Accuracy and reliability of postoperative radiographic measurements of glenoid anatomy and relationships in patients with total shoulder arthroplasty

Abstract

Radiographic imaging is the follow-up imaging modality most widely used for patients who have undergone total shoulder arthroplasty (TSA). However, its accuracy of measurement of component position has not been validated against a gold standard in a clinical series. Thirty-two x-ray images and computed tomography scans were taken within 1 month of each other in patients who had undergone TSA with an all-polyethylene glenoid component. The humeral glenoid alignment in the coronal superior-inferior (SI) plane (HGA-SI), humeral glenoid alignment in the axial anterior-posterior (AP) plane (HGA-AP), and humeral scapular alignment in the axial plane (HSA-AP) were measured with 21 pairs of images, and glenoid component retroversion was measured with all 32 pairs. Intraclass correlation coefficients (ICC) were calculated for HGA-SI, HGA-AP, HSA-AP, and version, and accuracy analysis criteria of the radiographs were assessed using predetermined criterion. We found fair-moderate agreement between x-ray images and CT scans for HGA-SI (ICC=0.42) and version (ICC=0.69), but poor agreement for HGA-AP (ICC=0.04) and HSA-AP (ICC=0.38). An average difference of overestimating HGA-SI by 0.06%±7.7%, with a precision 95% confidence interval of 7.6%, and overestimating version by -4.2° ± 5.1°, with a precision 95% confidence interval of 9.9°, was found. This validation study has defined the ability and limitation for these measurements using high-quality axillary and AP radiographs.