Telehealth‐aided outpatient management of acute heart failure in a specialist virtual ward compared with standard care

Abstract

AbstractAimsThe aim of this propensity score matched cohort study was to assess the outcomes of telehealth‐guided outpatient management of acute heart failure (HF) in our virtual ward (HFVW) compared with hospitalized acute HF patients.Methods and resultsThis cohort study (May 2022–October 2023) assessed outcomes of telehealth‐guided outpatient acute HF management using bolus intravenous furosemide in a HF‐specialist VW. Propensity score matching (PSM) was performed using logistic regression to adjust for potential differences in baseline patient characteristics between HFVW and standard care [Get With The Guidelines‐HF score, clinical frailty score (CFS), Charlson co‐morbidity index (CCI), NT‐proBNP, and ejection fraction]. Clinical outcomes (re‐hospitalizations and mortality) were compared at 1, 3, 6, and 12 months versus standard care‐SC (acute HF patients managed without telehealth in 2021). Five hundred fifty‐four HFVW ADHF patients (age 73.1 ± 10.9 years; 46% female) were compared with 404 ADHF patients (74.2 ± 11.8; P = 0.15 and 49% female) in the standard care‐SC cohort. After propensity score matching for baseline patient characteristics, re‐hospitalizations were significantly lower in the HFVW compared with SC (1 month‐HFVW 8.6% vs. SC‐21.5%, P < 0.001; 3 months‐21% vs. 30%, P = 0.003; 6 months‐28% vs 41%, P < 0.001 and 12 months‐47% vs. 57%, P = 0.005) and mortality was also lower at 1 month (5% vs. 13.7%; P < 0.001), 3 months (9.5% vs. 15%; P = 0.001), 6 months (15% vs. 21%; P = 0.03), and 12 months (20% vs. 26%; P = 0.04). Multivariate logistic regression analysis showed that compared with standard care, HFVW management was associated with lower odds of readmission (1‐month odds ratio (OR) = 0.3 [95% Confidence Interval CI 0.2–0.5], P < 0.0001; 3 month OR = 0.15 [0.1–0.3], P < 0.0001; 6‐month OR = 0.35 [0.2–0.6], P = 0.0002; 12‐month OR = 0.25 [0.15–0.4], P ≤ 0.001 and mortality (1‐month OR = 0.26 [0.14–0.48], P < 0.0001; 3‐month OR = 0.11 [0.04–0.27], P < 0.0001; 6‐month OR = 0.35, [0.2; 0.61], P = 0.0002; 12‐month OR = 0.6 [0.48; 0.73], P = 0.03. Higher GWTG‐HF score independently predicted increased odds of re‐hospitalization (1‐month OR = 1.2 [1.1–1.3], P < 0.001; 3‐month OR = 1.5 [1.37; 1.64], P < 0.0001; 6‐month OR = 1.3 [1.2–1.4], P < 0.0001; 12‐month OR = 1.1 [1.05–1.2], P = 0.03) as well as mortality (1‐month OR = 1.21 [1.1–1.3], P < 0.0001; 3‐month OR = 1.3 [1.2–1.4], P < 0.0001; 6‐month OR = 1.2 [1.1–1.3], P < 0.0001; 12‐month OR = 1.3 [1.1–1.7], P = 0.02). Similarly higher CFS also independently predicted increased odds of re‐hospitalizations (1‐month OR = 1.9 [1.5–2.4], P < 0.0001; 3‐month OR = 1.8 [1.3–2.4], P = 0.0003; 6‐month OR = 1.4 [1.1–1.8], P = 0.015; 12‐month OR 1.9 [1.2–3], P = 0.01]) and mortality (1‐month OR = 2.1 [1.6–2.8], P < 0.0001; 3‐month OR = 1.8 [1.2–2.6], P = 0.006; 6‐month OR = 2.34 [1.51–5.6], P = 0.0001; 12‐month OR = 2.6 [1.6–7], P = 0.02). Increased daily step count while on HFVW independently predicted reduced odds of re‐hospitalizations (1‐month OR = 0.85[0.7–0.9], P = 0.005), 3‐month OR = 0.95 [0.93–0.98], P = 0.003 and 1‐month mortality (OR = 0.85 [0.7–0.95], P = 0.01), whereas CCI predicted adverse 12‐month outcomes (OR = 1.2 [1.1–1.4], P = 0.03).ConclusionsTelehealth‐guided specialist HFVW management for ADHF may offer a safe and efficacious alternative to hospitalization in suitable patients. Daily step count in HFVW can help predict risk of short‐term adverse clinical outcomes.