The 2021 Australian Orthopaedic Association National Joint Replacement Registry report indicated that total shoulder replacement using both mid head (TMH) length humeral components and reverse arthroplasty (RTSA) had a lower revision rate than stemmed humeral components in anatomical total shoulder arthroplasty (aTSA) - for all prosthesis types and diagnoses. The aim of this study was to assess the impact of component variables in the various primary total arthroplasty alternatives for osteoarthritis in the shoulder.

Data from a large national arthroplasty registry were analysed for the period April 2004 to December 2020. The study population included all primary aTSA, RTSA, and TMH shoulder arthroplasty procedures undertaken for osteoarthritis (OA) using either cross-linked polyethylene (XLPE) or non-cross-linked polyethylene (non XLPE). Due to the previously documented and reported higher revision rate compared to other anatomical total shoulder replacement options, those using a cementless metal backed glenoid components were excluded. The rate of revision was determined by Kaplan-Meir estimates, with comparisons by Cox proportional hazard models. Reasons for revision were also assessed.

For a primary diagnosis of OA, aTSA with a cemented XLPE glenoid component had the lowest revision rate with a 12-year cumulative revision rate of 4.7%, compared to aTSA with cemented non-XLPE glenoid component of 8.7%, and RTSA of 6.8%. The revision rate for TMH was lower than aTSA with cemented non-XLPE, but was similar to the other implants at the same length of follow-up. The reason for revision for cemented aTSR was most commonly component loosening, not rotator cuff deficiency.

Long stem humeral components matched with XLPE in aTSA achieve a lower revision rate compared to shorter stems, long stems with conventional polyethylene, and RTSA when used to treat shoulder OA. In all these cohorts, loosening, not rotator cuff failure was the most common diagnosis for revision.