Vulnerable Patient Intensified Protocol to Reduce Readmission Disparities After Coronary Artery Bypass Grafting: Design, Implementation, and Lessons Learned from a Quality Initiative.

Abstract

Goal: In 2024, the Centers for Medicare & Medicaid Services (CMS) introduced equity metrics for healthcare systems to document social determinants of health (SDOH). Payment determinations were also linked to readmission performance measures. Readmission prevention programs for vulnerable patients, defined by CMS as dually eligible (DE) for Medicare and Medicaid, racial/ethnic minorities, and those with disabling conditions, have the potential to reduce readmission disparities. Our goal was to develop a systematic and pragmatic approach to collect, analyze, and utilize SDOH and insurance status to assign patients to an intensified protocol for reducing readmission disparities after coronary artery bypass grafting (CABG).

Methods: Patients admitted to a major urban medical center for isolated CABG from October 2023 to October 2024 (N = 286) completed a standardized clinician-assisted SDOH questionnaire. SDOH risk was adapted from ICD-10 Z codes that targeted social risk factors within the scope of healthcare providers' practices (i.e., language barriers, health literacy, access to a heart-healthy diet, lack of transportation to postoperative appointments, financial difficulties impeding access to prescription medications or medical care, and lack of a caregiver/social support). Project managers reviewed electronic health records and documented racial/ethnic categories using current CMS recommendations. SDOH+ (positive) and Medicaid or DE patients were assigned to a vulnerable patient intensified protocol (VPIP) readmission prevention program focused on education, deployment of hospital and community-based resources, transportation assistance, and increased frequency of postoperative follow-up. Non-VPIP patients continued their surgeons' usual care protocols.

Principal findings: Of the 286 isolated CABG patients, 55% were ≥65 years old, 80% were male, 47% were White, 24% had Medicaid, and 14% were DE. The most prevalent SDOH+ responses were the need for an interpreter (31%), low health literacy or less than a high-school education (23%), and transportation issues (17%). White patients were significantly (p < .05) less likely to qualify for VPIP than non-White patients, as were patients with Medicare compared to those with self-pay, commercial, or military insurance. Overall, 27% of patients had ≥2 SDOH risk factors. The need for an interpreter was 6.6 times more likely to be associated with having Medicaid or being DE than not. Low health literacy or less than a high school education, transportation issues, and the lack of access to a heart-healthy diet were all significantly associated with Medicaid or DE patients. White patients, compared to non-White patients or unknown or declined responses, were significantly less likely to be SDOH+ (13% vs. 39%, p < .0001). Asian patients were at a higher risk for SDOH+ compared to White patients (49% vs.13%, p < .0001), and among Asian patients, 29% had neither Medicaid nor DE status. The strongest predictor of SDOH+ status in logistic regression models was Medicaid or DE status compared to all other insurance types (OR = 5.4, confidence interval [3.0-9.9]) when adjusted for age, race/ethnicity, and gender.

Practical applications: Our findings demonstrate that a social-risk-informed care model is feasible in a hospital-based CABG readmission prevention program. Vulnerable patients can be identified through the standardized collection of SDOH, insurance status, and race/ethnicity data. Insurance status may be an excellent proxy to ascertain social risk and is readily accessible. Other healthcare organizations should consider regional demographics for possible SDOH risk. Our findings may support other initiatives to improve the collection of SDOH and demographic information. The VPIP CABG readmission prevention program could be adapted for other conditions and settings to achieve equitable care.