Anterior coracoid coverage and the glenoid roof morphometry: a radiographic model of directional shoulder instability

Background Quantification of anterior and posterior bony containment of the glenoid roof provides a reproducible radiographic framework for characterizing direction-specific shoulder instability on standard radiographs. However, simple and reproducible radiographic parameters integrating anterior and posterior roof morphology have not been systematically defined. The aim of this study was to compare posterior acromial coverage (PAC), anterior coracoid coverage (ACC), and glenoid roof coverage (GRC), defined as the sum of ACC and PAC, among shoulders with neutral alignment, static posterior subluxation (Walch B), and chronic anterior instability before Latarjet surgery. We hypothesized that posterior acromial coverage would be reduced in static posterior subluxation (Walch B), whereas anterior coracoid coverage would be reduced in chronic anterior instability compared with neutral shoulders. Methods A prospectively maintained institutional shoulder registry was retrospectively reviewed to identify 150 standardized preoperative Neer lateral Y-view radiographs, including 50 clinically stable arthritic shoulders with neutral glenoid alignment (Neutral), 50 shoulders with static posterior subluxation (Walch B), and 50 shoulders with chronic anterior instability treated with a Latarjet procedure (Latarjet). PAC and ACC were measured independently by two observers using predefined radiographic landmarks. Between-group comparisons were performed using Mann–Whitney U tests. Diagnostic performance was assessed using receiver operating characteristic (ROC) analysis, and associations with group allocation were explored using logistic regression. Interobserver reliability was evaluated using intraclass correlation coefficients (ICC). Results PAC was lower in Walch B than in Neutral (63.2° [55.3-68.9] vs 70.0° [65.9-77.9]; p<0.001) and discriminated Walch B from Neutral (AUC 0.76; 95% CI 0.66-0.85). ACC was lower in Latarjet than in Neutral (70.1° [61.2-77.1] vs 75.6° [69.6-83.7]; p=0.001) and moderately discriminated Latarjet from Neutral (AUC 0.70; 95% CI 0.59-0.79). GRC was highest in Neutral and reduced in Walch B (p=0.006) and Latarjet (p<0.001). Interobserver reliability was excellent for all parameters (ICC ≥0.85). Conclusion PAC and ACC quantify direction-specific alterations of superior glenoid roof morphology and provide a reproducible radiographic framework for identifying reduced anterior or posterior humeral head coverage on standardized Y-view radiographs. Level of Evidence Level III, Retrospective Cohort Comparison using Registry Data, Prognosis Study