Reproducibility of 135° humeral neck–shaft angle in stemless reverse shoulder arthroplasty: a single-surgeon radiographic study

Reverse shoulder arthroplasty (RSA) is widely used for cuff tear arthropathy, osteoarthritis, and fractures. To enhance function and longevity, implant designs have evolved toward lateralized glenoids, reduced neck-shaft angles (NSA), and bone-preserving stemless components. Nevertheless, accurate humeral positioning at a 135° of NSA without a diaphyseal reference can be challenging. We retrospectively analyzed 26 consecutive stemless RSAs performed between 2019 and 2025. Two independent observers analyzed postoperative radiographs, measuring NSA and Deltoid Tuberosity Index (DTI). Inter- and intra-observer reliability was assessed. Associations between NSA and patient factors (age, sex, diagnosis, bone quality) were examined using regression analysis. Postoperative complications were recorded. The mean NSA was 133.1° ± 9.9° (SE 1.9), with a mean absolute error of 7.6° from the target 135° (range 110.4°–150.3°). Excellent inter- and intra-observer reliability was observed for NSA (ICC > 0.93), while DTI reliability was moderate (ICC = 0.57). Simple regression showed no association between NSA and DTI (R 2 = 0.002, p = 0.831). Multiple regression including age, sex, diagnosis, and DTI explained 21.1% of NSA variability but was not statistically significant (p = 0.150). Varus malpositioning was more frequent, with three cases with early radiolucency lines and one case of dislocation (3.8%). Stemless RSA achieved a mean NSA of 133° with a 7° error, unaffected by patient factors. Variability may be linked to other factors such as surgical technique and/or implant design. Attention to humeral cut accuracy and controlled implant impaction is essential to minimize risks of malalignment and fixation failure.