Radiographic characterization of the B2 glenoid: is inclusion of the entirety of the scapula necessary?

Background Computed tomography (CT) scans are often obtained before total shoulder arthroplasty to assess glenoid deformity. To allow correction of the slice axis into the plane of the scapula, these scans have typically required inclusion of the entirety of the scapula. The purpose of this study was to determine whether inclusion of the medial border and inferior angle is necessary for accurate measurement of scapular version, inclination, and humeral subluxation. Methods Fourteen CT scans in preoperative total shoulder arthroplasty patients with Walch B2 type glenoids underwent a standardized measurement protocol. Glenoid version, inclination, depth, and humeral subluxation were measured on 2-dimensional CT images corrected to the plane of the scapula. These measurements were then repeated in randomized, blinded fashion after subtracting 12.5%, 25%, and 50% of the scapula from the medial border and 12.5%, 25%, and 50% of the scapula from the inferior angle. Results Measurement of retroversion did not significantly differ between measurement of the full scapula and measurement of any of the incomplete scapulas, with the exception of the subtraction of 50% of the scapular width, which caused retroversion to be overestimated by 4.7° (P = .006) and led to inaccurate measurement of subluxation and glenoid depth. Conclusion If at least 8 cm of scapular width is imaged on a CT scan, accurate glenoid measurements can be made. Even if 50% of scapular height is not imaged, accurate measurements can be made. Failure to include the medial border or inferior angle does not preclude accurate glenoid measurement.