LAUSR

BACKGROUND The purpose of this study was to compare long-term outcomes, complications, and reoperation rates of primary reverse total shoulder arthroplasty (RTSA) performed at a single institution using 2 implant designs: a Grammont medialized prosthesis (medialized [M] group) and a Frankle glenoid-based lateralized prosthesis (glenoid-lateralized [GL] group). METHODS Between 2004 and 2008, 100 consecutive single-institution primary RTSAs were performed by reconstructive shoulder surgeons who were not design consultants, with the aim of obtaining 10-year follow-up: 56 in the M group and 44 in the GL group. Patients were followed up until death, until revision surgery, or for a minimum of 10 years. RESULTS Of 100 patients, 87 had more than 2 years' follow-up (mean, 77 months). A subset analysis of 41 patients with an average of 10.2 years' follow-up showed sustained long-term outcomes. RTSA provided clinical improvements without significant differences between the M and GL groups, except for improved active forward elevation in the M group (144° in M group vs. 115° in GL group, P = .002). Reoperation was required in 6 shoulders (10-year cumulative incidence of 3 [5%] in M group vs. 3 [8%] in GL group) for a total of 16 complications (10-year cumulative incidence of 8 [14%] in M group vs. 8 [20%] in GL group). Notching rates were significantly higher in the M group (77% in M group vs. 47% in GL group, P = .013); differences in severe notching (grade 3 or 4) were clinically relevant but did not reach statistical significance (23% in M group vs. 9% in GL group, P = .22). CONCLUSION Primary RTSA using these first 2 prosthesis designs was associated with good outcomes and low reoperation (5%-8%) and complication (14%-20%) rates at 10 years. The M group had higher rates of notching. These results may provide a benchmark for comparison with newer implants, especially considering that these results include the early RTSA implantation learning curve.