Background: Heart failure affects 7 million United States adults. It remains a major contributor to morbidity, mortality, and hospital readmission. Optimizing a guideline-directed medical therapy (GDMT) regimen—particularly initiation and up-titration during the index hospitalization—has been shown to reduce rehospitalization rates, improve survival, and lower healthcare costs. There is also emerging evidence that the type of hospital community vs. academic (teaching) can influence prescribing practices and readmission outcomes. Our health system sought to improve heart failure outcomes through a combination of education, incentives, and systems interventions. Our study aims to investigate the inter-hospital variation in GDMT prescription at discharge across these intervention types.
Methods: Patients with primary heart failure discharge diagnosis across 10 enterprise hospitals were studied using a unified EHR (Epic) from July 2023 to June 2025. Hospitals were classified as academic (included 3 hospitals) or community (included 7 hospitals). The study period was divided into two seasonally-matched pre-intervention (Pre 1: July–October 2023; Pre 2: December 2023–June 2024) and post-intervention phases (Post 1: July–October 2024 with financial incentives and systemwide HF workgroups; Post 2: December 2024–June 2025 with added clinical-decision support systems [CDSS] tools including HF order sets and documentation prompts). The primary outcome was the mean number of GDMT agents prescribed at discharge for patients with HFrEF/HFimpEF in the community versus academic setting. Secondary outcomes included the proportion of eligible patients receiving all four GDMT classes and SGLT2 inhibitors in academic versus community hospitals.
Results: Between 7/01/2023-6/30/2025, 3,535 patients were included. The average number of GDMT medications per HFrEF/HFimpEF patients increased significantly in community hospitals in the Pre 1-> Post 1 analysis (2.19 v 2.44, p= 0.0028; 11.8% increase) vs AMC (2.32 v 2.38, p= 0.46; 2.7% increase); in Pre 2-> Post 2 analysis community hospitals increased from 2.35 to 2.47 (p= 0.053; 5.1% increase) compared to AMC increasing from 2.27 to 2.48 (p=0.0009; 9.1% increase) as seen in Figure 1. The proportion of HFpEF/HFmrEF patients on SGLT2is showed relative increase Pre 1-> Post 1 analysis 52.8% (community) versus 61.4% (AMC) and Pre 2-> Post 2 63.8% (community) versus 84.6% (AMC). The proportion of HFrEF/HFimpEF patients on all four-pillars of GDMT showed relative increase Pre 1-> Post 1 58.7% (community) versus 47.4% (AMC) and Pre 2-> Post 2 28.4% (community) versus 33.9% (AMC) as seen in Figure 2.
Conclusions: This data suggests community hospitals have a more pronounced response to education and incentives, whereas academic hospitals respond more to EHR interventions. This data underscores how institutional context (academic vs. community) may influence execution of guideline-based care. Quality improvement interventions for large systems are scalable to both community and academic hospitals. Thus, future investigations should evaluate why education-focused interventions, targeted incentive structures, and EHR-based CDSS might affect enhanced GDMT prescribing differently so iterative adjustments can be implemented to universally improve heart failure care.
.png)
.png)