Related Experiment Video
Updated: Jun 17, 2026
Measuring 3D In-vivo Shoulder Kinematics using Biplanar Videoradiography
Published on: March 12, 2021
Accuracy of visual estimation of shoulder abduction by expert physiotherapists: a comparative study using 3D motion
Martin Kenner1, Phillip Thies2, Annika Schwarz1
1Faculty of Social Sciences - Applied Therapeutic Sciences, City University of Applied Sciences Bremen, Bremen, Germany.
Background:
Accurate assessment of shoulder range of motion is essential for clinical decision-making in physiotherapy. While visual estimation of joint angles is commonly used in practice, its agreement with objective reference measurements under clinically relevant conditions remains unclear. This study aimed to assess the agreement between video-based expert visual estimation of shoulder abduction and three-dimensional (3D) motion capture as a reference standard.
Methods:
A cross-sectional method comparison study was conducted. Shoulder abduction movements of four healthy participants were recorded using a 3D motion capture system. Twenty-four video recordings were presented to experienced physiotherapists (≥ 10 years of clinical experience), who visually estimated joint angles via an online questionnaire. Differences between estimated and reference values were analysed using descriptive statistics, Bland-Altman analysis, and cross-classified mixed-effects modelling to account for repeated measurements across raters and video stimuli.
Results:
A total of 923 paired observations from 33 experts were analysed. Visual estimation showed a consistent tendency toward overestimation, with a mean difference of 14.7° (SD 11.6°) and limits of agreement ranging from - 8.0° to 37.5°. A cross-classified mixed-effects model estimated a mean bias of 14.5° (95% CI 11.8° to 17.3°), confirming systematic overestimation. Most estimates (89.9%) exceeded the reference value. Variability was substantial, with 27.3% of variance attributable to differences between video stimuli and 10.5% to differences between raters.
Conclusion:
Visual estimation of shoulder abduction angles by experienced physiotherapists showed limited agreement with a 3D motion capture reference and a consistent tendency toward overestimation. Given the observed variability, visual estimation alone may be insufficient for accurately detecting small changes in joint angle. The use of objective measurement tools may support more reliable assessment in clinical practice.
