168. A comparative analysis of racial disparities in nationally derived hospital data and two prospective multicenter surgical databases of adult spinal deformity surgery

Abstract

<h3>BACKGROUND CONTEXT</h3> While past research has shown racial disparities exist in the outcomes and utilization of spine surgery, those studies were often performed on nondisease-specific national databases, which may capture different patient populations and complication rates compared to prospective multicenter databases. <h3>PURPOSE</h3> Purpose: 1) Assess racial disparities for enrollment in two prospective multicenter databases (ASD-DB1 and ASD-DB2). 2) Assess racial disparities in demographics, outcomes, in three databases: ASD-DB1, ASD-DB2, and the National Surgical Quality Improvement Program (NSQIP). We hypothesize that race will not affect enrollment in prospective multicenter studies. Disparities in national databases will be different than those in prospective databases. <h3>STUDY DESIGN/SETTING</h3> Retrospective review. <h3>PATIENT SAMPLE</h3> ASD patients in ASD-DB1, ASD-DB2, and NSQIP. <h3>OUTCOME MEASURES</h3> Preoperative demographic and clinical characteristics, intraoperative surgical techniques utilized, postoperative complications, correction, and change in HRQOLs. <h3>METHODS</h3> ASD-DB1 (2-year follow-up) included ASD patients from 2008-2019. ASD-DB2 (6-week follow-up) included ASD patients from 2019-2021. NSQIP (30-day follow-up) included ASD patients from 2012-2018. Racial disparities were assessed in African Americans (AA), Asians (AN) and Hispanics (H), compared to Caucasians (CA). Enrollment racial disparities were evaluated in ASD-DB1 and ASD-DB2. <h3>RESULTS</h3> ASD-DB1: 1,356 patients. Compared to CA, AA and H were younger, AA had higher max Cobb and higher rates of diabetes, H had lower SRS Activity, Mental, and Total, and AN had higher SRS Pain (all p < 0.05). The counties that enrollment sites for ASD-DB1 were located in were overall 66% CA, 18% AA, 16% H, and 11% AN. ASD-DB1 was 91% CA, 5% AA, 0% H, 1% AN. ASD-DB2: 338 patients. Compared to CA, AAs were younger and more alone (p < 0.05 all). The counties that enrollment sites for ASD-DB2 were located in were 66% CA, 18% AA, 16% AA, and 11% AN. ASD-DB2 was 91% CA, 5% AA, 0% H, and 1% AN. On multivariable analysis (MVA), no disparity in complications, recovery in HRQOLs, surgical techniques, or correction were found in ASD-DB1 or ASD-DB2 (p GT 0.05 for all). NSQIP had 2,660 patients. AAs were younger and had higher rate of ASA Class 4, dependence and diabetes (all p< 0.05). On MVA, AAs more reoperation (OR 2.2; P=0.005), greater than 5 days length of stay (OR 1.63; P=0.010), septic shock (OR 10; P=0.000), renal failure (OR 6.5; P=0.022) and deep wound infection (OR 2.77;P=0.048) (all p<0.05). <h3>CONCLUSIONS</h3> All three databases found disparities in baseline patient presentation. Both prospective databases predominantly enrolled white patients compared to county demographics. Neither of the prospective databases found racial disparities in complications, HRQOLs, radiographic correction, or techniques utilized. The national database found racial disparities in 30-day postoperative outcomes. This suggests there exist differences in patient populations and data granularity between nationally collected and prospective databases. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.