Optimal glenosphere size cannot be determined by patient height

Background Glenosphere size remains 1 surgeon-controlled variable that can affect patient outcomes following reverse shoulder arthroplasty (RSA). There remains no objective criterion to guide surgeons in choosing glenosphere size. This study's purpose was to evaluate range of motion (ROM) as a function of patient height and glenosphere size to determine the optimal glenosphere size based on patient height. Methods We retrospectively reviewed 589 primary RSAs from a multicenter shoulder arthroplasty database of a single RSA system with multiple glenosphere sizes. Shoulders were separated into groups based on glenosphere size (38 or 42 mm). Predictive accuracy was calculated in relation to height and sex for predicting glenosphere size. Improvements in active ROM and patient-reported outcome measures (PROMs) were compared based on glenosphere size as a function of height. Results Logistic regression analysis demonstrated a strong association of height and sex with surgeon selection of glenosphere size, with shorter heights preferentially treated with 38-mm glenospheres and taller heights with 42-mm glenospheres. There were no statistically significant interaction effects of glenosphere size and height on improvements in ROM or PROMs. These results indicate that for a given glenosphere size, there is not an optimal height range to maximize improvements in postoperative outcome measures. Discussion Height and sex are highly correlated with a surgeon's choice of glenosphere size. However, on the basis of improvements in ROM and PROMs, no recommendation can be made for surgeons to select a particular glenosphere size based on a patient's height. Surgeons should consider other variables when selecting a glenosphere size.