Eric West - Department of Orthopaedic Surgery and Biomedical Engineering, University of Tennessee Health Science Center-Campbell Clinic, Memphis, TN, USA
Derek Dixon - Department of Orthopaedic Surgery and Biomedical Engineering, University of Tennessee Health Science Center-Campbell Clinic, Memphis, TN, USA
Thomas Throckmorton - Department of Orthopaedic Surgery and Biomedical Engineering, University of Tennessee Health Science Center-Campbell Clinic, Memphis, TN, USA
David Bernholt - Department of Orthopaedic Surgery and Biomedical Engineering, University of Tennessee Health Science Center-Campbell Clinic, Memphis, TN, USA
Frederick Azar - Department of Orthopaedic Surgery and Biomedical Engineering, University of Tennessee Health Science Center-Campbell Clinic, Memphis, TN, USA
Tyler Brolin - Department of Orthopaedic Surgery and Biomedical Engineering, University of Tennessee Health Science Center-Campbell Clinic, Memphis, TN, USA

Objectives: Reverse total shoulder arthroplasty (rTSA) has shown success in the treatment of end-stage glenohumeral pathology. However, one major shortcoming has been the lack of internal rotation (IR), which can have significant functional consequences. Much research has been conducted to maximize IR after rTSA, but the literature is unclear which measurement of IR represents the “gold standard” between vertebral level and goniometer-based measurements.Methods: Patients were prospectively enrolled into one of three groups: postoperative from rTSA, subacromial pain (SA), and normal. IR measurements were obtained either by the vertebral body level, by which radiographic markers indicated the highest level that the patient was able to reach on the body midline; or by using a goniometer while the shoulder was in ۹۰-degree abduction as the patient stood upright.Results: Comparisons between the radiographic vertebral level and goniometer IR measurements showed significant correlations within the normal (r = - ۰.۴۳, P = ۰.۰۲) and SA pain groups (r = - ۰.۴۴, P = ۰.۰۲). The rTSA group did not quite reach statistical significance (P = ۰.۱۱), but had a moderate correlation coefficient (r = - ۰.۳۳). Accuracy of visual IR measurements was also significant. All rTSA group vertebral level measurements were within two vertebral levels, while only ۸۴.۶% of IR measurements by goniometer were within ۱۵ degrees. Visual vertebral level measurements were found to be more accurate for the SA pain group (۸۶.۲ vs ۶۶.۷%).Conclusion: A comparison of the two primary IR measurement methods for shoulders was shown to have a correlation. This would allow for direct comparison of different literature using only one measurement method. While the correlation is not yet strong enough to allow for conversion between the two measurement types, creating a matched cohort taking into account other factors may lead to the correlation reaching this point.        Level of evidence: III

internal rotation, Range of motion, Rotator Cuff, Shoulder Arthroplasty, Shoulder pain