Black and Hispanic Patients Do Not Stay Longer After Primary Total Knee Arthroplasty: Results From an Urban Center Serving a Predominantly Minority Cohort

Abstract

Introduction: Previous reports identified minority race/ethnicity to be an independent risk factor for prolonged length of stay (LOS); however, these cohorts consisted of predominantly White patients. This study sought to evaluate minority status as an independent risk factor for prolonged LOS after primary total knee arthroplasty (TKA) in a predominantly Hispanic and Black cohort. Methods: This was a retrospective study using an institutional database of patients who underwent primary TKA between the years 2016 and 2019. Demographic and socioeconomic data, smoking, body mass index (BMI), medical comorbidities, discharge disposition, and 30-day readmission rates were collected. Patients were first categorized into racial/ethnic groups (Hispanic, Black, or White). An univariate analysis was performed comparing patient characteristics between racial/ethnic groups using the Wilcoxon rank sum, chisquared, and Fisher exact tests. We then categorized patients into two groups-normal LOS (discharged on postoperative day 1 to 2) and prolonged LOS (discharged after postoperative day 2). An univariate analysis was again performed comparing patient characteristics between LOS groups using Wilcoxon rank sum, chi-squared, and Fisher exact tests. After identifying risk factors markedly associated with LOS, a multivariate logistic regression analysis was performed to identify independent risk factors for prolonged LOS. Results: A total of 3,093 patients were included-47.9% Hispanic and 38.3% Black. Mean LOS was 2.9 6 1.6 days. An univariate analysis found race/ethnicity, age, low socioeconomic status (SES), discharge disposition, insurance type, weekday of surgery, BMI .40, smoking, increased American Society of Anesthesiologists (ASA)/Charlson Comorbidity Index (CCI) and several medical comorbidities to be associated with prolonged LOS (P , 0.05). A multivariate logistic