Abstract 4143228: Uncovering Hidden Disparities in Surgical Mortality among Texas Medicaid-Insured Infants with Critical Congenital Heart Disease

Background: Data shows that publicly insured patients with critical congenital heart disease (CCHD) have higher surgical mortality than privately insured patients. However, little is known about disparities within Medicaid-insured cohorts. We hypothesized that disparities in surgical mortality exist within the Medicaid population, with the highest mortality being among the most socially vulnerable patients. Methods: We conducted a retrospective cohort study using Texas Birth Defects Registry records of infants with CCHD born 2010-2014 who underwent cardiac surgery <1 year old (CCHDs). Patients with CCHD as defined by the Centers for Disease Control or ductal dependent anatomy were included, and records for study inclusion were reviewed by a pediatric cardiologist. Cases were linked with death certificates and the Medicaid database, and Medicaid patients were stratified into Primary (PM) and Secondary (SM) cohorts. The PM cohort was comprised of patients with the highest scores (most socioeconomically vulnerable) based on the census tract derived Social Vulnerability Index (SVI). The primary predictor was race/ethnicity and primary outcome was infant mortality. Descriptive sociodemographics were analyzed using chi-square. Infant mortality was analyzed using stratified univariable Cox Proportional-Hazards models. Results: We identified 2525 infants with CCHD S , of whom 14.5% had PM and 56.9% had SM. The Medicaid-cohort had more single ventricle, chromosomal, and extracardiac defects than the Non-Medicaid (NM) cohort (p<0.0001). Compared to the SM cohort, CCHD S infants with PM were more likely to be born preterm, and be born to mothers who are non-White, have < a high school education (p<0.0001, respectively), and who drove further to surgical facility (p<0.005). Cox regression showed no racial/ethnic mortality disparities in the PM cohort. In the SM cohort, infant mortality was 2.2x higher among Black patients (HR 2.21, CI 1.41-3.45). In the NM cohort, Hispanic patients had nearly 2x (HR 1.93, CI 1.12-3.33) and Black patients had 3x (HR 3.01, CI 1.42-5.96) higher risk of infant mortality than White CCHDs infants. Conclusions: We completed one of the largest analyses of Medicaid-insured CCHD S infants using a novel SVI stratification. Despite worse SVI, racial/ethnic disparities in CCHD S mortality were not observed among PM-insured patients. Future database work should include stratified Medicaid populations to accurately interpret outcomes in patients with CCHD.