Evaluation of the Circles Measurement and the ABC Classification of Acromioclavicular Joint Injuries

• Published in: The American journal of sports medicine 2021-05, Vol.49 (6), p.1619-1625
• Main Authors: Murphy, Richard J., Moor, Beat K., Lesniewski, Piotr J., Hayoz, Annabel, Alcantara, Wolfan, Zumstein, Matthias A.
• Format: Article
• Language: English
• Subjects: Classification; Injuries; Sports medicine; Validity
• ISSN: 0363-5465; 1552-3365
• DOI: 10.1177/03635465211003300

Background: Acromioclavicular joint (ACJ) injuries are common. Despite this, it remains unclear how best to assess, classify, and manage these cases. A simple, reliable, valid, and accurate radiographic parameter to measure ACJ displacement would allow improved consistency of diagnosis and subsequent treatment pathways. Purpose: To evaluate “the circles measurement” and associated “ABC classification” as a tool for assessing ACJ displacement and injury classification. Study Design: Descriptive laboratory study. Methods: The circles measurement is taken from a lateral Alexander radiograph of the shoulder. The measurement is the center-to-center distance between 2 circles drawn to define the lateral extent of the clavicle and the anteromedial extent of the acromion; it is independent of the displacement plane, judging total ACJ displacement in any direction rather than trying to quantify vertical and/or horizontal displacement. When utilized clinically, the circles measurement is a single measurement calculated as the difference between values recorded for the injured and uninjured sides. Validation of the circles measurement was performed using lateral Alexander radiographs (including ±20° projection error in all planes) and computed tomography of standardized ACJ injury simulations. We assessed inter- and intrarater reliability, convergent validity, and discriminant validity of the circles measurement and subsequently generated a classification of ACJ injury based on displacement. Results: Reliability and validity of the circles measurement was excellent throughout. Interrater reliability (ICC [intraclass correlation coefficient] [2,1], 95% CI; n = 78; 4 observers) was 0.976 (0.964-0.985). Intrarater reliability (ICC [2,1]; 95% CI; n = 78; 2 measures) was 0.998 (0.996-0.998). Convergent validity (Pearson correlation coefficient, r) was 0.970 for ideal radiographs and 0.889 with ±20° projection error in all planes. Discriminant validity, with 1-way analysis of variance, showed a P value of