Internal Rotation Behind-the-Back Angle: A Reliable Angular Measurement for Shoulder Internal Rotation Behind the Back

BACKGROUND: The hand-behind-back method is the accepted technique to evaluate shoulder internal rotation, is highly popular, and is endorsed by the American Academy of Orthopaedic Surgeons. It remains, however, subject to intra- and interexaminer discrepancy and has been challenged by several recent publications. HYPOTHESIS: Internal rotation behind the back can be evaluated with a measurable angle, which eliminates the need to estimate spinal level, decreases the effect of unrelated joints, and allows collection of numeric rather than categoric data. STUDY DESIGN: Descriptive laboratory study. METHODS: We defined the internal rotation behind-the-back (IRB) angle as that between the ulna and the line of gravity. A pendulum is attached to a standard goniometer. The patient is asked to reach the highest point along the midline of the back. The goniometer is centered over the pisiform, and the angle between the ulnar axis and the pendulum is measured. Two examiners assessed both shoulders of 60 volunteers with no shoulder pathology using this technique on 2 occasions. Both examiners were blinded to each other’s values. We applied the paired Student t test and calculated Pearson correlation coefficients and weighted Cohen kappa values. RESULTS: The IRB angles ranged from 50° to 125°. The difference of the mean, as measured with the Student t test, was 0.6° (95% confidence interval: 0.1°, 1.3°) and 0.6° (95% confidence interval: −0.8°, 1.8°); the Pearson correlation coefficients were 0.98 and 0.92; and the weighted kappa values were 0.88 and 0.77 for interobserver and intraobserver analyses, respectively. CONCLUSION: The IRB angle is easy to measure, is reproducible, and does not rely on determination of spinal level. It provides numeric data and may eliminate some of the uncertainty associated with the estimation of spinal level. CLINICAL RELEVANCE: The IRB angle may eliminate some of the uncertainty associated with the estimation of spinal level.