Background:
Despite guideline recommendations, initiation of guideline‑directed medical therapy (GDMT) for heart failure with reduced ejection fraction (HFrEF) remains suboptimal, and contemporary routine care analysis of the ambulatory and hospital setting, continue to show gaps in comprehensive use across classes. The ESC heart failure guidelines emphasize the hospital-based discharge management in optimizing pharmacological care. Therefore, the Helios hospitals initiated a GDMT program with structured physician training.
Purpose:
The primary aim of PHRASE‑HF (NCT06675552) is to evaluate the use of GDMT at hospital discharge and the in‑hospital initiation of the four drug classes (renin–angiotensin–aldosterone system inhibitor or angiotensin receptor–neprilysin inhibitor, RAASi/ARNi; beta‑blocker, BB; sodium–glucose cotransporter‑2 inhibitor, SGLT2i; mineralocorticoid receptor antagonist, MRA) as indicated, with 12‑month follow‑up for clinical and patient‑reported outcomes. Patients were assessed for phenotype-specific pharmacotherapy, accounting for substance-specific contraindications. Here we present the study design and results of the first interim analysis.
Methods: PHRASE‑HF is a prospective, multicentre, non‑interventional study (planned N=438) enrolling adults hospitalised with HFrEF (LVEF ≤ 40%) suspected of being pharmacologically undertreated, defined by receipt of ≤2 GDMT classes at admission. Univariable and multivariable logistic regression (covariates selected for univariable p<0.20) assessed factors associated with GDMT concordance (GDMTc) at discharge, defined as receipt of all four classes when eligible or 3 classes when one class is contraindicated. Follow‑up is ongoing at 3, 6, and 12 months capturing GDMT, clinical, and patient‑reported outcomes. GDMTc will be assessed for temporal trends, including routine care data of the Helios Heart Failure Registry before starting the GDMT program.
Results:
Of 200 patients enrolled at six Helios hospitals, 189 constituted the analysis set meeting all inclusion/exclusion criteria. GDMTc rate was 63.5% (120/189). Compared with discordance, diabetes (DM) was less frequent in GDMTc patients (23.3% vs. 40.6%; p=0.020), and systolic blood pressure (SBP) tended to be higher (138.7 ± 25.4 vs. 133.4 ± 24.1 mmHg; p=0.162), while age, sex, BMI, LVEF, NYHA class, atrial fibrillation, ischaemic heart disease, renal impairment, alcohol intake and smoking were similar. Length of stay tended to be shorter (median 6.5 vs. 8.0 days; p=0.056). In the multivariable model, each 10 mmHg higher SBP was independently associated with GDMTc (OR 1.15, 95% CI 1.00–1.32; p=0.049), absence of DM (OR 1.849, 95% CI 0.902–3.789; p=0.093) showed a non‑significant trend and other characteristics were not associated. GDMT eligibility for all four classes was similar in both groups (89.2% vs. 91.3%; p=0.803). The most frequent gaps (n=69) were MRA (absent in 75.4%), followed by SGLT2i (33.3%), BB (26.1%), and RAASi/ARNi (20.3%).
Conclusions:
At hospital discharge in this interim analysis, GDMTc was observed in most patients previously receiving suboptimal GDMT and was associated with increased SBP, despite tendencies for comorbidities and length of stay. Discordance predominantly reflected missed opportunities to prescribe MRA —and, to a lesser extent, other classes—rather than ineligibility. Follow-up will assess how rapid in-hospital GDMT initiation translates to outpatient care and outcomes.
