recorded video-based assessments.

Results: For concurrent validity, exact agreement was observed in 28 of 35 cases (80%; κ = 0.35), indicating fair agreement. Potential agreement was achieved in 33 of 35 cases (94%; κ = 0.64), indicating substantial agreement. Interrater reliability demonstrated exact agreement in 25 of 29 cases (86%; κ = 0.52), corresponding to moderate agreement. Potential agreement for interrater reliability was observed in 27 of 29 cases (93%; κ = 0.63), corresponding to substantial agreement.

Conclusions: Video-based physiotherapy assessment appears feasible and may provide preliminary indications of validity for diagnosing KOA in individuals with nontraumatic knee pain. The results suggest acceptable interrater agreement and highlight the need for more standardized digital assessment protocols to ensure consistent and reliable use in clinical practice.

Trial registration: ISRCTN Registry (ISRCTN41057250), 09/05/2025. Retrospectively registered. Prospectively registered in FoU in VGR (researchweb.org) 282608, Date of registration 26/03/2024.