Abstract
Background
Heart failure with preserved ejection fraction (HFpEF) remains challenging to diagnose, and several scoring systems have been developed to aid evaluation. Emerging evidence suggests that these scores provide prognostic information. This study assessed the prognostic utility of HFpEF scoring systems and compared their performance.
Methods
We analyzed data of 448 consecutive patients with HFpEF from a nationwide, multicenter, prospective, observational heart failure registry (TSOC [Taiwan Society of Cardiology] Heart Failure Registry 2020). We proposed the TSOC‐HFpEF score, which incorporated left ventricular mass index, left atrial volume index, E/e’ ratio, tricuspid regurgitation velocity, and N‐terminal pro‐brain natriuretic peptide level. Its prognostic performance was compared with that of H2FPEF and Heart Failure Association Pre‐test Assessment, Echocardiography and Natriuretic Peptide, Functional Testing, Final Etiology scores.
Results
The mean age of the cohort was 77.0±10.6 years, and 60.7% were women. Mortality rates increased across TSOC‐HFpEF scores, ranging from 0.0 to 23.9 per 100 person‐years. High‐risk categories defined by the Heart Failure Association Pre‐test Assessment, Echocardiography and Natriuretic Peptide, Functional Testing, Final Etiology and TSOC‐HFpEF scores were associated with higher all‐cause mortality than their respective non‐high‐risk categories (hazard ratio [HR], 2.18 [95% CI, 1.17–4.07] and HR, 3.35 [95% CI, 1.97–5.70], respectively), whereas the association for the high‐risk H2FPEF category was of borderline statistical significance (HR, 1.55 [95% CI, 0.99–2.42]; P=0.053). The TSOC‐HFpEF score had the highest discrimination for mortality (C‐index, 0.65) compared with the H2FPEF (C‐index, 0.53) and Heart Failure Association Pre‐test Assessment, Echocardiography and Natriuretic Peptide, Functional Testing, Final Etiology (C‐index, 0.57) scores (P=0.001 and P=0.018, respectively, by DeLong test). Renin‐angiotensin system inhibitor use was associated with lower mortality during a median follow‐up of 2.0 years (interquartile range, 1.5–2.0 years).
Conclusions
Diagnostic scores for HFpEF can provide prognostic value. The proposed simple echocardiography and natriuretic peptide‐based scoring systems offered superior risk discrimination.
Keywords: heart failure with preserved ejection fraction, mortality, N‐terminal pro‐brain natriuretic peptide, renin–angiotensin system inhibitors, Taiwan society of cardiology
Subject Categories: Echocardiography, Heart Failure, Biomarkers
Nonstandard Abbreviations and Acronyms
- HFA‐PEFF
Heart Failure Association Pre‐test Assessment, Echocardiography and Natriuretic Peptide, Functional Testing, Final Etiology
- HFpEF
heart failure with preserved ejection fraction
- LAVi
left atrial volume index
- LVMi
left ventricular mass index
- TSOC
Taiwan Society of Cardiology
Research Perspective.
What Is New?
This study introduces the Taiwan Society of Cardiology‐heart failure with preserved ejection fraction (HFpEF) score, a novel, simplified prognostic tool specifically validated in an ethnic Asian cohort with a high comorbidity burden.
The Taiwan Society of Cardiology‐heart failure with preserved ejection fraction score, which integrates standard echocardiographic parameters (left ventricular mass, left atrial volume, E/e’, and tricuspid regurgitation velocity) with NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) levels, demonstrated superior discriminative ability for 2‐year mortality (C‐index: 0.65) compared with the established H2FPEF and Heart Failure Association Pre‐test Assessment, Echocardiography and Natriuretic Peptide, Functional Testing, Final Etiology scores.
What Question Should Be Addressed Next?
Future research should evaluate the performance of Taiwan Society of Cardiology‐HFpEF score in younger or lower‐risk HFpEF populations outside of the advanced, high‐comorbidity Asian registry setting; the prospective implementation of this scoring system in clinical workflows to improve physician adherence to foundational therapies in patients with HFpEF and subsequent outcomes.
Despite advances in pharmacological therapies, heart failure (HF) remains a leading cause of hospitalization and mortality worldwide, substantially contributing to the global economic burden. 1 , 2 , 3 HF with preserved ejection fraction (HFpEF), defined by a normal or near‐normal left ventricular ejection fraction (LVEF ≥50%), has emerged as a common HF phenotype worldwide. Its rising prevalence is largely driven by population aging and the increasing burden of comorbidities. 4 , 5 , 6 Senescence and conditions such as hypertension, diabetes, chronic kidney disease (CKD), atrial fibrillation (AF), and microvascular dysfunction play key roles in HFpEF pathogenesis, progression, and outcomes.
Because HFpEF is a complex syndrome, accurate diagnosis in clinical practice remains challenging. Several risk‐scoring systems have been developed to support diagnostic workflows. Two widely used scoring systems are the H2FPEF and Heart Failure Association Pre‐test Assessment, Echocardiography and Natriuretic Peptide, Functional Testing, Final Etiology (HFA‐PEFF) scores, which integrate clinical demographics, echocardiographic parameters, and natriuretic peptide levels (Brain natriuretic peptide [BNP] or N‐terminal proBNP [N‐terminal pro‐B‐type natriuretic peptide]) to estimate the likelihood of HFpEF across risk categories. These scoring systems have also been shown to predict adverse outcomes, with the HFA‐PEFF score demonstrating a stronger prognostic value than the H2FPEF score. In recently published local guideline, 7 we proposed diagnostic criteria based on clinical presentation, abnormal natriuretic peptide levels, and echocardiographic measures of cardiac structure and function, adapted from The American Society of Echocardiography diastolic dysfunction criteria. However, the prognostic implications of these criteria remain uncertain.
Accordingly, this study aimed to evaluate the feasibility of applying HFpEF diagnostic scores for prognostic purposes and compare the prognostic performances of different scoring systems using data from a nationwide, prospectively enrolled, multicenter cohort.
METHODS
Study Population
Data were derived from the TSOC (Taiwan Society of Cardiology) Heart Failure Registry 2020, a nationwide, multicenter, prospective, observational registry. The registry was approved by the Institutional Review Board or ethics committee at each participating hospital and was conducted in accordance with the Declaration of Helsinki (Data S1). All participants provided written informed consent to this study. The data that support the findings of this study are available from the corresponding author upon reasonable request. The median follow‐up was 2.0 years (interquartile range: 1.5–2.0 years). 8 Eligible patients were adults (age >20 years) with a clinical diagnosis of HF, either hospitalized or discharged within 6 months for acute decompensated HF, and meeting the Framingham criteria. Exclusion criteria included unwillingness to participate, severe valvular disease, recent acute myocardial infarction (≤2 weeks), acute severe arrhythmia, CKD stage 5 or regular dialysis, constrictive pericarditis, congenital heart disease, confirmed hypertrophic cardiomyopathy or amyloidosis, Takotsubo cardiomyopathy, thyroid storm, isolated right ventricular HF, severe liver cirrhosis, and life‐threatening conditions with a life expectancy <1 year.
According to the pre‐specified enrollment criteria, the registry aimed to enroll 59%, 12%, and 29% of patients with HF with reduced ejection fraction (HFrEF), HF with mildly reduced ejection fraction (HFmrEF), and HFpEF, respectively. However, the final enrollment proportions were 69%, 15%, and 16% of patients with HFrEF, HFmrEF, and HFpEF, respectively. Our analyses focused on the HFpEF cohort. LVEF was assessed by echocardiography within 3 months of enrollment. Inclusion in the HFpEF cohort required the following: LVEF ≥50%; modified H2FPEF score ≥5 (Tables S1–S5), considering the general tendency toward a lean HFpEF phenotype with lower body mass index (BMI) in Asian populations 9 , 10 ; and diagnostic NT‐proBNP >125 pg/mL in sinus rhythm, NT‐proBNP >250 pg/mL in AF, or BNP >35 pg/mL regardless of rhythm. Overall, the entire cohort enrolled 45.3% of inpatients in the HFpEF cohort. Furthermore, among those enrolled in outpatient settings, the median discharge‐to‐enrollment interval was only 8 days.
Clinical Data Collection and Definition
Baseline demographics including age, sex, and BMI; comorbidities; and New York Heart Association functional class at baseline were extracted from electronic medical records. Medical histories and comorbidities included hypertension, type 2 diabetes, hypercholesterolemia, hypertriglyceridemia, peripheral artery disease, cerebral vascular accident, AF, coronary artery disease, prior percutaneous coronary intervention or coronary artery graft bypass surgery, CKD, chronic obstructive pulmonary disease, sleep apnea, and malignancy.
Echocardiographic Measures
LVEF was measured using the Simpson method. Other parameters including geometry (left ventricular hypertrophy, chamber diameter, and relative wall thickness), cardiac chamber size (left atrial [LA] and left ventricular [LV]), conventional hemodynamic Doppler information, myocardial tissue Doppler data, and LV filling pressures (E/e’) were obtained through a pre‐specified structural questionnaire according to standardized international recommendation guidelines. 11 , 12
HFpEF Scoring Systems
The H2FPEF score incorporates clinical and echocardiographic parameters, including age >60 years (1 point), BMI >30 kg/m2 (2 points), hypertension treated with ≥2 antihypertensive medications (1 point), history of paroxysmal or persistent AF (3 points), systolic pulmonary artery pressure (sPAP) >35 mm Hg assessed on Doppler echocardiography (1 point), and E/e’ ratio >9 (1 point). 13 The HFA‐PEFF score integrates morphological, functional, and biomarker domains. Major morphological parameters include left atrial volume index (LAVi) >34 mL/m2, relative wall thickness >0.42, or left ventricular mass index (LVMi) ≥149 g/m2 in men or ≥122 g/m2 in women. Minor morphological parameters include LAVi 29–34 mL/m2, relative wall thickness >0.42, LVMi 116–149 g/m2 in men or 96–122 g/m2 in women, or LV wall thickness ≥12 mm. Functional criteria for patients aged <75 years include septal e’ <7 cm/s or lateral e’ <10 cm/s, with average E/e’ ≥15, tricuspid regurgitation velocity >2.8 m/s, or sPAP >35 mm Hg classified as major, and average E/e’ 9–14 classified as minor. For biomarkers, NT‐proBNP, >220 pg/mL is considered major, while 125–220 pg/mL is minor. 14 Patients were stratified into high‐, intermediate‐, or low‐risk categories according to the recommendations and classified into high‐risk and non‐high‐risk (intermediate to low risk) categories in the present study.
Our society also proposed a local HFpEF diagnostic workflow adapted from ASE diastolic dysfunction criteria. 7 , 11 This workflow includes 4 components: left ventricular hypertrophy defined by sex‐specific LVMi cutoffs (≥95 gm/m2 in women and ≥115 gm/m2 in men), LA enlargement (LAVi >29 mL/m2 in sinus rhythm or LAVi >40 mL/m2 in AF), elevated LV filling pressures (average E/e’ >14), and elevated pulmonary arterial pressure (tricuspid regurgitation velocity >2.8 m/s) (Tables S1–S5). Given the high NT‐proBNP values in our cohort (median: 2938 pg/mL, interquartile range: 1340–5634 pg/mL), we set the prognostic NT‐proBNP cut‐off as >3000 pg/mL. Based on these parameters, we developed the TSOC‐HFpEF score (range 0–5), assigning 1 point for each abnormal echocardiographic criterion and 1 point for elevated NT‐proBNP levels.
Medication Usage
Information on the adoption of foundational HF medications, classes, and uptake doses at discharge was obtained. Discharge prescriptions for renin–angiotensin system inhibitors (RASis), beta‐blockers, mineralocorticoid receptor antagonists (MRAs), and sodium‐glucose cotransporter‐2 inhibitors (SGLT2i)—and their dosages was obtained from the predefined structured case report form. 15 , 16 , 17
Clinical Outcome Measures
The primary outcome all‐cause mortality at 2 years, analyzed by Kaplan–Meier survival. Secondary outcomes included cardiovascular and non‐cardiovascular mortality and HF hospitalizations for 2 years.
Statistical Analysis
Continuous variables are presented as mean values±SDs; categorical variables are presented as percentages. Between‐group differences were tested using Chi‐square test for categorical variables and independent Student' t‐test for continuous variables. Prognostic performance for all‐cause mortality was assessed by area under the receiver operating characteristic curve. The optimal TSOC HFpEF cutoff point was determined using Youden index from the maximum area under the receiver operating characteristic curve (cutoff point=2) and further categorized as high‐risk (>2) and non‐high‐risk (≤2) categories. Missing data were handled using multiple imputations via fully conditional specifications in SPSS (version 26). Each variable with missing data was imputed using regression models appropriate for the measurement level (eg, linear regression for continuous variables and logistic regression for categorical variables). Five imputed data sets were generated and subsequently pooled for statistical analysis, in accordance with Rubin's rules. We pre‐specified the clinical endpoint as all‐cause mortality, with secondary endpoints as cardiovascular death, non‐cardiovascular death, and heart failure hospitalization. Survival analysis was performed using uni‐ and multivariate Cox proportional hazards regression models to evaluate the associations between various risk scores and pre‐specified clinical endpoints. Time‐dependent Harrell C statistics (C‐index) was used to evaluate the predictive performance of various scores used in current study. DeLong's test was used to compare the C‐index of each score against the TSOC‐HFpEF score. The proportional hazards assumption was assessed using Schoenfeld residual test, which showed no non‐proportionality. Kaplan–Meier curves were compared using log‐rank tests. Finally, to internally validate the discrimination performance of the TSOC HFpEF score and to compare it with the H2FPEF and HFA‐PEFF scores, we performed bootstrap resampling with methods further detailed elsewhere (Data S2). For heart failure hospitalization, recurrent event counts were analyzed using Poisson regression, with the number of heart failure hospitalizations as the dependent variable and the logarithm of follow‐up time included as an offset. Results are presented as incidence rate ratios with 95% CIs. Harrell C‐index was not applicable to this analysis. Analyses were performed with Stata version 14 (StataCorp, College Station, TX, USA). The number of decimal places was selected according to the scale and precision of each measure and the information supported by the available sample and event counts. Counts were reported as integers; percentages to 1 decimal place; continuous variables generally to 1 decimal place, unless greater precision was inherent to the measurement scale; incidence rates per 100 person‐years to 1 decimal place; effect estimates and 95% CIs to 2 decimal places; C‐indices and their 95% CIs to 2 decimal places; and exact P values to 3 decimal places, with values <0.001 reported as P<0.001. A 2‐tailed P<0.05 was considered statistically significant.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
RESULTS
Study Participants
The mean age of the HFpEF cohort was 77.0±10.6 years; 60.7% were women. The mean BMI was 27.0±5.7 kg/m2, and the median NT‐proBNP level was 2938 pg/mL (interquartile range: 1340–5634). Overall, 34.8% of our HFpEF cohort presented with New York Heart Association functional class III or IV. Baseline characteristics by scoring system are shown in Table 1. The cohort had high rates of AF (44.9%), prior HF hospitalization (68.3%), hypertension (80.4%), diabetes (49.1%), and CKD (63.4%); 25.0% had ischemic heart disease. Patients in the high‐risk TSOC‐HFpEF score category (>2) were older, more often women, had lower heart rates, and higher coronary artery disease prevalence compared with those in the non‐high‐risk category (Table 1). Based on H2FPEF, HFA‐PEFF, and TSOC‐HFpEF scores, 69.6%, 82.1%, and 66.5% of patients, respectively, were classified as high‐risk category.
Table 1.
Baseline Characteristics of Patients With HFpEF Stratified by H2FPEF, HFA‐PEFF, and TSOC‐HFpEF Scores
| Characteristic | High‐risk H2FPEF (n=312) | Non‐high‐risk H2FPEF (n=136) | High‐risk HFA‐PEFF (n=368) | Non‐high‐risk HFA‐PEFF (n=80) | High‐risk TSOC‐HFpEF (n=298) | Non‐high‐risk TSOC‐HFpEF (n=150) |
|---|---|---|---|---|---|---|
| Age, y | 78.2±10.2 | 74.3±11.2 | 77.7±10.1 | 74.0±12.3 | 78.3±9.9 | 74.5±11.7 |
| Female sex | 194 (62.2%) | 78 (57.4%) | 232 (63.0%) | 40 (50.0%) | 201 (67.5%) | 71 (47.3%) |
| High HFA‐PEFF score (5 or 6) | 261 (83.7%) | 107 (78.7%) | … | … | 270 (90.6%) | 98 (65.3%) |
| High H2FPEF score (6–9) | … | … | 261 (70.9%) | 51 (63.7%) | 220 (73.8%) | 92 (61.3%) |
| BMI, kg/m2 | 27.4±5.9 | 26.1±5.2 | 26.9±5.8 | 27.3±5.7 | 26.6±5.5 | 27.8±6.1 |
| Heart rate, beats/min | 78.7±17.2 | 77.7±17.5 | 77.9±16.9 | 80.7±19.0 | 76.7±16.5 | 81.7±18.3 |
| Systolic blood pressure, mm Hg | 127.7±20.2 | 128.1±20.0 | 128.4±20.5 | 125.4±18.6 | 128.1±20.6 | 127.4±19.3 |
| NYHA functional class I/II | 197 (63.1%) | 95 (69.9%) | 244 (66.3%) | 48 (60.0%) | 186 (62.6%) | 106 (70.7%) |
| NYHA functional class III | 96 (30.8%) | 29 (21.3%) | 99 (27.1%) | 26 (31.3%) | 87 (29.3%) | 38 (25.3%) |
| NYHA functional class IV | 19 (6.1%) | 12 (8.8%) | 24 (6.6%) | 7 (8.4%) | 24 (8.1%) | 6 (4.0%) |
| Prior hospitalization for HF | 222 (71.2%) | 84 (61.8%) | 247 (67.1%) | 59 (73.8%) | 207 (69.5%) | 99 (66.0%) |
| History of CAD | 79 (25.3%) | 30 (22.1%) | 96 (26.1%) | 13 (16.3%) | 85 (28.5%) | 24 (16.0%) |
| History of PCI | 57 (18.3%) | 22 (16.2%) | 68 (18.5%) | 11 (13.8%) | 60 (20.1%) | 19 (12.7%) |
| History of MI | 26 (8.3%) | 12 (8.8%) | 31 (8.4%) | 7 (8.8%) | 25 (8.4%) | 13 (8.7%) |
| History of PAD | 9 (2.9%) | 5 (3.7%) | 13 (3.5%) | 1 (1.3%) | 12 (4.0%) | 2 (1.3%) |
| History of VTE/PE | 5 (1.6%) | 4 (2.9%) | 8 (2.2%) | 1 (1.3%) | 4 (1.3%) | 5 (3.3%) |
| History of stroke/TIA | 41 (13.1%) | 16 (11.8%) | 52 (14.1%) | 5 (6.3%) | 37 (12.4%) | 20 (13.3%) |
| History of hypertension | 274 (87.8%) | 86 (63.2%) | 301 (81.9%) | 59 (73.8%) | 246 (82.6%) | 114 (76.0%) |
| History of AF | 286 (91.7%) | 34 (25.0%) | 261 (70.9%) | 59 (73.8%) | 210 (70.5%) | 110 (73.3%) |
| History of CIED | 21 (6.7%) | 10 (7.4%) | 27 (7.3%) | 4 (5.0%) | 22 (7.4%) | 9 (6.0%) |
| History of dyslipidemia | 87 (27.9%) | 42 (30.9%) | 108 (29.3%) | 21 (26.3%) | 85 (28.5%) | 44 (29.3%) |
| History of diabetes | 159 (51.0%) | 61 (44.9%) | 180 (48.9%) | 40 (50.0%) | 154 (51.7%) | 66 (44.0%) |
| History of COPD | 43 (13.8%) | 8 (5.9%) | 45 (12.2%) | 6 (7.5%) | 38 (12.8%) | 13 (8.7%) |
| History of cancer | 37 (11.9%) | 20 (14.7%) | 46 (12.5%) | 11 (13.8%) | 34 (11.4%) | 23 (15.3%) |
| History of depression | 5 (1.6%) | 3 (2.2%) | 6 (1.6%) | 2 (2.5%) | 5 (1.7%) | 3 (2.0%) |
| History of sleep apnea | 5 (1.6%) | 2 (1.5%) | 7 (1.9%) | 0 (0.0%) | 2 (0.7%) | 5 (3.3%) |
Values are mean±SD or N (%), with percentages shown in parentheses. High H2FPEF: (6–9) vs Intermediate H2FPEF (2–5); High HFA‐PEFF (5 or 6) vs Intermediate HFA‐PEFF (2–4); High TSOC (3–6) vs Low TSOC (0–2). AF indicates atrial fibrillation; CAD, coronary artery disease; CIED, cardiac implantable electronic device; COPD, chronic obstructive pulmonary disease; MI, myocardial infarction; NYHA Fc, New York Heart Association functional class; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; PE, pulmonary embolism; TIA, transient ischemic attack; and VTE, venous thromboembolism.
Cardiac Structure and Functions
In the study cohort, the mean LVEF, LVMi, average E/e’, LAVi, and TVR velocity were 61.4±8.8%, 126.8±50.8 g/m2, 16.2±9.0, 54.4±33.1 mL/m2, and 2.9±0.6 m/s, respectively. LV hypertrophy was present in 69.0% (n=309) and LA enlargement in 71.9% (n=322). Patients in high‐risk categories generally showed more adverse remodeling (Table 2). Higher LVMi, average E/e’, and TVR velocity—but not LAVi—were statistically significant with modest correlations with higher NT‐proBNP levels (r=0.12, 0.16, and 0.12; P=0.009, 0.009, and 0.014, respectively). Patients classified as high‐risk category using HFA‐PEFF and TSOC‐HFpEF scores, but not H2FPEF, had more unfavorable ventricular remodeling patterns (Figures S1–S5). Distribution of major TSOC‐HFpEF score components is shown in Figure 1.
Table 2.
Baseline Echocardiographic Measurements of Patients With HFpEF Stratified by H2FPEF, HFA‐PEFF, and TSOC‐HFpEF Scores
| Echocardiographic variable | High‐risk H2FPEF (n=312) | Non‐high‐risk H2FPEF (n=136) | P value | High‐risk HFA‐PEFF (n=368) | Non‐high‐risk HFA‐PEFF (n=80) | P value | High‐risk TSOC‐HFpEF (n=298) | Non‐high‐risk TSOC‐HFpEF (n=150) | P value |
|---|---|---|---|---|---|---|---|---|---|
| LAVi, mL/m2 | 56.3±35.9 | 50.0±31.4 | 0.098 | 57.3±34.7 | 38.1±30.1 | <0.001 | 60.9±35.2 | 41.7±24.0 | <0.001 |
| IVS, mm | 12.0±2.7 | 11.9±2.7 | 0.836 | 12.1±2.7 | 11.1±2.4 | 0.014 | 12.1±2.7 | 11.6±2.8 | 0.086 |
| LVPW, mm | 11.2±2.4 | 11.2±2.5 | 0.853 | 11.4±2.5 | 10.0±1.9 | <0.001 | 11.4±2.5 | 10.6±2.2 | <0.001 |
| LVEDD, mm | 48.2±9.2 | 47.8±7.3 | 0.723 | 47.8±7.5 | 49.4±12.8 | 0.138 | 48.7±9.2 | 46.9±7.4 | 0.042 |
| LVEDVi, mL/m2 | 52.4±24.7 | 57.4±22.7 | 0.051 | 54.0±24.5 | 54.3±22.5 | 0.933 | 56.7±25.4 | 49.1±20.0 | 0.002 |
| LVESVi, mL/m2 | 20.6±12.6 | 21.8±10.5 | 0.374 | 20.9±12.1 | 21.4±10.6 | 0.762 | 21.9±12.6 | 19.7±9.5 | 0.067 |
| TR velocity, m/s | 3.0±0.6 | 2.6±0.6 | <0.001 | 2.9±0.6 | 2.6±0.5 | <0.001 | 3.0±0.6 | 2.5±0.5 | <0.001 |
| Mitral E/A ratio | 1.8±1.2 | 1.3±0.9 | <0.001 | 1.6±1.1 | 1.6±1.2 | 0.784 | 2.9±19.2 | 1.3±0.8 | 0.322 |
| Mean TDI e’ velocity, cm/s | 7.5±2.9 | 7.1±3.0 | 0.188 | 7.2±2.8 | 9.3±3.4 | <0.001 | 7.2±2.9 | 7.8±2.6 | 0.029 |
| E/e’ | 16.5±8.1 | 16.1±11.8 | 0.653 | 17.0±9.6 | 11.6±5.3 | <0.001 | 18.4±10.0 | 11.7±3.7 | <0.001 |
| LVEF, % | 60.9±8.8 | 62.8±8.6 | 0.034 | 61.6±8.8 | 60.5±8.7 | 0.279 | 61.9±8.9 | 60.4±8.4 | 0.090 |
| LVMi, g/m2 | 127.0±54.3 | 126.4±42.1 | 0.901 | 129.7±44.6 | 113.8±72.0 | 0.043 | 135.9±55.0 | 108.7±35.0 | <0.001 |
| Relative wall thickness | 0.48±0.14 | 0.48±0.15 | 0.756 | 0.49±0.14 | 0.43±0.13 | <0.001 | 0.50±0.14 | 0.49±0.15 | 0.529 |
Values are mean±SD. EDV indicates end‐diastolic volume; EF, ejection fraction; IVS, inter‐ventricular septum; LAVi, left atrial volume index; LVEDD, left ventricular end‐diastolic diameter; LVEDV, left ventricular end‐diastolic volume; LVEF, left ventricular ejection fraction; LVESV, left ventricular end‐systolic volume; LVMi, left ventricular mass index; LVPW, left ventricular posterior wall; Mitral E/A, mitral inflow peak E and A velocities; TDI, Tissue Doppler imaging; and TR, tricuspid regurgitation.
Figure 1. Distribution of key TSOC‐HFpEF score components.
Distribution of key TSOC‐HFpEF score components (echocardiography structural or functional), including LV mass index, LA volume index, E/e’, TR velocity, and NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) level, stratified by those with (orange) or without (blue) clinical events of death defined at the 2‐year follow‐up period. The levels of all the components were significantly higher in patients who died. TSOC indicates Taiwan Society of Cardiology; and HFpEF, heart failure with preserved ejection fraction.
Prognostic Implications by HF2PEF/HFA‐PEFF and TSOC‐HFpEF Scores
During a median follow‐up of 2.0 years (interquartile range: 1.5–2.0 years), 109 patients died from any cause, and 92 experienced ≥1 HF hospitalization. The mean H2FPEF, HFA‐PEFF, and TSOC‐HFpEF scores in this study were 6.2±1.8, 5.2±0.9, and 3.0±1.3, respectively. Compared to their non‐high‐risk counterparts, those classified into high‐risk HFA‐PEFF category showed doubled mortality risk (HR: 2.18, 95% CI: 1.17–4.07, P=0.012), whereas those classified into high‐risk H2FPEF category showed borderline significance (HR: 1.55, 95% CI: 0.99–2.42, P=0.053). Those classified into high‐risk TSOC‐HFpEF scores nearly tripled mortality risk compared with non‐high‐risk category (HR: 3.35, 95% CI: 1.97–5.70, P<0.001) (Table 3 and Figure 2). Mortality risk increased progressively with higher TSOC‐HFpEF scores, from 0.0% at a score of 0 to 17.4% and 23.9% at scores of 4 and 6, respectively (p for trend, <0.001) (Figure 3, right; Table 3). The Harrell's C‐index for all‐cause mortality was 0.65 (95% CI: 0.59–0.70) for the TSOC‐HFpEF score, 0.57 (95% CI: 0.51–0.62) for the HFA‐PEFF score, and 0.53 (95% CI: 0.47–0.59) for the H2FPEF score. Using the TSOC‐HFpEF score as the reference, DeLong's test showed that its C‐index was significantly higher than that of the HFA‐PEFF score (0.65 vs 0.57, P=0.018) and the H2FPEF score (0.65 vs 0.53, P=0.001). Bootstrapping sampling (n=1000 replicates) confirmed the consistency of superior discrimination performance by TSOC HFpEF compared with the H2FPEF score and the HFA‐PEFF score for predicting 2‐year mortality (Data S2) (Figure S2).
Table 3.
Clinical Outcomes According to H2FPEF, HFA‐PEFF, and TSOC‐HFpEF Score Categories
| Outcome | Score group | Risk category | Incidence rate (per 100 person‐years) | Harrell C‐index | HR (95% CI) |
|---|---|---|---|---|---|
| All‐cause mortality | H2FPEF | Non‐high‐risk | 10.8 | 0.53 | Reference: 1.00 |
| High‐risk | 16.9 | 1.55 (0.99–2.42) | |||
| HFA‐PEFF | Non‐high‐risk | 7.5 | 0.57 | Reference: 1.00 | |
| High‐risk | 16.9 | 2.18 (1.17–4.07) | |||
| TSOC‐HFpEF | Non‐high‐risk | 5.8 | 0.65 | Reference: 1.00 | |
| High‐risk | 20.4 | 3.35 (1.97–5.70) | |||
| cardiovascular death | H2FPEF | Non‐high‐risk | 2.2 | 0.53 | Reference: 1.00 |
| High‐risk | 6.2 | 2.86 (1.11–7.36) | |||
| HFA‐PEFF | Non‐high‐risk | 2.0 | 0.56 | Reference: 1.00 | |
| High‐risk | 5.7 | 2.70 (0.83–8.81) | |||
| TSOC‐HFpEF | Non‐high‐risk | 2.2 | 0.61 | Reference: 1.00 | |
| High‐risk | 6.6 | 2.90 (1.21–6.97) | |||
| Non‐cardiovascular death | H2FPEF | Non‐high‐risk | 8.6 | 0.52 | Reference: 1.00 |
| High‐risk | 10.7 | 1.22 (0.73–2.04) | |||
| HFA‐PEFF | Non‐high‐risk | 5.4 | 0.57 | Reference: 1.00 | |
| High‐risk | 11.2 | 1.99 (0.95–4.14) | |||
| TSOC‐HFpEF | Non‐high‐risk | 3.6 | 0.63 | Reference: 1.00 | |
| High‐risk | 13.8 | 3.63 (1.86–7.07) |
| Outcome | Score group | Risk category | Incidence rate (per 100 person‐years) | Harrell C‐index | IRR (95% CI) |
|---|---|---|---|---|---|
| Total HFH Frequencies | H2FPEF | Non‐high‐risk | 11.6 | NA | Reference: 1.00 |
| High‐risk | 25.2 | 2.01 (1.18–3.42) | |||
| HFA‐PEFF | Non‐high‐risk | 16.3 | NA | Reference: 1.00 | |
| High‐risk | 22.0 | 1.27 (0.70–2.31) | |||
| TSOC‐HFpEF | Non‐high‐risk | 12.8 | NA | Reference: 1.00 | |
| High‐risk | 25.7 | 1.80 (1.15–2.70) |
For heart failure hospitalization, recurrent event counts were analyzed using Poisson regression, with the number of heart failure hospitalizations as the dependent variable and the logarithm of follow‐up time included as an offset. Results are presented as incidence rate ratios with 95% confidence intervals. Harrell C‐index was not applicable to this analysis. CI, confidence interval; CV, cardiovascular; HFA‐PEFF, Heart Failure Association Pre‐test assessment, Echocardiography and natriuretic peptide, Functional testing, Final etiology; HFH, heart failure hospitalization; HFpEF, heart failure with preserved ejection fraction; HR, hazard ratio; IRR, incidence rate ratio; TSOC, Taiwan Society of Cardiology.
Figure 2. Kaplan–Meier survival curves by various score categories (high‐risk vs non‐high‐risk).
Kaplan–Meier survival curves with hazard ratio (HR) and 95% CI shown for all‐cause mortality by various score categories (high‐risk vs non‐high‐risk), including H2FPEF, HFA‐PEFF, and TSOC‐HFpEF scores, respectively. Borderline significance of survival difference was observed for the H2FPEF score categories (P=0.053), with both high‐risk HFA‐PEFF and TSOC HFpEF categories demonstrating significantly worse survival compared with their non‐high‐risk counterparts, respectively. 0 and 1 represent non‐high‐risk and high‐risk categories. and HR, hazard ratio.
Figure 3. Mortality rates according to different TSOC‐HFpEF score cutoff points (left) and Kaplan–Meier survival curves by TSOC‐HFpEF scores strata (right).
Mortality rates (per 100 person‐years) (left) according to different TSOC‐HFpEF score cutoff points and Kaplan–Meier survival curves (right) shown for different TSOC‐HFpEF score categories. Increased TSOC‐HFpEF scores were associated with higher risks of events (all‐cause death), ranging from 0 (when the score is 0) to as high as 23.9 (when the score is 6). TSOC indicates Taiwan Society of Cardiology; and HFpEF, heart failure with preserved ejection fraction.
Detailed report on the sensitivity and specificity of overall mortality by the TSOC‐HFpEF score is provided in Table S2. Kaplan–Meier curves showed consistent trends across TSOC‐HFpEF score categories (Figure 3). No differences were observed in cardiovascular versus non‐cardiovascular mortality across TSOC HFpEF categories (Figure S3). The TSOC‐HFpEF score demonstrated the highest Harrell C index for death‐related events, regardless of etiologies. Prognostic impact of the NT‐proBNP strata is further detailed in Figure S4. For HF hospitalization, those who were classified into both high‐risk H2FPEF and TSOC‐HFpEF categories—but not high‐risk HFA‐PEFF category—doubled the risk compared with non‐high‐risk categories (Table S3). Only TSOC‐HFpEF score strata significantly predicted recurrent HF hospitalization (IRR: 1.80, 95% CI: 1.15–2.70). Overall, TSOC‐HFpEF score showed the highest area under the receiver operating characteristic curve for HF‐related events (Table S3).
Medication Use
Beta‐blockers were prescribed in 63.8%, RASi in 51.8%, MRA in 44.4%, and SGLT2i in 11.8% of patients in the cohort (Figure S5). The number of foundational therapies is displayed in Figure 4; 12.1% received no foundational therapy, 63.4% received 1–2 classes, and 24.5% received 3–4 classes. No significant differences in therapy use were observed across H2FPEF or HFA‐PEFF risk categories (χ2 P values for differences: 0.68 and 0.89, respectively), although a marginally higher rate of 1–2 classes was noted in the high TSOC HFpEF group (χ2 p value for differences: 0.093) (Figure 4).
Figure 4. Overall proportion and distribution of uptake rate (percentage) of various GDMT classes according to various score categories.
Overall proportion and distribution of uptake rate (in percentage [%]) of various GDMT classes in the HFpEF cohort (left upper) (gray: no use of any class, blue: 1–2 classes, and green: 3–4 classes) and data stratified by different scores strata (others) (left bars: non‐high‐risk category; right bars: high‐risk category). No significant differences were observed in the distribution of GDMT use by different score categories. GDMT indicates guideline‐directed medical therapy.
Use of RASi (including 9.5% angiotensin receptor/neprilysin inhibitor) was associated with reduced all‐cause mortality (HR: 0.56, 95% CI 0.38–0.82, P=0.003; Figure 5). This association remained significant after adjustment for age, sex, BMI, LVEF, prior HF hospitalization, and renal function based on estimated glomerular filtration rate (adjusted HR: 0.65, 95% CI: 0.44–0.95, P=0.028) and consistent across risk categories. No mortality benefit was observed for other therapies, and no significant differences in HF‐related outcomes were noted by therapy class.
Figure 5. Kaplan–Meier survival curves from different classes of GDMT uptake on overall mortality.
Kaplan–Meier survival curves with hazard ratio (HR) and 95% CI for all‐cause mortality from uptake of different GDMT classes, including RASi, beta‐blocker (BB), MRA, and sGLT2i. Use of RASi was significantly associated with better survival (HR: 0.56, 95% CI: 0.38–0.82, P=0.003), whereas other classes of GDMT did not show overall survival differences. BB indicates beta‐blocker; GDMT indicates guideline‐directed medical therapy; MRA, mineralocorticoid receptor antagonist; RASi, renin‐angiotensin system inhibitors; and sGLT2i, sodium‐glucose cotransporter‐2 inhibitors.
DISCUSSION
In this multicenter, prospectively enrolled, high‐risk HFpEF ethnic Asian cohort, we evaluated the feasibility and prognostic performance of established HFpEF scoring systems (H2FPEF and HFA‐PEFF) compared with a newly developed score that integrates structural parameters and NT‐proBNP levels, the TSOC‐HFpEF score. Our findings demonstrate that the TSOC‐HFpEF score provided the strongest prognostic discrimination for survival, whereas H2FPEF and HFA‐PEFF showed modest performance. The original H2FPEF score was designed to estimate the probability of HFpEF in ambulatory patients with unexplained exertional dyspnea, primarily in White populations, using clinical and echocardiographic parameters such as E/e’ and pulmonary artery pressure. Differences in study setting (mixed inpatient population), lack of structural criteria, and distinct baseline characteristics (HFpEF in the Asian population tends to be leaner than in the White population) may explain its lower prognostic utility in our cohort. In contrast, the HFA‐PEFF score incorporates functional, morphological, and biomarker domains, 14 but its complexity and reliance on weighted criteria may limit applicability without external validation in diverse populations. Our cohort likely represented a population with advanced HFpEF, as nearly half of the patients were enrolled during hospitalization and a higher proportion (34.8%) presenting with New York Heart Association functional class III or IV. Consequently, the baseline NT‐proBNP levels in our study were 3‐fold higher than those reported in recent clinical trials, 18 , 19 a factor that may have influenced the relative performance of the evaluated scores. By combining structural and functional measures with NT‐proBNP, the TSOC‐HFpEF score offers a practical and objective tool for risk stratification in clinical practice. Importantly, mortality risk in our study increased sharply when the TSOC‐HFpEF score reached ≥4, with no deaths observed at score 0.
Previous studies have reported modest prognostic value of HFA‐PEFF in Asian HFpEF populations, 20 with area under the receiver operating characteristic curve values (0.63) higher than those observed in our cohort (0.57). Both the HFA‐PEFF and H2FPEF have demonstrated prognostic relevance, but HFA‐PEFF appears more strongly associated with all‐cause mortality. 20 , 21 Meanwhile, H2FPEF is useful for identifying HF events in the HFpEF population. 22 Likewise, differences between these 2 scores have been reported in cardiac amyloidosis manifesting HFpEF phenotype. In that study, the HFA‐PEFF score also demonstrated numerical superiority in prognosis. 23 Although our TSOC‐HFpEF score demonstrated superiority in prognosis compared with HFA‐PEFF and H2FPEF scores, we acknowledged that the components of the TSOC‐HFpEF score were largely defined a priori based on established consensus and guideline‐recommended echocardiographic features. As both TSOC HFpEF and HFA‐PEFF scores are composed predominantly of the extent of cardiac remodeling and natriuretic peptide levels, they may more accurately reflect HFpEF severity and subsequent outcomes. More importantly, we also recognize that the natriuretic peptide cutoff was calibrated within this data set to reflect the high‐risk clinical profile of the enrolled population, which partly explained the superiority of TSOC‐HFpEF score against others. The natriuretic peptide cutoff may therefore need to be further optimized based on the clinical profiles in future studies. Notably, the utilization of pre‐specified enrollment criteria for patients with HFpEF in this study led to a higher AF percentage and potentially more comorbid conditions. Given the frequent coexistence of comorbidities in HFpEF, natriuretic peptide levels 24 should be considered surrogate prognostic markers beyond diagnostic thresholds.
Regarding pharmacological therapy, approximately half of the cohort received RASi, whereas uptake of SGLT2 inhibitors was low (11.8%). In the CHARM (Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity)‐Preserved trial, ARB demonstrated borderline benefit (covariate‐adjusted HR: 0.86 [0.74–1.00], P=0.051) in patients with New York Heart Association class II–IV HF and LVEF >40%. 25 By contrast, the TOPCAT (Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist) trial found no significant effect of MRA in chronic HFpEF. 26 The STRONG‐HF (Safety, Tolerability, and Rapid Optimization, Helped by N‐Terminal Pro‐Brain Natriuretic Peptide [G‐protein/Guidance, noted formally as N‐Terminal Pro‐Brain Natriuretic Peptide testing] of Heart Failure therapies) trial showed that intensive initiation and up‐titration of RASi, BB, and MRA in acute HF significantly reduced all‐cause mortality in patients with LVEF >40%. 27 The relatively small percentage of SGLT2 inhibitors prescription rate may not be sufficient to show therapeutic benefit and prone to survivor bias in a study without randomization design, which likely explained the lack of outcomes benefit of such class of foundational therapy in current registry. These findings also suggest that early initiation of RASi may improve outcomes in acute HF across the LVEF spectrum, although confirmation in future studies is needed.
This study has some limitations. First, the cohort was derived from a Taiwanese HFpEF population characterized by advanced age, high comorbidity burden (AF, hypertension, diabetes, and CKD), and elevated natriuretic peptide levels, which may limit the generalizability to younger or less comorbid populations. Second, nearly half of the patients were enrolled during hospitalization, potentially biasing cohort toward more severe presentations and higher event rates. External validation in larger and independent populations will be essential to determine the robustness and clinical relevance of these observations. Third, although we compared our TSOC‐HFpEF score with 2 widely used HFpEF scoring systems, our results suggest incremental prognostic stratification within a high‐risk HFpEF cohort, rather than definitive discrimination, and should be interpreted as exploratory and hypothesis‐generating. Additionally, the TSOC‐HFpEF score was developed and validated within the same cohort, raising concerns about overfitting and limiting external validity. Fourth, echocardiographic parameters were obtained from routine practice and collected from local, site‐specific measurements, and potential limitations about inter‐observer variability could not be fully addressed.
CONCLUSION
In this nationwide prospective registry of Asian patients, the proposed TSOC‐HFpEF score—derived from 5 easily accessible echocardiographic and biomarker criteria—offered significantly better risk stratification than the H2FPEF or HFA‐PEFF systems. These findings underscore that diagnostic scoring systems for HFpEF can also be effectively repurposed for prognosis, providing clinicians with a practical objective tool to identify high‐risk individuals.
Sources of Funding
This research was funded by Novartis Taiwan. Novartis also provided expert input in the development of the study design but had no role in data collection, data analysis and data interpretation nor in the writing of the report. TSOC holds final responsibility for the decision to submit for publication.
Disclosures
None.
Supporting information
Figures S1–S5. Tables S1–S3.
Data S1. STROBE checklist.
Acknowledgments
We would like to express our gratefulness for Dr Nguyen Ngoc Khoi Truong contribution in the graphs used in this manuscript.
This manuscript was sent to Peng Li, PhD, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.125.047728
For Sources of Funding and Disclosures, see page 12.
References
- 1. Darvish M, Shakoor A, Feyz L, Schaap J, van Mieghem NM, de Boer RA, Brugts JJ, van der Boon RMA. Heart failure: assessment of the global economic burden. Eur Heart J. 2025; 46: ehaf323. doi: 10.1093/eurheartj/ehaf323 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Shahim B, Kapelios CJ, Savarese G, Lund LH. Global public health burden of heart failure: an updated review. Card Fail Rev. 2023;9:e11. doi: 10.15420/cfr.2023.05 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Wei C, Heidenreich PA, Sandhu AT. The economics of heart failure care. Prog Cardiovasc Dis. 2024;82:90–101. doi: 10.1016/j.pcad.2024.01.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Chien SC, Chen YC, Hung CL. Microvascular dysfunction and heart failure with preserved ejection fraction in senescence. Int J Gerontol. 2024;18:134–135. [Google Scholar]
- 5. Dunlay SM, Roger VL, Redfield MM. Epidemiology of heart failure with preserved ejection fraction. Nat Rev Cardiol. 2017;14:591–602. doi: 10.1038/nrcardio.2017.65 [DOI] [PubMed] [Google Scholar]
- 6. Ziaeian B, Fonarow GC. Epidemiology and aetiology of heart failure. Nat Rev Cardiol. 2016;13:368–378. doi: 10.1038/nrcardio.2016.25 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Li YH, Wang CC, Hung CL, Wu YW, Hsu CH, Tsou YL, Wang CH, Wu CK, Lin PL, Chang HY, et al. 2024 guidelines of the Taiwan Society of Cardiology for the diagnosis and treatment of heart failure with preserved ejection fraction. Acta Cardiol Sin. 2024;40:148–171. doi: 10.6515/ACS.202403_40(2).20240206A [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Chang HY, Lee CM, Hung CL, Sung SH, Lin TH, Wu YW, Hwang JJ, Chen WJ, Wang CC. Taiwan society of cardiology heart failure registry 2020: rationale and design. Acta Cardiol Sin. 2024;40:235–241. doi: 10.6515/ACS.202403_40(2).20230822A [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Tromp J, Tay WT, Ouwerkerk W, Teng T‐HK, Yap J, MacDonald MR, Leineweber K, McMurray JJV, Zile MR, Anand IS, et al. Multimorbidity in patients with heart failure from 11 Asian regions: a prospective cohort study using the ASIAN‐HF registry. PLoS Med. 2018;15:e1002541. doi: 10.1371/journal.pmed.1002541 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Tromp J, Teng TH, Tay WT, Hung CL, Narasimhan C, Shimizu W, Park SW, Liew HB, Ngarmukos T, Reyes EB, et al. Heart failure with preserved ejection fraction in Asia. Eur J Heart Fail. 2019;21:23–36. doi: 10.1002/ejhf.1227 [DOI] [PubMed] [Google Scholar]
- 11. Nagueh SF, Smiseth OA, Appleton CP, Byrd BF 3rd, Dokainish H, Edvardsen T, Flachskampf FA, Gillebert TC, Klein AL, Lancellotti P, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2016;29:277–314. doi: 10.1016/j.echo.2016.01.011 [DOI] [PubMed] [Google Scholar]
- 12. Lang RM, Badano LP, Mor‐Avi V, Afilalo J, Armstrong A, Ernande L, Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015;28:1–39.e14. doi: 10.1016/j.echo.2014.10.003 [DOI] [PubMed] [Google Scholar]
- 13. Reddy YNV, Carter RE, Obokata M, Redfield MM, Borlaug BA. A simple, evidence‐based approach to help guide diagnosis of heart failure with preserved ejection fraction. Circulation. 2018;138:861–870. doi: 10.1161/CIRCULATIONAHA.118.034646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Pieske B, Tschöpe C, de Boer RA, Fraser AG, Anker SD, Donal E, Edelmann F, Fu M, Guazzi M, Lam CSP, et al. How to diagnose heart failure with preserved ejection fraction: the HFA‐PEFF diagnostic algorithm: a consensus recommendation from the heart failure association (HFA) of the European Society of Cardiology (ESC). Eur Heart J. 2019;40:3297–3317. doi: 10.1093/eurheartj/ehz641 [DOI] [PubMed] [Google Scholar]
- 15. Peikert A, Solomon SD. Contemporary treatment options in heart failure with preserved ejection fraction. Eur Heart J Cardiovasc Imaging. 2024;25:1517–1524. doi: 10.1093/ehjci/jeae201 [DOI] [PubMed] [Google Scholar]
- 16. Desai AS, Lam CSP, McMurray JJV, Redfield MM. How to manage heart failure with preserved ejection fraction: practical guidance for clinicians. JACC Heart Fail. 2023;11:619–636.doi: 10.1016/j.jchf.2023.03.011 [DOI] [PubMed] [Google Scholar]
- 17. Chiang CE, Hung CL, Wu YW, Lin TH, Ueng KC, Sung SH, Wu CK, Chao TH, Lin HJ, Lin YH, et al. 2023 consensus of Taiwan Society of Cardiology on the pharmacological treatment of chronic heart failure. Acta Cardiol Sin. 2023;39:361–390. doi: 10.6515/ACS.202305_39(3).20230301A [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Anker SD, Butler J, Filippatos G, Ferreira JP, Bocchi E, Böhm M, Brunner‐La Rocca H‐P, Choi D‐J, Chopra V, Chuquiure‐Valenzuela E, et al. Empagliflozin in heart failure with a preserved ejection fraction. N Engl J Med. 2021;385:1451–1461.doi: 10.1056/NEJMoa2107038 [DOI] [PubMed] [Google Scholar]
- 19. Solomon SD, McMurray JJV, Anand IS, Ge J, Lam CSP, Maggioni AP, Martinez F, Packer M, Pfeffer MA, Pieske B, et al. Angiotensin‐Neprilysin inhibition in heart failure with preserved ejection fraction. N Engl J Med. 2019;381:1609–1620. doi: 10.1056/NEJMoa1908655 [DOI] [PubMed] [Google Scholar]
- 20. Egashira E, Sueta D, Komorita T, Yamamoto E, Usuku H, Tokitsu T, Fujisue K, Nishihara T, Oike F, Takae M, et al. HFA‐PEFF scores: prognostic value in heart failure with preserved left ventricular ejection fraction. Korean J Intern Med. 2022;37:96–108.doi: 10.3904/kjim.2021.272 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Verbrugge FH, Reddy YNV, Sorimachi H, Omote K, Carter RE, Borlaug BA. Diagnostic scores predict morbidity and mortality in patients hospitalized for heart failure with preserved ejection fraction. Eur J Heart Fail. 2021;23:954–963. doi: 10.1002/ejhf.2142 [DOI] [PubMed] [Google Scholar]
- 22. Sueta D, Yamamoto E, Nishihara T, Tokitsu T, Fujisue K, Oike F, Takae M, Usuku H, Takashio S, Arima Y, et al. H2FPEF score as a prognostic value in HFpEF patients. Am J Hypertens. 2019;32:1082–1090. doi: 10.1093/ajh/hpz108 [DOI] [PubMed] [Google Scholar]
- 23. Tomasoni D, Aimo A, Merlo M, Nardi M, Adamo M, Bellicini MG, Cani D, Franzini M, Khalil A, Pancaldi E, et al. Value of the HFA‐PEFF and H2FPEF scores in patients with heart failure and preserved ejection fraction caused by cardiac amyloidosis. Eur J Heart Fail. 2022;24:2374–2386.doi: 10.1002/ejhf.2616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Böhm M, Burri H, Butler J, Čelutkienė J, Chioncel O, et al. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42:3599–3726. [Google Scholar]
- 25. Yusuf S, Pfeffer MA, Swedberg K, Granger CB, Held P, McMurray JJ, Michelson EL, Olofsson B, Ostergren J; CHARM Investigators and Committees . Effects of candesartan in patients with chronic heart failure and preserved left‐ventricular ejection fraction: the CHARM‐preserved trial. Lancet. 2003;362:777–781.doi: 10.1016/S0140-6736(03)14285-7 [DOI] [PubMed] [Google Scholar]
- 26. Pitt B, Pfeffer MA, Assmann SF, Boineau R, Anand IS, Claggett B, Clausell N, Desai AS, Diaz R, Fleg JL, et al. Spironolactone for heart failure with preserved ejection fraction. N Engl J Med. 2014;370:1383–1392. doi: 10.1056/NEJMoa1313731 [DOI] [PubMed] [Google Scholar]
- 27. Pagnesi M, Metra M, Cohen‐Solal A, Edwards C, Adamo M, Tomasoni D, Lam CSP, Chioncel O, Diaz R, Filippatos G, et al. Uptitrating treatment after heart failure hospitalization across the Spectrum of left ventricular ejection fraction. J Am Coll Cardiol. 2023;81:2131–2144. doi: 10.1016/j.jacc.2023.03.426 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figures S1–S5. Tables S1–S3.
Data S1. STROBE checklist.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.