Osseous lesions of the glenoid cavity and humeral head are predictive of recurrence after Bankart arthroscopic procedures. The objective of this study was to analyze, for each glenoid and humeral defect plain x-ray criteria of the Instability Severity Index Score (ISIS), two aspects: inter- and intraobserver reliability of their qualitative and quantitative assessment and correlations between positive criteria and their quantitative measurement. Thirty-one medical files were retained for evaluation of the glenoid and 26 for humeral notch assessment. The yes or no response for the ISIS criterion was completed by its quantitative measurement using the Griffiths and Sugaya CT methods for the glenoid and the P/R index calculation on plain x-rays with internal rotation for the Hill-Sachs lesion. Three observers provided two consecutive readings for each criterion. The analysis of the glenoid radiological criterion of the ISIS seems sufficiently reproducible for daily practice. When the evaluation is positive, bone loss is greater than 15%, without a maximum value established. In this study, the analysis of the ISIS humeral notch criterion was not reproducible. It can be improved using the P/R index and should be completed by CT imaging.

Level IV: retrospective diagnostic study.