Each assessor performed imaging evaluations using the Horos picture archiving and communication system (Version 3.3.6; Horos Project). The purpose of the evaluators was to classify AC joint injuries using both the Rockwood and Kraus classifications and to select nonoperative management or surgical treatment according to each categorization. For each set of images, the evaluators were instructed to measure the CCD on the AP view using the digital caliper tool of the software. The CCD was defined as the vertical distance between the upper border of the coracoid process and the lower border of the clavicle on the injured and contralateral sides (Figure 2).
Subsequently, the CCD ratio ({[CCD injured – CCD healthy]/CCD healthy} × 100) was calculated to (1) classify the injury using the Rockwood system and (2) select a treatment option between nonoperative management and surgical treatment. After the qualitative assessment of the Alexander radiographs, 2 additional inquiries were made: (3) to classify the injury using the Kraus system and (4) to select a treatment option accordingly. The number and order of the set of images were randomly modified in the second stage, and the 12 investigators repeated the analysis 6 weeks later using the same approach to determine intraobserver reliability. This time period was chosen to avoid visual recall by the surgeon.