Patient race and ethnicity are associated with higher unplanned 90-day emergency department visits and readmissions but not 10-year all-cause complications or reoperations: a matched cohort analysis of primary shoulder arthroplasties

Abstract

Background

Within orthopedic surgery, there remain limited data evaluating the impact of racial and socioeconomic disparities on outcomes of primary shoulder arthroplasty (SA) over time. As such, we analyzed both short- and longer-term outcome differences in complications, reoperations, and revision surgery of primary SA when performed in non-White patients when compared to a matched cohort of White patients who had undergone SA.

Methods

Over a 39-year period (1981-2020), an institutional Total Joint Registry Database was utilized to identify all non-White patients (Asian/Pacific Islander, Black, Hispanic or Latino, American Indian/Alaska Native, other) who underwent primary SA with a minimum of 2 years of follow-up. The search identified 275 primary SA (46 hemiarthroplasties, 97 anatomic total shoulder arthroplasties, and 132 reverse total shoulder arthroplasties). The ethnicity composition was 8.7% Asian, 27.3% Black, 37.8% Hispanic, 12.4% American Indian, and 13.8% other. This cohort was matched 1:2 according to age, sex, diagnosis, implant, and surgical year to a control group of 550 White patients who had undergone SA. The rates of medical and surgical complications, reoperations, revisions, and implant survivorship were assessed. The mean follow-up time was 6.3 years (range, 2 to 40 years).

Results

Comparisons between the non-White and White matched cohorts demonstrated a higher rate of tobacco use (14.2% vs. 10.5%; P < .001), diabetes (21.5% vs. 11.8%; P < .001), length of stay (1.9 vs. 1.6 days; P = .014), and a lower rate of private commercial insurance (27.3% vs. 44.5%; P < .001 in the non-White cohort. Within the first 90 days after surgery, non-White patients had a higher rate of emergency department visits (5.5% vs. 0.9%; P < .001) and unplanned readmissions (2.9% vs. 0.7%; P = .014). After the first 90 postoperative days, there were no differences regarding medical (1.8% vs. 0.7%; P = .135) or surgical complications (12.0% vs. 13.6%; P = .446). Ten-year survivorship free of all-cause complication (76.8% vs. 81.5%; P = .370), reoperation (84.9% vs. 89.8%; P = .492), and revision (89.3% vs. 91.4%; P = .715) were similar between the non-White and White cohorts.

Discussion

After accounting for age, sex, and surgical indication, patient race and ethnicity were not associated with an increased risk of long-term all-cause complications, reoperations, or revision surgery after primary SA. However, within the first 90 postoperative days, non-White patients had a higher likelihood of unplanned emergency room visits and readmissions.