Abstract 164: Neighborhood Socioeconomic Status is Significantly Associated with Outcomes After Coronary Artery Bypass Grafting

Abstract

Objective: Area Deprivation Index (ADI) is a geographically-derived measure of socioeconomic status composed of 17 education, employment, housing-quality, and poverty measures derived from census data. Block group level ADI has recently been made widely accessible to clinicians. A block group is the smallest geographical unit from which census data is collected, representing a homogenous group of people living on a single city block or rural equivalent. Previous work analyzing the relationship of socioeconomic status with cardiac surgery outcomes has been limited by the use of zip codes as the geographic unit, which comprise a larger and more diverse population. This study evaluates the association of ADI with outcomes after coronary artery bypass grafting (CABG). Methods: Patients who underwent isolated CABG in the University of Pennsylvania health system between 2002 and 2018 with a valid home address were assigned an ADI based on their corresponding 2010 Census Block Group. The association between ADI and each STS post-operative outcome was assessed with univariable analysis using a Wilcoxon rank sum test with continuity correction. A multivariable logistic regression model was created for each STS post-operative outcome using variables in the 2008 STS risk model for CABG with ADI as an additional variable. Results: The 6,751 patients who met the inclusion criteria were from 19 states with a median predicted STS risk of post-operative major morbidity and mortality of 16.6% (IQR 7.0% to 19.9%) and median ADI of 31 (IQR 17 to 50). ADI was not normally distributed by Wilcoxon signed rank test (p < 0.01). ADI was significantly associated with stroke (p = 0.03), renal failure (p < 0.01), prolonged ventilation (p < 0.01), reoperation (p = 0.05), and composite morbidity and mortality (p < 0.01) in univariable analysis. When included in a multivariable logistic regression model ADI was significantly associated with renal failure (p = 0.02), prolonged ventilation (p < 0.01), and composite morbidity and mortality (p < 0.01). Conclusions: ADI is associated with STS post-operative outcomes. When included in a multivariable model, ADI remains highly associated with outcomes. Inclusion of block group level socioeconomic status in prediction models may significantly improve risk prediction.