Which shoulder functions correlate with patient satisfaction after primary shoulder arthroplasty?

BACKGROUND

A principal goal of shoulder arthroplasty is to achieve patient satisfaction. While patient self-assessment of shoulder function is commonly used to assess operative success, it is not known how improvements in shoulder function correlate with patient satisfaction after primary anatomic total shoulder arthroplasty (aTSA), reverse total shoulder arthroplasty (rTSA), and ream-and-run (RnR) arthroplasty. The objectives of this study were (1) to determine the preoperative to postoperative improvement in the individual functions assessed by the Simple Shoulder Test (SST) for patients having aTSA, rTSA, and RnR arthroplasty and (2) to relate these improvements to patient satisfaction after each of these 3 arthroplasty types.

METHODS

We conducted a retrospective cohort study of a longitudinally maintained arthroplasty database from a single academic center. Patients undergoing aTSA, rTSA, and RnR arthroplasty with a minimum of 2-year follow-up data were included for analysis. Patients having arthroplasty for fracture and revision arthroplasty were not included. Preoperative SST scores and patient satisfaction questionnaires were obtained and compared with 2-year postoperative data. Patients who improved from neutral or negative satisfaction scores to positive scores on the questionnaire were determined to be "satisfied," while others were "unsatisfied." Individual SST questions were analyzed by arthroplasty type with receiver operator characteristic curves to determine their relationship to patient satisfaction.

RESULTS

In total, 1,048 patients were included. Of them, 468 (44.7%) underwent aTSA, 164 (15.6%) rTSA, and 416 (39.7%) RnR arthroplasty. Patients undergoing aTSA and RnR arthroplasty had higher rates of overall satisfaction at 2 years compared with rTSA patients (83.5% and 80.5% compared with 68.9%, respectively, P < .001). Shoulder comfort while sleeping demonstrated the highest or second Youden index (metric of the best-fit curve based on maximization of sensitivity and specificity) across all 3 arthroplasty types. Shoulder comfort at rest demonstrated high positive predictive value in all cohorts. The relationship of the other SST functions to patient satisfaction varied among the 3 arthroplasty types.

CONCLUSION

For all shoulder arthroplasty types, patient comfort with the arm at rest and the ability to sleep comfortably were strongly related to patient satisfaction at 2 years after surgery. The relationship between satisfaction and the other shoulder functions varied among the different types of arthroplasties. The results of this study may be useful in setting expectations for patients having shoulder joint replacement.