Failure of total shoulder arthroplasty: why does it occur?

From a historical perspective, shoulder arthroplasty has evolved significantly from its inception. Much like arthroplasty of the lower extremities, shoulder implant design had its roots in a constrained device. Unfortunately, the inherently high loads that these devices generated across the implant articulation and the glenoid bone interface resulted in an unacceptably high failure rate. Beginning with Neer's original hemiarthroplasty of the 1950s and the more recent implant designs, there has been a strong trend toward less constraint and a greater emphasis on soft tissue preservation in shoulder replacement. In the ideal arthroplasty patient with an intact rotator cuff and a stable, yet degenerative, glenohumeral joint, arthroplasty invariably yields good-to-excellent results 90% of the time. However, owing to the enormous functional range of motion of the joint and its inherent dependence on soft tissues for both stability and motion, there are many areas for potential complications. This article addresses the etiology, recognition, and treatment of these problems.