This retrospective study compared glenoid version measured on axillary radiographs versus CT in 50 shoulders. Half were evaluated for instability and half had a total shoulder prosthesis. Three independent observers assessed each image to test which method is reliable.
When you need to know glenoid version, order a CT and do not trust the axillary radiograph.
This matters most in two settings the paper studied: working up posterior or recurrent instability, and planning or auditing glenoid component position in total shoulder arthroplasty. Excessive retroversion drives abnormal glenoid loading and poor arthroplasty outcomes, so getting the number right changes surgical decisions.
The reason axillary films fail is anatomic. The neck blocks medial film placement, so the scapular reference line follows the neck rather than the true blade axis, adding about 6.4° of false retroversion. A practical pearl: the Friedman CT method requires the glenoid to be perpendicular to the slice, because 20° of scapular malrotation alone can shift the reading by 10.5°.
Postoperative CT is limited by metal artifact from the prosthetic head, which is why interobserver agreement dropped in the arthroplasty group, though it still beat plain films.
This retrospective study compared glenoid version measured on axillary radiographs versus CT in 50 shoulders. Half were evaluated for instability and half had a total shoulder prosthesis. Three independent observers assessed each image to test which method is reliable.
When you need to know glenoid version, order a CT and do not trust the axillary radiograph.
This matters most in two settings the paper studied: working up posterior or recurrent instability, and planning or auditing glenoid component position in total shoulder arthroplasty. Excessive retroversion drives abnormal glenoid loading and poor arthroplasty outcomes, so getting the number right changes surgical decisions.
The reason axillary films fail is anatomic. The neck blocks medial film placement, so the scapular reference line follows the neck rather than the true blade axis, adding about 6.4° of false retroversion. A practical pearl: the Friedman CT method requires the glenoid to be perpendicular to the slice, because 20° of scapular malrotation alone can shift the reading by 10.5°.
Postoperative CT is limited by metal artifact from the prosthetic head, which is why interobserver agreement dropped in the arthroplasty group, though it still beat plain films.