P146. The impact of race/ethnicity on perioperative outcome measures: a longitudinal analysis of patients undergoing lumbar fusion surgery in the United States

Abstract

<h3>BACKGROUND CONTEXT</h3> Previous studies have described racial and ethnic disparities in outcomes after lumbar fusion surgery. It is unclear to what extent these disparities have changed over time among patients undergoing lumbar fusion surgery in the United States (US). <h3>PURPOSE</h3> Evaluate racial/ethnic disparities for lumbar fusion surgery outcomes as measured by length of hospital stay in days (LOS) and discharge disposition (DD). <h3>STUDY DESIGN/SETTING</h3> This retrospective study examined US discharge summaries compiled by the National Inpatient Sample from 2003-2018. <h3>PATIENT SAMPLE</h3> Discharge summaries representing 2,438,062 Black non-Hispanic, White non-Hispanic, and Hispanic adult inpatients with a diagnosis of lumbar stenosis, disc herniation/degeneration, spondylosis and/or spondylolisthesis who underwent a lumbar fusion were included. <h3>OUTCOME MEASURES</h3> The primary outcomes were LOS and DD for each hospitalization, stratified by race/ethnicity. DD was classified as ‘routine' (reference), or ‘non-routine' which included ‘Transfer to Short-term Hospital' (TSTH), ‘Transfer to Other [Facility]' (TOF) (including Skilled Nursing Facilities and Intermediate Care Facilities), and discharge to ‘Home Health Care' (HHC). <h3>METHODS</h3> A multivariable generalized linear model with a gamma distribution and a multinomial logistic regression model were utilized to measure the association between race/ethnicity and LOS/DD, respectively. Models were adjusted for gender, age, primary payor, zip code-specific income quartile, geographical region, Deyo-Charlson Comorbidity Index, year of the procedure, hospital bed size and location/setting of the procedure. To assess racial/ethnic disparities over time, an interaction term between race/ethnicity and year was introduced. Unstandardized coefficients (B), odds ratios (OR) and 95% confidence intervals (95% CI) are reported. <h3>RESULTS</h3> Mean LOS for White, Black, and Hispanic patients declined from 4.4, 4.8 and 4.6 days in 2003 to 3.7, 4.4 and 4.0 days in 2018, respectively. After adjustment for relevant covariates, Black (B:1.097; 95% CI:1.095-1.100) and Hispanic (B:1.056; 95% CI:1.053-1.059) patients had significantly longer LOS compared to White patients (Reference B:1.000) when analyzing the 16-year study period. When evaluating DD, non-routine discharges for White and Hispanic patients increased from 27%, 35% and 25% in 2003 to 39%, 54% and 43% in 2018, respectively. Black patients had increased odds of non-routine discharges, including TSTH (OR:2.029, 95% CI:1.895-2.172), TOF (OR:1.854, 95% CI:1.826-1.881) and discharge to HHC (OR:1.420, 95% CI:1.401-1.439) when compared to White patients (Reference OR:1.000). Similarly, Hispanic patients also experienced increased odds of TSTH (OR:1.213, 95% CI:1.100-1.337), TOF (OR:1.399, 95% CI:1.374-1.424), and discharge to HHC (OR:1.371, 95% CI:1.351-1.392) compared with White patients. Utilizing the interaction term between race/ethnicity and year, we identified that from 2009 to 2018 Black patients progressively experienced increasing odds of TOF (2009 OR:0.609, 95% CI:0.560-0.663; 2018 Reference OR:1.000) and discharge to HHC (2009 OR:0.630, 95% CI:0.585-0.679). <h3>CONCLUSIONS</h3> Significant racial/ethnic disparities in LOS persisted within the context of falling mean LOS for all racial/ethnic groups. The percentage of patients experiencing non-routine DD grew for all racial/ethnic groups over the study period, however, from 2009 to 2018 increases in non-routine DD for Black patients outpaced those of White patients, resulting in increased levels of disparities. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.