Predictors of medium-term clinical outcomes after total shoulder arthroplasty

The study aims to measure the relationship of potential clinical and radiographic predictors and medium-term clinical outcomes after total shoulder arthroplasty in primary osteoarthritis.

In a prospective follow-up study, preoperatively collected clinical and radiographic variables were recorded in 103 patients with an average age of 66 years (range 37-83 years) to measure their influence on the clinical outcome by use of the Constant score. The average length of follow-up was 6 years (range 3-12 years). Radiographs and a computed tomography (CT) of the shoulders were obtained preoperatively for evaluating the acromiohumeral distance, the lateral glenohumeral offset and the morphology of the glenoid, which was classified according to the method by Walch (type-A1 to type-C glenoids).

The mean Constant score improved from its preoperative value of 25 points (range 6-54 points) to 65 points (range 10-86 points) postoperatively (p < 0.001). The mean age- and sex-normalized Constant score was similarly improved from 34 points (range 8-78 points) preoperatively to 90 points (range 14-130 points) at the time of follow-up (p < 0.001). The Walch classification of glenoid wear had a significant negative effect (r = -0.32, p < 0.001). The other predictors showed no significant influence on the Constant score (p > 0.05).

A higher Walch classification is a negative predictor for the postoperative clinical function. The proven significant negative effect  on outcomes in total shoulder arthroplasty emphasizes the importance of the preoperative evaluation of humeral head subluxation and glenoid erosion, which are associated with less favorable postoperative results. The measured internal patient variables gave no negative predictions on the medium-term clinical outcomes in the study population.