When patients require reoperation after primary shoulder arthroplasty, revision reverse total shoulder arthroplasty (rTSA) is most commonly performed. However, defining clinically important improvement in these patients is challenging because benchmarks have not been previously defined. Furthermore, while the minimal clinically important difference (MCID) and substantial clinical benefit (SCB) are commonly utilized to assess clinically relevant success, these metrics are limited by ceiling effects which may cause inaccurate estimates of patient success. Our purpose was to define the minimal and substantial clinically important percentage of maximal possible improvement (MCI-%MPI and SCI-%MPI) for commonly utilized pain and functional outcome scores after revision rTSA and to quantify the proportion of patients achieving clinically relevant success.
This retrospective cohort study used a prospectively collected single-institution database of patients undergoing first revision rTSA between August 2015 and December 2019. Patients with a diagnosis of periprosthetic fracture or infection were excluded. Outcomes scores included the ASES, raw and normalized Constant, SPADI, SST, and UCLA scores. We utilized an anchor-based method to calculate the MCI-%MPI and SCI-%MPI. Additionally, we calculated the MCI-%MPI using a distribution-based method for historical comparison. The proportions of patients achieving each threshold were assessed. The influence of sex, type of primary shoulder arthroplasty, and reason for revision rTSA were also assessed by calculating cohort-specific thresholds.
Ninety-three revision rTSAs with minimum 2-year follow-up were evaluated. Mean age was 67 years, 56% were female, and average follow-up was 54 months. Revision rTSA was performed most commonly for failed anatomic TSA (n=47), followed by hemiarthroplasty (n=21), rTSA (n=15), and humeral head resurfacing (n=10). The indication for revision rTSA was most commonly glenoid loosening (n=24), followed by rotator cuff failure (n=23), subluxation and unexplained pain (n=11 for both). The anchor-based MCI-%MPI thresholds (% of patients achieving) were: ASES=33% (49%), raw Constant=23% (64%), normalized Constant=30% (61%), UCLA=51% (53%), SST=26% (68%), SPADI=29% (58%). The anchor-based SCI-%MPI thresholds (% of patients achieving) were: ASES=55% (31%), raw Constant=41% (27%), normalized Constant=52% (22%), UCLA=66% (37%), SST=74% (25%), SPADI=49% (34%).
This study is the first to establish thresholds for the MCI-%MPI and SCI-%MPI at minimum 2-years after revision rTSA, providing physicians an evidence-based method to assess patient outcomes postoperatively.

Copyright © 2023. Published by Elsevier Inc.