Clinical Orthopaedics and Related Research ®

A Publication of The Association of Bone and Joint Surgeons ®

What Are the Complications, Survival, and Outcomes After Revision to Reverse Shoulder Arthroplasty in Patients Older Than 80 Years?

Eduard Alentorn-Geli MD, Nicholas J. Clark MD, Andrew T. Assenmacher MD, Brian T. Samuelsen MD, Joaquín Sánchez-Sotelo MD, Robert H. Cofield MD, John W. Sperling MD



By the time patients with a failed shoulder arthroplasty require revision surgery, a substantial number are older than 80 years. The risk of complications of revision arthroplasty in this elderly population is largely unknown and needs to be considered when contemplating whether these patients are too frail for revision surgery.


(1) What are the 90-day medical and surgical complications after revision to reverse shoulder arthroplasty (RSA) in patients older than 80 years? (2) What are the 2- and 5-year survival rates after revision? (3) Was there an improvement in pain at rest or with activity, range of motion (ROM), and strength after revision surgery?


Between 2004 and 2013, 38 patients who were older than 80 years (84 ± 3 years) underwent revision surgery to a RSA. Of those, five were lost to followup before 2 years, and two had died within 2 years of revision surgery, leaving 31 for analysis of our survivorship, pain, ROM, and strength endpoints at a minimum of 2 years or until revision surgery had occurred (mean, 28 months; range, 1–77 months); all 38 patients were included for purposes of evaluating medical and surgical complications at 90 days. During the period in question, our general indication for using RSA included failure of previous shoulder arthroplasty because of instability, glenoid loosening with bone loss, or rotator cuff insufficiency. The indication for revision to RSA did not change during the study period. The index procedure (revision to RSA at the age of 80 years or older) was the first revision arthroplasty in 33 (87%) patients and the second in five (13%) patients. We tallied 90-day medical and surgical complications by performing a retrospective chart and institutional joint registry review. The cumulative incidence of implant loosening (implant migration or tilting, or complete radiolucent lines present) and revision surgery was calculated at 2 and 5 years using competing risk of death method. Pain levels at rest or with activity (rated in a 1 to 5 Likert-type scale) were collected through a retrospective chart review and values before and after surgery were compared.


Medical complications occurred in three of 38 (8%) patients and surgical complications occurred in five of 38 (13%) patients. The 90-day mortality was 3% (one of 38 patients), and the total mortality was 26% (10 of 38 patients). The cumulative incidence of revision was 11% (95% CI, 0%–20%) at 2 years and 16% (95% CI, 1%–30%) at 5 years; the cumulative incidence of loosening was 8% (95% CI, 0%–20%) at 2 years and 16% (95% CI, 1%–30%) at 5 years. Pain at rest or with activity improved from pre- to postoperation (preoperative: median, 4 [range, 2–5]; postoperative: median, 1 [range, 1–4]; median difference: -2, 95% CI -3 to 0; p < 0.000). The active ROM improved during the preoperative compared with postoperative periods: mean ± SD forward flexion of 52° ± 40° to 109° ± 44°, respectively (mean difference: 56; 95% CI, 40–72; p < 0.000), and mean ± SD external rotation of 15° ± 22° to 31° ± 21°, respectively (mean difference: 16; 95% CI, 8–25; p < 0.000).


Age should not be used as a reason to not consider revision surgery to RSA in patients older than 80 years. Further studies with a prospective design, larger sample size, investigating risk factors for complications or poor outcome, and incorporation of functional scores are required.

Level of Evidence

Level IV, therapeutic study.

Back to top