Effectiveness of Virtual Care Team Guided Management of Hospitalized Patients with HFrEF by Ethnicity

Background: IMPLEMENT-HF demonstrated a virtual team-based care strategy was safe and improved prescription of guideline directed medical therapy (GDMT) in hospitalized patients with heart failure and reduced ejection fraction (HFrEF). We evaluated differences in efficacy and safety outcomes by ethnicity in IMPLEMENT-HF.

Methods: IMPLEMENT-HF evaluated a provider-facing virtual team-based care strategy vs. usual care in hospitalized patients with HFrEF from October 2021 to June 2022. The primary outcome was change in a guideline directed medical therapy (GDMT) optimization score from hospital admission to discharge, with positive changes reflecting net optimization. In this post-hoc analysis, we assessed heterogeneity in treatment effects by ethnicity (Hispanic vs. non-Hispanic). Outcomes included prespecified primary and secondary effectiveness outcomes and adjudicated safety events.

Results: Of 808 screened patient admissions, 252 (31%) from 198 unique patients met inclusion criteria. Hispanic patients (n=43) were more likely to have diabetes and end-stage kidney disease than non-Hispanics; 70% spoke Spanish as a primary language. GDMT optimization score was lower among Hispanic vs. non-Hispanic patients (-0.44, 95% CI: -1.88 to 0.99 vs. +1.62, 95% CI: +1.02 to +2.21; p-value of interaction by ethnicity= 0.002). Allocation to the virtual care team intervention vs. usual care increased the proportion of patients experiencing >1 new initiation or dose up-titration among non-Hispanic patients but did not among Hispanic patients (absolute difference non-Hispanic vs. Hispanic: +31% vs. -19%; p-value of interaction = 0.003). Similar trends were seen among individual HF therapy and for the proportion of patients with optimization score >0 (absolute difference non-Hispanic vs. Hispanic: +29% vs. -20%; p-value of interaction = 0.005). Safety outcomes were similar among Hispanic and non-Hispanic patients.

Conclusion: A provider-facing, virtual care team guided strategy for HFrEF GDMT optimization was less effective in Hispanic patients. Efforts to identify and reduce bias contributing to this observed difference and more diverse implementation studies are needed.