Reliability of Implant- vs. Bone-Based Postoperative Measurements After Anatomic Total Shoulder Arthroplasty Using CT...

In anatomic total shoulder arthroplasty (aTSA) research, postoperative anatomical measurements are often referenced to the reamed glenoid bone surface because the glenoid component is not fully visible on CT. However, the reliability of measurements based directly on the implant (implant-based method) and their agreement with the commonly used bone-based method remain unclear. In this study, CT scans from 29 patients undergoing primary aTSA were used to calculate postoperative glenoid positioning (translations, version, inclination) and HHS in the anterior-posterior (HHS_AP) and superior-inferior (HHS_SI) directions using both methods. The bone-based method referenced measurements to 3D postoperative scapular bone reconstructions, whereas the implant-based method used CT-registered digital glenoid component models. Inter- and intra-rater reliability of implant-based measurements were assessed using intraclass correlation coefficients (ICCs), and agreement between methods was evaluated using ICCs and Bland-Altman analyses. The implant-based approach demonstrated high inter- and intra-rater reliability (ICCs > 0.88), with minimal detectable changes below commonly cited clinical thresholds. The agreement between implant- and bone-based methods was moderate-to-excellent overall (ICC > 0.52), although significant bias and/or wide limits of agreement were observed for version, inclination, superior-inferior implant positioning, and HHS_SI. Compared to implant-based measures, bone-based estimates showed a bias toward more superior implant positioning and superior HHS_SI, greater retroversion, and more medialized implant positioning. In conclusion, direct implant-to-CT registration is a reliable method for 3D postoperative measurements after aTSA. Implant- and bone-based measurements can be used interchangeably for anterior-posterior implant positioning and HHS_AP, but not for other measures when high precision is required. LEVEL OF EVIDENCE: Basic science study.