Summary

The presentation of a promising model for quantitative measurements and objective assessment of scapular kinematics for clinical use.

Abstract

Background

Primary or secondary scapular dyskinesis is characterized by abnormal scapular rotations. It can be painful and impair the shoulder function. A skin marker-based motion capture model providing quantitative measures of the rotations of the scapula was recently developed and can potentially be used to diagnose and monitor scapular dyskinesis.

Aim

To investigate the concurrent validity and the interrater reliability of a new model for analysis of scapular kinematics.

Materials And Methods

Twelve infrared cameras were used to track reflections from moving skin markers in the motion capture model. A strict protocol for placement of the skin markers was followed. Shoulder range of motion (ROM) and activities of daily living (ADL) were tested in healthy volunteers. To investigate the validity, the skin marker-based model was compared to a gold standard through simultaneous data collection from markers fitted to an intracortical pin in the scapula. Interrater reliability was tested in a different group of healthy volunteers by comparing the skin marker-based protocol performed by two investigators blinded to each other’s results. The mean Root Mean Square Error (RMSE) was calculated for each tested motion to determine the validity. The interrater reliability was determined as Intraclass Correlation (ICC2,1) for each tested motion.

Results

Eight subjects were included in the validity test: female/male=2/6, mean (standard deviation) age 35.0 (3.0) and BMI 23.4 (3.3). The mean RMSE of all scapular rotations ranged 2.3-6.7° during shoulder ROM and 2.4-7.6° during ADL. The highest errors were seen for movements in front of the body: sagittal/scapular plane flexion, hair combing and eating. In 19 out of 24 measurements, the model showed larger range of rotation than the gold standard. In the reliability test, 20 subjects were included: female/male=8/12, mean (standard deviation) age 31.4 (4.9) and BMI 22.9 (1.7). ICC for measuring protraction ranged 0.07-0.60 during ROM and 0.27-0.69 for ADL. Correspondingly, ICC ranged 0.01-0.64 and 0.38-0.60 for upward rotation, and 0.25-0.83 and 0.25-0.62 for anterior tilt.

Conclusion

Our results indicate that the model’s validity and reliability are task dependent and interpretation should be made with caution. The model provides quantitative measurements for objective assessment of the scapular movements and can be an important supplement to the clinical examination. Taking the inherent limitations of the method into consideration, the model is promising for clinical use.