Abstract 14442: Longitudinal Cardio-Oncology Trends in Sex and Racial Disparities in PCI and Inpatient Mortality: Machine Learning Augmented Propensity Score Analysis of Over 101 Million Hospitalizations

Abstract

Introduction: The longitudinal trends of cardio-oncology healthcare disparities is unknown. Methods: We performed the first nationally representative longitudinal analysis using Machine Learning-augmented Propensity Score adjusted multivariable regression (ML-PSr) and the 2016-2018 National Inpatient Sample (NIS), the United States’ largest all-payer inpatient dataset. Results: Among 101,521,656 hospitalizations from 2016-2018, 3,233,249 (3.18%) were female with active cancer, 18,310 (0.57%) had STEMI, 4,670 (0.14%) received LHC, 3,360 (0.10%) received PCI, and 160 (0.005%) received percutaneous heart pump (Impella). The most common active primary malignancies significantly differed for females (lung [15.19%], breast [14.10%], uterus [10.60%], leukemia [8.88%], and non-Hodgkin lymphoma [NHL] [6.92%]) versus males (prostate [15.56%], lung [15.21%], leukemia [10.05%], NHL [8.50%], and colon [6.57%]) (p<0.001). Among those with active cancer and STEMI, white females versus white males had similar likelihood of PCI, but nonwhite females versus nonwhite males were significantly less likely to receive PCI in 2016 (43.86% versus 52.20%, p=0.009), 2017 (25.64% versus 30.52%, p=0.039) and 2018 (11.10% versus 14.73%, p=0.001). In ML-PS multivariable regression fully adjusted for age, sex, race, income, region, STEMI, metastasis, NIS-calculated mortality risk by DRG, and the likelihood of receiving PCI, in 2016 PCI had less mortality reduction for females (OR 0.68, 95%CI 0.55-0.85; p<0.001) than males (OR 0.58, 95%CI 0.49-0.68; p<0.001) with nearly equal in 2017 for females (OR 0.50, 95%CI 0.34-0.74; p<0.001) and males (OR 0.50, 95%CI 0.38-0.66; p<0.001), but in 2018 there was a comparable non-significant mortality reduction for both females or males. Conclusions: Our study provides novel evidence that female racial minorities receive post-STEMI PCI at lower rates than male racial minorities, white females, and white males, though mortality disparities appear to have a favorable trend over time.