Does long-term follow-up and monitoring of primary shoulder arthroplasty identify failing implants requiring revision?

Abstract

Published scoping review has identified evidence paucity related to long-term follow-up of shoulder arthroplasty. We aim to report effectiveness of elective primary shoulder arthroplasty surveillance in identifying failing implants requiring revision. A prospective database recording shoulder arthroplasty and subsequent follow-up surveillance in a shoulder unit was analyzed. Shoulder arthroplasty was performed by 4 fellowship-trained shoulder surgeons for accepted elective indications including the use of anatomic arthroplasty in arthritic shoulders with intact rotator cuff and a reverse prosthesis being used in rotator cuff-deficient shoulders and rotator cuff-competent arthritic shoulders when deemed preferable by the treating surgeon. All shoulder arthroplasty implants used had achieved a minimum 7A Orthopaedic Data Evaluation Panel (ODEP) rating. The included shoulder arthroplasties were performed between May 1, 2004, and December 31, 2021, with minimum 1-year follow-up. Surveillance program involves specialist physiotherapist review at 1, 2, 3, 5, 8, 10, and 15 years postoperatively, including clinical examination, outcome scoring, and radiographs. Patient-initiated review occurred between time points if a patient requested assessment because of symptoms. Outcome measures include ratio of failing implants identified by surveillance and patient-initiated review, with number of surveillance reviews offered and proportion that identified a failing implant requiring revision calculated. A total of 1002 elective primary shoulder arthroplasty with minimum 1-year follow-up were performed (547 reverse total shoulder arthroplasty [rTSA], 234 anatomic total shoulder arthroplasty [aTSA], and 221 hemiarthroplasty [HA]). A total of 238 patients died prior to December 31, 2022, resulting in 4019 surveillance appointments offered. Thirty-eight prostheses required revision ≥1 year postoperatively (6 rTSA, 9 aTSA, and 23 HA), with surveillance identifying requirement in 53% (33% rTSA, 56% aTSA, and 57% HA) and patient-initiated review in 47%. Mean years from implantation to revision was 5.2 (2.7 rTSA, 3.6 aTSA, and 6.6 HA). Revision indications included rotator cuff failure (56% aTSR and 43% HA) and glenoid erosion (57% HA). This is the first series reporting effectiveness of shoulder arthroplasty surveillance in identifying implants requiring revision. Surveillance identified more than half of implants requiring revision, although only 0.5% of appointments identified revision requirement. Surveillance enrolment may influence patient-initiated review utilization; therefore, similar studies using only patient-initiated follow-up would help inform recommendations.