Revision shoulder arthroplasty: does the stem really matter?

Background: The management of a failed shoulder arthroplasty represents a complex and difficult problem for the treating surgeon, with potential difficulties and complications that are related to the need to remove a well-fixed stem. The aim of this study is to compare the intraoperative complications, postoperative complications, and outcome of revisions from stemmed arthroplasties (STAs) with those from surface replacement arthroplasties (SRAs). Methods: From 2005 to 2012, 40 consecutive revision shoulder arthroplasties were performed at our institute: 17 from STAs and 23 from SRAs. Perioperative events, operation time, blood loss, intraoperative fractures, and use of structural allograft were recorded. Clinical and radiologic outcomes were analyzed. Results: Operation time, need for humeral osteotomy, need for structural allograft, and number of intraoperative fractures were significantly higher in the STA group. Blood loss, drop in hemoglobin level, need for blood transfusion, and hospitalization time were also higher in the STA group, but these differences were not statistically significant. Reoperation was performed in 3 patients in the SRA group. A significant clinical improvement was observed in both groups. The Constant score was higher in the SRA group. Conclusion: Revision of STAs is a more demanding procedure. The postoperative complication rate was slightly higher in the SRA group. The group with revision of SRAs showed slightly better clinical and radiographic results, but there was no statistically significant difference between the groups. Level of evidence: Level III; Retrospective Cohort Design; Treatment Study (C) 2016 Journal of Shoulder and Elbow Surgery Board of Trustees.