Reliability, Precision, Accuracy, and Validity of Posterior Shoulder Tightness Assessment in Overhead Athletes

Neuromuscular Research Laboratory, Sports Medicine and Nutrition, School of Health and Rehabilitation Sciences, University of Pittsburgh, Pennsylvania

^* To whom correspondence should be addressed. E-mail: myersjb{at}upmc.edu.

	Abstract

Background: Posterior shoulder tightness with subsequent loss of humeral internal rotation range of motion has been linked to upper extremity lesions in overhead athletes. A valid clinical assessment is necessary to accurately identify posterior shoulder tightness as a contributor to injury.

Purpose: To describe a modified supine assessment of posterior shoulder tightness by establishing the reliability, precision, clinical accuracy, and validity of the assessment.

Study Design: Cohort study (diagnosis); Level of evidence, 2.

Methods: Intrasession, intersession, and intertester reliability and precision were established by comparing the commonly used side-lying assessment of posterior shoulder tightness and the described modified supine assessment. Clinical accuracy of both methods was obtained using an electromagnetic tracking device to track humeral and scapular motion. Construct validity was established by identifying posterior shoulder tightness in a group of overhead athletes (baseball pitchers and tennis players) reported in the literature to have limited humeral internal rotation and posterior shoulder tightness.

Results: The side-lying intrasession intraclass correlation coeffecient (standard error of measurement), intersession intraclass correlation coeffecient (standard error of measurement), and intertester intraclass correlation coeffecient (standard error of measurement) were 0.83 cm (0.9), 0.42 cm (1.7), and 0.69 cm (1.4), respectively. The supine intrasession intraclass correlation coeffecient (standard error of measurement), intersession intraclass correlation coeffecient (standard error of measurement), and intertester intraclass correlation coeffecient (standard error of measurement) were 0.91°(1.1°), 0.75°(1.8°), and 0.94°(1.8°), respectively. In side-lying, the clinical accuracy expected was 0.9 ±0.6 cm of error while, when measured supine, it was 3.5°±2.8°of error. Both assessments resulted in minimal scapular protraction (~3.5°). Between groups, baseball pitchers and tennis players had significantly less internal rotation range of motion (P <.0001) and greater posterior shoulder tightness (P =.004) when measured in supine, but not in side-lying (P = .312).

Conclusion: Both methods resulted in good clinician accuracy and precision, suggesting that both can be performed accurately. The supine method can be assessed more reliably than side-lying between both sessions and testers.

Clinical Relevance: Clinicians may want to consider use of the supine method given the higher reliability, validity, and similar precision and clinical accuracy.