Introduction

Globally, 64 million people suffer from heart failure (HF), with a prevalence of 1–3% in adults and more than 6% in Taiwan [1]. Half of acute decompensated heart failure (ADHF) patients died within five years while ischemic heart disease (IHD) is associated with poorer prognoses, insurance burden, and worsening mortality [2]. IHD was more common in Southeast Asia, and the Western Pacific region (more than 50% in ADHF) in REPORT-HF registry [2, 3]. Even with the complexity of etiology in ADHF, left ventricular ejection fraction (LVEF) can be a simple tool for phenotyping in many randomized controlled trials and observative studies [4]. HFmrEF is an intermediate phenotype between HFrEF and HFpEF sharing heterogenous outcomes [1, 3,4,5]. The HFmrEF is similar to the HFrEF on mortality and HF rehospitalization, but it is also similar to the HFpEF on comorbidities [6,7,8,9,10,11]. The use of guideline-directed medical therapy (GDMT) [4, 12] confirmed a survival effect on chronically stable HFrEF while the PARAGON-HF [13] and DELIVER trials [14] had conflicting outcomes in patients with EF > 40%, particularly when it came to ADHF.

As with HFrEF, HFmrEF was associated with a significantly higher incidence of IHD than HFpEF [15], and established IHD had a considerably poorer prognosis [6, 8, 9, 16, 17]. Sudden cardiac death may be more prevalent among patients with HFrEF and HFmrEF due to underlying coronary events [18]. The results of the ESC HF LT Registry indicate that myocardial ischemia contributes to hospitalization in patients with HFmrEF [19]. As the EF decreased by 45% in the CHARM study, the risk of myocardial infarction increased linearly [20]. Further, Vedin O and colleagues demonstrated that HFmrEF was associated with similar MI risks as HFrEF, as well as higher risks than HFpEF [2, 9]. After acute compensation, sequential MI may increase in HFmrEF with potential IHD risk; however, it is relevant to note that most previous studies lacked adequate propensity matching, short-term follow-up, mixed with chronic or acute HF, and a diverse distribution of EF categories. A comprehensive cardiovascular outcome study and long-term observation of ADHF were therefore needed beyond heart failure and cardiovascular mortality.

To resolve the conflict in HFmrEF outcomes, we compared three cohorts by ejection fraction and balanced covariables through propensity matching. Furthermore, ischemic events including subsequent myocardial infarction and stroke after discharge were also surveyed using long-term and large-scale observation. Our goal was to compare admission characteristics, in-hospital course, discharge medication, and cause-specific post-discharge outcomes for patients with acute HFpEF, HFmrEF, and HFrEF.

Methods

Database

A total of 10,000 beds are available at Taiwan’s CGMH Medical System, which includes four tertiary care hospitals and three major teaching hospitals. The CGMH medical database contains records from 1 to 2001 to 31 March 2021, including diagnoses, laboratory data, medications, echocardiography, imaging, and detailed charts. In spite of the fact that the patient’s identity (i.e. the chart number or national identification number) was encrypted, every patient was assigned a personal identification number (PIN), which could be linked to their medical records. The CGMH medical database is described in greater detail elsewhere [21]. Research protocol approved by CGMH Institutional Review Board (IRB no. 202100393B0C601).

Study design

The retrospective, multicenter cohort study examined 38,069 hospitalized patients with acute decompensated heart failure who were discharged during the period of January 1, 2005, through December 31, 2019. It excludes individuals under the age of 18, type I diabetes mellitus, HIV, malignancy, autoimmune diseases, infective endocarditis, major organ transplants, cardiac resynchronization therapy devices, cardioverter defibrillators (ICD), permanent pacemakers (PPM), and absence of in-hospital echocardiography. The echographic examination and the ESC/AHA guidelines definition [4, 12] was used to classify 12,852 patients into three categories: HFrEF, HFmrEF, and HFpEF (Fig. 1).

Fig. 1
figure 1

Enrollment and flowchart of the study

Full size image

Baseline characteristics

There were a number of HF-related comorbidities present, including hypertension, diabetes, dyslipidemia, atrial fibrillation, peripheral arterial disease, venous thromboembolism, chronic obstructive pulmonary disease, gouty arthritis, gastrointestinal bleeding, intracranial hemorrhage, ischemic heart disease, old ischemic stroke, status following coronary artery bypass surgery, and valve surgery. At least once in the one-year period before discharge, comorbidities were identified from outpatient, emergency, or inpatient records.

Discharge medications

A review of medications prescribed within six months of discharge was conducted. The drugs prescribed included antiplatelets, oral anticoagulants, renin-angiotensin system inhibitors (RASi), beta-blockers, dihydropyridine calcium channel blockers (DCCB), calcium channel blockers (CCB), mineralocorticoid receptor antagonists (MRA), diuretics, oral glycemic agents (OHA), insulin therapy, and Satin. In accordance with World Health Organization Anatomical Therapeutic Chemical codes and Taiwan National Health Insurance reimbursement codes, guidelines-directed medical therapy (GDMT) includes RASi, beta-blockers, and MRA.

Survival and cardiovascular outcomes

The primary outcomes are cardiovascular death and rehospitalization for heart failure. As well as the primary outcomes of all-cause mortality, individual HF rehospitalizations, CV death, acute myocardial infarction (AMI), and ischemic stroke (IS), secondary outcomes are also evaluated. According to ICD-9-CM and ICD-10-CM diagnostic codes for any inpatient diagnosis (eAppendix 1 in the Additional File), the occurrence of cardiovascular death and heart failure rehospitalization was calculated. As defined by the National Registry of Death, CV death includes death from heart disease, hypertension, and cerebrovascular disease. Among the principal discharge diagnoses of hospitalization were ischemic stroke, acute myocardial infarction, and heart failure rehospitalization. From the index date to the date of an event or death, or March 31, 2021, whichever occurred first, each patient was followed.

Statistical analysis

The characteristics of patients with each HFrEF, HFmrEF, and HFpEF grade were compared in this study. A Chi-Squared test was used to compare categorical variables between HFrEF, HFmrEF, and HFpEF. ANOVA was used to compare continuous variables between HFrEF, HFmrEF, and HFpEF. Over a 5-year follow-up period, Kaplan–Meier plots were constructed comparing the primary outcome, all-cause mortality, CV death, heart failure rehospitalization, AMI and ischemic stroke. Events are defined as the occurrence of the outcomes of interest within five years of discharge. For survival curve analysis, patients with no events and no deaths during follow-up were censored. A Cox proportional hazard model was used to compare the risks of fatal outcomes (all-cause and cardiovascular death). In comparing non-fatal outcomes (HF rehospitalization, AMI, and ischemic stroke) between groups, Fine and Gray subdistribution hazard models were used. SAS version 9.4 (SAS Institute, Cary, NC) was used for all statistical analyses.

Results

Baseline characteristic of ADHF among three subgroups

Table 1 summarizes the baseline clinical characteristics of the patients. The total number of subjects enrolled was 12,852 (HFrEF 37.6%, HFmrEF 17.3%, HFpEF 45.1%), with a mean in-hospital mortality of 11.7% (13.8% in HFrEF, 10.3% in HFmrEF, 10.5% in HFpEF) (eTable 1 and eFigure 1 in the Additional File). HFrEF was younger than HFpEF and HFmrEF and had a higher rate of ischemic heart disease and post-CABG. In addition to severe pulmonary and hepatic congestion, HFrEF had significantly higher levels of ALT, AST, and NT-proBNP. As with HFmrEF, HFpEF patients had multicomorbidity, including AF and advanced CKD. The HFrEF had a higher proportion of RASi, BBs, and MRA prescriptions before discharge than the HFpEF and HFmrEF, who also had more antiplatelet agents, OAC, and statins.

Table 1 Baseline characteristics of the study patients
Full size table

Outcomes analysis by left ventricular ejection fraction

In the Table 2, all outcomes are analyzed according to the individual definition based on 2.8 years of follow-up (interquartile range, 0.9 to 5.8 years). HFrEF has a significantly higher incidence event of primary endpoint (HFrEF 58.2%; AHR, 1.55; 95% CI, 1.42 to 1.69; HFmrEF 56.4%; AHR, 1.27; 95% CI, 1.14 to 1.41; HFpEF 51.2%, respectively) (Fig. 2 and eTable 2). Additionally, HFrEF has the highest risk for individual outcomes on CV death (AHR, 1.35; 95% CI, 1.24 to 1.47; P < 0.001) and HF rehospitalization (AHR, 1.34; 95% CI, 1.25 to 1.43; P < 0.001) (Fig. 2 and eTable 2). HFpEF has a significantly higher proportion of all-cause mortality compared with the other two groups (HFrEF 56.8%; HFmrEF 61.3%; HFpEF 63.8%; P < 0.001) (Table 2). After adjustment, HFrEF (AHR, 1.24; 95% CI, 1.18 to 1.31; P < 0.001) and HFmrEF (AHR, 1.11; 95% CI, 1.04 to 1.18; P = 0.002) have significant risk compared to HFpEF (eTable 2 and Fig. 3). For AMI, HFmrEF is the highest risky group compared with others (HFrEF 9.9%; AHR, 0.99; 95% CI, 0.87 to 1.13; P = 0.881; HFmrEF 13.6%; AHR, 1.15; 95% CI, 0.99 to 1.32; P < 0.001; HFpEF 9.3%, respectively.); however, HFpEF has significant more events of sequent stroke than other two groups (HFrEF 7.8%; HFmrEF 11.2%; HFpEF 12.0%; P < 0.001). After adjustment, HFrEF still has a significantly lower risk than HFpEF (AHR, 0.72; 95% CI, 0.63 to 0.82; P < 0.001) (eTable 2 and Fig. 3). Despite all-cause mortality, all outcomes initially had similar trends and differences during the first 6 months (Table 2).

Table 2 Statistical analysis of ADHF outcomes according to the ejection fraction at various time points
Full size table
Fig. 2
figure 2

Kaplan–Meier curves for the primary outcome (A), all-cause mortality (B), CV death (C), HF rehospitalization (D), acute myocardial infarction (E), and ischemic stroke (F) of patients with decompensated heart failure by left ventricular ejection fraction

Full size image
Fig. 3
figure 3

 A comparison of the prognostic factors for outcomes based on left ventricular ejection fraction after adjustment

Full size image

Predicting model for AMI in HFmrEF

We analyze the factors predicting sequent AMI in hospitalized HFmrEF in eTable 3 in the Additional File. We found history of hypertension (AHR 1.39; 95% CI 1.00-1.92; p = 0.048) dyslipidemia (AHR, 1.32; 95% CI, 1.02 to 1.70; P = 0.032), Gastrointestinal bleeding (AHR, 1.33; 95% CI, 1.03 to 1.71; P = 0.031), Peripheral arterial disease (AHR 1.53; 95% CI 1.04–2.25; p = 0.033), ischemic heart disease (AHR, 2.04; 95% CI, 1.41 to 2.95; P < 0.001), discharged medication of statin (AHR, 1.44; 95% CI, 1.10 to 1.87; P = 0.007) could be independently associated with further AMI attack. Moreover, a history of atrial fibrillation (AHR, 0.63; 95% CI, 0.46 to 0.87; P = 0.005) seems to be negatively correlated with incidences of AMI. We investigate LDL-C and cholesterol levels while both are relatively higher among those subjects with hypertension, dyslipidemia, gastrointestinal bleeding, peripheral arterial disease, ischemic heart disease and statin use; however, those levels in HFmrEF with AF are significantly lower than those without AF (eTable 4 in the Additional File). The change of EF% during sequent AMI (n = 141) revealed 15.6% maintained at mildly reduced, 46.8% with HFpEF and 37.6% into HFrEF (eFigure 2 in the Additional File).

Discussion

As a result of the present study, significant differences were found between ADHF subgroups with different EF% based on the ESC/AHA definitions [4, 12]. Contrary to previous studies [3, 5-11], this investigation is based on a large-scale, long-term observation, with a secondary comparison using propensity scores. HFpEF had the highest five-year mortality (50.7%), as opposed to HFrEF, which had significantly higher in-hospital mortality (13.8%), HF rehospitalization, and CV death. In both short- and long-term studies, survival rates of survivors with ADHF were similar regardless of EF% after discharge [2, 10]. A further observation is that HFmrEF has a significantly increased rate of AMI, while HFpEF has a slightly increased rate of stroke as a result of ischemic events.

ADHF distribution by EF% and ischemic heart burden

A number of registries maintain HFmrEF at about 10–25% of entries [10, 22]. Our results shows HFpEF as the largest subgroup (45.1%), followed by HFrEF (37.6%), and HFmrEF (17.3%), consistent with the CHARM study (17% for HFmrEF) [23], Spain study (16%) [24], ESC-HF-LT Registry of AHF (18%) [19], CHART-2 (17.1%) [5], KorAHF (16%) [8] and Swedish SwedeHF registry (21%); The disparity between HFpEF and HFrEF can be explained by (1) ADHF versus CHF or mixed; and (2) diverse racial profiling and (3) different healthcare system. In our study, HFmrEF were younger and have more patients with AF and hypertension than HFpEF. HFmrEF and HFpEF had a higher percentage of women than HFrEF. Consequently, HFmrEF has similar characteristics to HFrEF as the number of patients comorbid with ischemic heart disease (IHD) and CABG. CHART-2 Study [5] showed prevalence of IHD 38.8% in HFmrEF and 37.2% in HFrEF; whereas higher prevalence in our ADHF study (HFmrEF 67.6%, HFrEF 61.0%). A significant impact of IHD burden on ADHF could indicate the importance of ASCVD outcomes in addition to the hospitalization for HF.

Prognosis and ADHF across the spectrum of ejection fraction

A study conducted by our group showed that in-hospital mortality was 11.7% as compared to previous studies in the MIMIC-III database (13.5%) [25] and the ARIC community study (6 ~ 12%) [26]. However, the rate is higher than that of the 3 ~ 13% in the China PEACE study [27], 4.8 ~ 7.6% in the KorAHF Registry [7, 28], 2.9 ~ 3.9% in the OPTIMIZE-HF Study [29], ESC-HF-LT Registry (Mortality in HFrEF 3.4% was higher than HFpEF 2.2% or HFmrEF 2.1%) [19, 30]. Higher proportion of diabetes (more than 40%), eGFR < 60 mL/min/1.73m2 (> 40%), AF burden (26.0 ~ 36.2%) in our study influence the in-hospital mortality while average diabetes (34 ~ 38%), CKD (23 ~ 26%), or AF (20 ~ 32%) in ESC-HF-LT registry [19, 30]. As a result of our study, HFrEF was significantly associated with all-cause mortality, CV death, and HF rehospitalization, whereas HFpEF exhibited higher rates of ischemic strokes, but not HFmrEF. The proportion of atrial fibrillation in HFpEF (34.4%) was higher than HFrEF (27.5%) and HFmrEF (29.6%) as SwedeHF (Swedish Heart Failure Registry) registry [31]. Although, AF was associated with similarly increased risk of death, HF hospitalization, and stroke or TIA in all ejection fraction groups [31]. An metanalysis of retrospective data also demonstrated that AF increases all-cause mortality among patients with HFpEF but not among those with HFrEF [32]. The findings of our study suggest that more comorbidities and less discharge OAC use in HFpEF may increase the risk of embolic stroke and AF burden [33]. Furthermore, strokes may have an impact on all-cause mortality by impairing function and causing fatal complications in HFpEF, while clinicians should pay great attention to ASCVD events following ADHF.

Risk of MI and ischemic burden in HFmrEF

Consequently, HFmrEF appears potentially risky following MI compared to HFrEF and HFpEF. Even after adjusting for several patient characteristics, the correlation remains insignificant trend in our study (eTable 2). In spite of the CHART-2 Study finding that AMI, or stroke did not differ significantly between the three groups of people with CHF, the majority of the HFmrEF could transition to HFrEF as a result of IHD burden [5]. There were fewer patients in in CHART-2 Study and a shorter follow-up period than in ours, which enrolled patients with ADHF, not CHF. REPORT-HF registry reported significant IHD burden (53%) in global ADHF and 17% were admitted with new-onset IHD [2]. After discharge, ADHF with IHD had higher incidental AMI events after discharge (IHD 2% vs. non-IHD < 1%, p < 0.001). SwedeHF registry [34] showed HFmrEF has a greater proportion of IHD and AMI history as HFrEF compatible with previous studies (7 to 25% had a history of AMI) [35, 36] and recent meta-analysis [10]. As the results of the ESC HF LT Registry, the Gulf CARE registry reported acute coronary syndrome is a major precipitating factor for De novo ADHF compared to acute decompensated CHF (39% vs. 17%, p < 0.001) [3]. Ola Vedin et al. also found HFmrEF resembled HFrEF rather than HFpEF with regard to both a higher prevalence of IHD and a greater risk of new IHD events [9]. Additionally, we found that higher lipid profiles in patients with histories of IHD, discharge statin use may increase the risk of subsequent MI in patients with HFmrEF (eTable3), and this association decreased after being adjusted for other factors. According to multi-parametric cardiovascular magnetic resonance (CMR), Brown et al. reported a similar degree of fibrosis and microvascular impairment (hyperaemic myocardial blood flow) in HFpEF and HFmrEF, who also exhibit a high prevalence of occult ischemic heart disease as HFrEF [37]. Researchers also found that HFmrEF might be more sensitive to mild ischemic injury than patients with HFrEF or HFpEF based on hs-cTnT levels (HR HFrEF vs. HFmrEF vs. HFpEF: 1.71 vs. 3.76 vs. 1.87) [38, 39]. As a result of our study, 46.1% of patients with HFmrEF transitioned to HFrEF during an AMI attack, while nearly 54% of patients had LVEFs greater than 40%. According to a study done by Farr, 62% of HFmrEF patients remained at LVEF 40–50% while 24% and 33% transitioned to HFrEF and HFpEF, respectively. Since one-third of patients with HFmrEF exhibited decreased LVEF or decreased EF%, underlying coronary artery disease should be investigated [40].

Furthermore, we could confirm a correlation between IHD, particularly between new MI events and ADHF with HFmrEF, which should further reinforce research efforts into the possibly beneficial effects of revascularization in the subgroup. Ischemia type HFmrEF should be closet to HFrEF, and non-ischemic type could be closed to HFpEF focus on comorbidities prevention.

Limitations

Retrospective data were collected from electronic records at multiple centers with hereditary bias. In addition to operator factors, echocardiographic parameters can be affected by interaction factors and the absence of multiple checks during hospitalization. In view of co-existing conditions, such as rapid AF, ADHF did not adjudicate the EF. When the patient was hospitalized, echocardiography (including an assessment of the EF) was performed in accordance with local protocol and routine (there was no core laboratory). There was a lack of information available regarding coronary artery disease and ischemic heart disease. In terms of major diagnoses, it was difficult to differentiate between AMI types. The immediate hemodynamic status, personal function status, and severity of noncardiac complications, such as sepsis, were not reported. Medical care quality and records varied from hospital to hospital, whereas validation analysis was carried out in accordance with previous reports [21]. We were unable to obtain detailed information on the clinical condition of the patient after discharge, as well as his compliance with his medications. In contrast to previous studies, our study makes use of a large cohort of ADHF and propensity scoring for variable balances, which is a unique feature.

Conclusion

Despite appearing intermediate between HFrEF and HFpEF, our study found HFmrEF has a significant risk of subsequent myocardial infarction after ADHF. We considered vascular events as the essential outcome instead of mortality, and those results suggested that aggressive revascularization might reduce HF rehospitalization of HFmrEF patients with ischemia as well as preserve cardiac function. These findings are of importance to future research strategies on prevention and treatment of different HF types and IHD.