Do shoulder kinematics change after reverse total shoulder arthroplasty?

Reverse total shoulder arthroplasty (rTSA) is believed to alter shoulder kinematics. However, it remains unclear whether the kinematic changes observed in shoulders following rTSA were present preoperatively or developed as a result of the surgery. Furthermore, the impact of preoperative scapulohumeral rhythm (SHR) on the postoperative SHR after rTSA is also poorly understood. The primary aim of this study was to compare shoulder kinematics before and after rTSA using the 3-dimensional (3D) to 2-dimensional model-image registration with dynamic digital radiography images. The secondary purpose was to evaluate whether preoperative SHR correlates with postoperative SHR after rTSA.

Twenty shoulders in 19 patients that underwent rTSA were included. Dynamic digital radiography images were performed preoperatively and 6 months or later postoperatively to assess shoulder motion during scapular plane abduction. 3D surface models of preoperative scapula and humerus, and postoperative scapula and humerus with implants were created from the preoperative and postoperative computed tomography images. Scapular and humeral kinematics were evaluated using a 3D to 2-dimensional model image registration method. Each kinematics parameter was calculated at 10° increments from 20° to 90° of humeral abduction. The mean SHR was also calculated from the humeral position at side to 90° or to the maximum humeral abduction if it was less than 90°. Linear mixed-effects models for repeated measures were used to compare kinematics at each humeral abduction between preoperative and postoperative conditions. The correlation between the preoperative mean SHR and the postoperative mean SHR was assessed using Pearson's product-moment correlation statistic.

A significant postoperative increase in scapular posterior tilt was observed (P = .019). However, changes in SHR, scapular upward rotation, or scapular and humeral internal/external rotation from preoperative to postoperative conditions were not statistically significant. Additionally, a significant correlation was not detected between the preoperative and postoperative mean SHR.

rTSA significantly increases scapular posterior tilt postoperatively compared to the preoperative condition. However, significant changes from the preoperative to postoperative conditions were not demonstrated for other kinematic parameters.