These authors used the average and incremental cost-effectiveness ratio (ACER and ICER) and the procedure value index (PVI) to examine differences in outcomes and value between TSA and RSA for treatment of glenohumeral osteoarthritis with an intact rotator cuff. They performed a retrospective matched-cohort study of patients treated with primary shoulder arthroplasty for osteoarthritis with an intact rotator cuff who had a minimum 2-year follow-up.
Patients treated with aTSA were matched 4:1 to those treated with RSA based on sex, age, and preoperative SST score.
Two hundred and fifty-two aTSA-treated patients were matched to 63 RSA-treated patients with no significant differences in sex, age, or preoperative SST score. Total hospitalization costs, charges, and reimbursements along with outcome improvements in units of minimal clinically important differences (MCIDs) and patient satisfaction did not differ between the groups. For RSA, the implant cost was significantly higher than that for aTSA, but the operating room, anesthesia, and cement costs were lower. The aTSA group had a 3.2% rate of gross glenoid loosening and a 2.4% revision rate at an average followup of 54 (24-122) months. There was no loosening or revision in the RSA group at an average of 38 (24-85) months.
None of the value analytics differed between groups even after inclusion of the outcomes and costs of early aTSA revisions. The patients having revision were 72, 80, 76, 77, 75, and 88 years of age at the time of revision and had an A2, B2, A1, A2, D, and B3 glenoid prior to the primary aTSA, respectively. The reasons for failure in these individual patients were (1) subscapularis tear, (2) massive rotator cuff tear, (3) massive rotator cuff tear, (4) subscapularis tear with gross loosening, (5) massive rotator cuff tear with gross loosening, and (6) infection with gross loosening. A second comparative value analysis including the outcomes and costs of these revisions resulted in no significant differences in any ACER or PVI between aTSA and RSA.
Comment: This is a well-done study showing that under ideal circumstances both aTSA and RSA can yield comparable value. The complication and reoperation rates in this study for both procedures were remarkably low.
The authors are thorough in their presentation of its limitations:
(1) These results are from a high volume shoulder arthroplasty surgeon at a high volume medical center. They may not be generalizable to lower volume practices.
In comparison to low surgeon volume, high surgeon volume can result in a number of factors that enter the value calculation
(a) lower negotiated costs of implants (sometimes <25% of the "sticker price")
(b) more refined patient selection (with consideration of patient health, motivation, social support and insurance status) resulting in fewer consultations and shorter hospital stays
(c) better surgical technique with fewer complications
(d) better operating room efficiency (lower operating room and anesthesia costs)
(e) consistent and experienced anesthesia, nursing, and therapy resulting in decreased complication rates, shorter lengths of hospital stay, and lower hospitalization costs
(2) There was inherent selection bias in this retrospective cohort analysis, as there were a variety of justifications for selecting RSA or aTSA. Patients were not randomized to RSA or aTSA and
decisions were based on surgeon and patient preference.
(3) The average followup duration was shorter for the RSA (38 months) patients than for the aTSA patients (54 months).
(4) The implants used for aTSA and RSA used are not presented.
