Race, gender, and income negatively impact patient-reported outcomes following total shoulder arthroplasty

Background

Social determinants of health (SDOH) refer to social and economic factors that influence a patient’s health status. The purpose of this study was to investigate the impact of SDOH on preoperative and postoperative Patient-Reported Outcomes Measurement Information System (PROMIS) scores and postoperative resource utilization following primary shoulder arthroplasty (SA).

Methods

This retrospective chart review evaluated data from all patients who underwent primary SA (including anatomic, reverse, and hemiarthroplasty) at a single health system between May 2020 and May 2022. Patients without at least 6-month postoperative PROMIS questionnaires and those undergoing revision surgeries were excluded. The electronic medical record was used to identify SDOH for each patient, and PROMIS scores for Upper Extremity (PROMIS-UE), Pain Interference (PROMIS-PI), and Depression (PROMIS-D) were completed electronically at respective preoperative and postoperative visits. Univariate analysis using independent 2-group t-tests and Chi-squared tests were used to analyze the mean difference between patient groups based on SDOH. Multivariate linear regressions were performed with all predictors used in the univariate model using the least squares method.

Results

The study included 248 patients who underwent SA, with a mean age of 67.9 years. Caucasian patients were over-represented in the highest quartile of median household income (MHI) compared to Black patients (35.1% vs. 17.2%) who were over-represented in the lowest MHI quartile (37.9% vs. 8.3%). At 6-month postoperative, black patients had significantly lower UE (33.8 ± 6.2 vs. 38.1 ± 9.0; P = .03) and greater PI scores (59.1 ± 6.1 vs. 55.6 ± 8.6; P = .145) compared to Caucasian patients. Similarly, at 6-month follow-up, the lowest MHI quartile had lower UE (33.8 ± 7.7 vs. 39.6 ± 8.8; P = .01) and higher PI scores (58.7 ± 6.5 vs. 54.3 ± 8.2; P < .01) compared to the highest MHI quartile, and females demonstrated lower UE (36.3 ± 7.9 vs. 38.6 ± 9.8; P = .04) and higher D scores (46.2 ± 9.1 vs. 42.0 ± 8.6; P = .046) compared to males. Government/public insurance demonstrated lower UE (36.8 ± 8.0 vs. 39.8 ± 10.4; P = .03) and higher D scores (45.9 ± 9.2 vs. 40.6 ± 7.6; P = .03) compared to private insurance. At 12-month follow-up, females demonstrated lower UE scores compared to males (36.0 ± 10.2 vs. 40.1 ± 11.3; P = .03). Black patients harbored lower UE, PI, and D scores compared to Caucasian patients, although not statistically significant.

Conclusion

Several socioeconomic factors such as race, gender, and insurance status are associated with differential outcomes after SA. Patients who are Black, female, current smokers, and from the lowest income quartile are associated with inferior PROMIS function, pain, and depression outcomes following SA.