Author + information
- Received November 10, 2018
- Revision received November 17, 2018
- Accepted November 26, 2018
- Published online March 25, 2019.
- Gianluigi Savarese, MD, PhDa,∗,
- Ola Vedin, MD, PhDb,c,∗∗ (, )
- Domenico D'Amario, MD, PhDd,
- Alicia Uijl, MD, PhDa,e,
- Ulf Dahlström, MD, PhDf,
- Giuseppe Rosano, MD, PhDg,
- Carolyn S.P. Lam, MD, PhDh and
- Lars H. Lund, MD, PhDa
- aDepartment of Medicine, Karolinska Institutet and Department of Cardiology, Karolinska University Hospital, Stockholm, Sweden
- bDepartment of Medical Sciences, Uppsala University and Uppsala Clinical Research Center, Uppsala, Sweden
- cBoehringer Ingelheim AB, Stockholm, Sweden
- dInstitute of Cardiology, Fondazione Policlinico Universitario A. Gemelli Institute of Scientific Research and Treatment, Università Cattolica del Sacro Cuore, Rome, Italy
- eJulius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands
- fDepartment of Cardiology and Department of Medical and Health Sciences, Linkoping University, Linkoping, Sweden
- gDepartment of Medical Sciences, IRCCS San Raffaele Hospital, Rome, Italy
- hNational Heart Centre Singapore, Duke-NUS Medical School, and University Medical Centre Groningen, Groningen, the Netherlands
- ↵∗Address for correspondence:
Dr. Ola Vedin, Uppsala Clinical Research Centre, Uppsala Science Park, Dag Hammarskjölds väg 38, 751 85 Uppsala, Sweden.
Objectives This study sought to evaluate the incidence, the predictors, and the associations with outcomes of changes in ejection fraction (EF) in heart failure (HF) patients.
Background EF determines therapy in HF, but information is scarce about incidence, determinants, and prognostic implications of EF change over time.
Methods Patients with ≥2 EF measurements registered in the Swedish Heart Failure Registry were categorized as heart failure with preserved ejection fraction (HFpEF) (EF ≥50%), heart failure with midrange ejection fraction (HFmrEF) (EF 40% to 49%), or heart failure with reduced ejection fraction (HFrEF) (EF <40%). Changes among categories were recorded, and associations among EF changes, predictors, and all-cause mortality and/or HF hospitalizations were analyzed using logistic and Cox regressions.
Results Of 4,942 patients at baseline, 18% had HFpEF, 19% had HFmrEF, and 63% had HFrEF. During follow-up, 21% and 18% of HFpEF patients transitioned to HFmrEF and HFrEF, respectively; 37% and 25% of HFmrEF patients transitioned to HFrEF and HFpEF, respectively; and 16% and 10% of HFrEF patients transitioned to HFmrEF and HFpEF, respectively. Predictors of increased EF included female sex, cases of less severe HF, and comorbidities. Predictors of decreased EF included diabetes, ischemic heart disease, and cases of more severe HF. Use of renin-angiotensin-system inhibitors was associated with lower likelihood of EF increase, but not with EF decrease (i.e., stable EF). Increased EF was associated with a lower risk (hazard ratio [HR]: 0.62; 95% confidence interval [CI]: 0.55 to 0.69) and decreased EF with a higher risk (HR: 1.15; 95% CI: 1.01 to 1.30) of mortality and/or HF hospitalizations. Prognostic implications were most evident for transitions to and from HFrEF.
Conclusions Increases in EF occurred in one-fourth of HFrEF and HFmrEF patients, and decreases occurred in more than one-third of patients with HFpEF and HFmrEF. EF change was associated with a wide range of important clinical and organizational factors as well as with outcomes, particularly transitions to and from HFrEF.
Heart failure (HF) phenotyping and treatment decisions rely profoundly on the assessment of ejection fraction (EF), categorizing patients as having heart failure with reduced ejection fraction (HFrEF) (EF <40%), heart failure with midrange ejection fraction (HFmrEF) (EF 40% to 49%), or heart failure with preserved ejection fraction (HFpEF) (EF ≥50%) (1). In addition, EF is not only a marker of cardiac function but also retains independent prognostic information (2). Thus, EF has become a fundamental part of daily clinical protocol and an inclusion criterion and surrogate endpoint in trials (1,3).
For purposes of diagnosis, prognostication, and treatment assignment, a baseline EF assessment is mandatory in every HF patient (1,2). However, EF is not a static measurement but may increase or decrease over time (4–6), warranting renewed classification and prognostic evaluation. Furthermore, indications for HF therapy may arise with deteriorating EF, whereas data for benefits and risks of continuation versus withdrawal of therapy with improving EF are lacking. Given these important implications, comprehensive assessments of changes in EF in large HF populations are warranted. However, most contemporary studies focus solely on recovery of EF (7–13) and do not evaluate the whole spectrum of EF change across all EF categories, including determinants of change and associated prognosis.
The current study was undertaken, therefore, to assess: 1) the incidence and type of EF change (increase or decrease); 2) the predictors associated with different types of EF change; and 3) the prognostic implications of different types of EF change in the large contemporary and unselected SwedeHF (Swedish Heart Failure) study population.
Study protocol and setting
The SwedeHF study was described previously (14). The only inclusion criterion was the clinician judged that the patient had HF. Approximately 80 variables were entered at hospital discharge or after an outpatient clinic visit to complete the Internet-based case report form. The Uppsala Clinical Research Center (Uppsala, Sweden) manages the database. The prevalence of HF in the SwedeHF study was 54% (15), but the first HF onset incidence was only 10% (16).
The Swedish Board of Health and Welfare administers the Population Registry that provided date of death, and the Patient Registry that supplied baseline comorbidities, beyond those available in the SwedeHF cohort, and the outcome HF hospitalization defined according to International Statistical Classification of Diseases and Related Health Problems (revision 10) (ICD-10) codes in the first position. ICD-10 coding has been validated in Sweden. The positive predictive value for most diagnoses is 85% to 95% (17). Diagnosis of HF was verified in 86% to 91% of cases (18).
Socioeconomic data were provided by Statistics Sweden SCB (Stockholm, Sweden). All permanent residents of Sweden, regardless of citizenship, have unique personal identification numbers that allow linking of disease-specific health registries and governmental health and statistical registries.
Establishment of the HF registry and this analysis with linking of the above registries were approved by a multisite ethics committee. Individual patient consent was not required, but patients were informed of entry into national registries and allowed to opt out.
In the current study, patients with at least 2 consecutive EF assessments were enrolled. When the same patient reported more than 2 EF assessments, the first and last assessments were considered in order to calculate the change in EF. EF data were collected, categorized, and defined as HFpEF ≥50%, HFmrEF 40% to 49%, and HFrEF <40%. Transitions from HFpEF to HFmrEF, HFpEF to HFrEF, and HFmrEF to HFrEF were pooled and defined as EF decrease. Transitions from HFrEF to HFmrEF, HFrEF to HFpEF, and HFmrEF to HFpEF were pooled and defined as EF increase. The absence of any changes among EF groups was defined as stable EF. Primary outcome of the analysis was a composite of all-cause mortality and HF hospitalization.
Baseline characteristics of patients according to increasing, decreasing, or stable EF were compared by using Student’s t-test or Wilcoxon-Mann-Whitney test for continuous variables and by chi-square test for categorical variables. Missing data were managed by multiple imputation, using the chained equations method (n = 20). Variables noted in Table 1 were considered for the imputation procedure. All analyses, except for descriptive statistics, were performed using imputed data.
Predictors of EF changes
To assess the independent predictors of EF increase or decrease, multivariate logistic regression models were applied, including EF increase or decrease as a dependent variable, all the variables reported in Table 1 and Online Tables 1 and 2 as covariates with the addition of HF type at baseline, the time between EF measurements, and the year of registration. Because predictors of EF increase or decrease were mostly unknown and because the sample size was sufficiently large, variables for the multivariate models were not selected according to any stepwise variable selection but rather included all variables from the SwedeHF study and the National Patient Registry and Statistics Sweden that were clinically relevant and that could potentially affect EF over time.
Kaplan-Meier curves and multivariate Cox regression models including the same covariates as in the multivariate logistic regression model were fitted to evaluate the association among EF changes (increase, decrease, or stable) and outcomes. The index date was defined as the date of the follow-up EF assessment. The end of follow-up was December 31, 2012. The same models were applied including EF changes categorized as pEF to pEF, pEF to mrEF, pEF to rEF, mrEF to pEF, mrEF to mrEF, mrEF to rEF, rEF to pEF, rEF to mrEF, and rEF to rEF.
Statistical analyses were performed by using Stata version 14.2 software (StataCorp, College Station, Texas). A p value <0.05 was considered statistically significant.
Sample size and EF change patterns
Between May 11, 2000, and December 31, 2012, 51,060 patients were included. Of these, 4,942 patients had undergone at least 2 consecutive EF assessments (median time between EF assessments was 1.4 years [interquartile range (IQR): 0.5 to 3.0 years]), and follow-up was ≥1 day. Overall, 1,027 patients (21%) reported increased EF; 689 (14%) reported decreased EF; and 3,226 (65%) reported unchanged EF. Specifically, 21% and 18% of baseline HFpEF patients transitioned to HFmrEF and HFrEF, respectively; 37% and 25% of baseline HFmrEF patients transitioned to HFrEF and HFpEF, respectively; and 16% and 10% of baseline HFrEF patients transitioned to HFmrEF and HFpEF, respectively (Figure 1).
Baseline characteristics by EF pattern of change
In the overall population, the mean age was 72 ± 12 years of age, 31% were female, 18% had HFpEF, 19% had HFmrEF, and 63% had HFrEF. Table 1 lists characteristics according to an increase, decrease, or no change in EF. Those patients with decreasing EF were among the oldest. Patients with decreasing EF had lower estimated glomerular filtration rate and hemoglobin levels, whereas those with increasing EF had lower N-terminal pro–B-type natriuretic peptide (NT-proBNP) and shorter HF duration. Patients with increasing EF had fewer comorbidities (e.g., hypertension, diabetes, ischemic heart disease) than those with stable or decreasing EF. Additionally, patients with increasing EF were more likely to receive specialized rather than primary care follow-up, to receive scheduled HF nurse-led clinic follow-up, to be married or cohabitating, and to have the highest education level and annual income. Use of angiotensin-converting enzyme inhibitors and angiotensin receptor blockers (ACE inhibitor/ARB), beta-blockers, and mineralocorticoid receptor blockers was highest among patients whose EF increased or did not change, whereas diuretics were more likely to be used in patients with worsening EF, and use of cardiac resynchronization therapy and implantable cardioverter defibrillator was highest in those with stable EF.
Predictors of EF increase and decrease
The differences in Table 1 are unadjusted. Adjusted odds ratios (ORs) for EF increase and decrease after multivariate logistic regression are shown in Figure 2. Variables independently associated with increased EF included shorter HF duration (<6 vs. ≥6 months); absence of ischemic heart disease and coronary revascularization; lower New York Heart Association functional classes (I to II vs. III to IV); lower NT-proBNP levels; higher estimated glomerular filtration rates; higher mean arterial pressures; higher body mass indexes; being married or cohabitating versus living alone; history of hypertension; higher income; history of anemia; planned follow-up in an HF nurse-led clinic; history of atrial fibrillation; history of chronic obstructive pulmonary disease; being an outpatient versus an inpatient; and female sex. On the other hand, variables predicting a decrease in EF included no history of peripheral artery disease; being an inpatient versus an outpatient; male sex; no history of anemia; no planned follow-up in an HF nurse-led clinic; higher NT-proBNP levels; and a history of ischemic heart disease and diabetes. Use of ACE inhibitors/ARBs was associated with less likelihood of increased EF but not with decreased EF (i.e., stable EF). Variables not associated with EF increase or decrease (including beta-blocker and cardiac resynchronization therapy) are reported in Online Tables 1 and 2.
Unadjusted event rates for the composite outcome were 18.0 (95% confidence interval [CI]: 16.2 to 19.9) per 100 patient-years in those who experienced an increase in EF, 43.1 (95% CI: 41.2 to 45.1) per 100 patient-years in those who showed no change in EF, and 57.8 (95% CI: 52.6 to 63.4) per 100 patient-years in those with decreasing EF (log-rank p < 0.0001). After adjustments, compared with stable EF, an increase in EF was associated with statistically significant reduced risk of the composite outcome (hazard ratio [HR]: 0.62; 95% CI: 0.55 to 0.69), whereas a decrease in EF was associated with an increase in risk (HR: 1.15; 95% CI: 1.01 to 1.30) (Figure 3).
Event rates for all types of transitions are shown in Table 2. Additionally, using stable HFrEF as reference, HRs for all types of transition were calculated, showing improved outcomes only for the transition from HFrEF to HFmrEF (HR: 0.55; 95% CI: 0.47 to 0.64) or to HFpEF (HR: 0.42; 95% CI: 0.33 to 0.53) or to stable HFmrEF (HR: 0.73; 95% CI: 0.62 to 0.85).
The effects of change from and to each EF category were also assessed, using no change as the reference. Compared with stable HFpEF, downward transition from HFpEF to HFrEF but not to HFmrEF was associated with harm, whereas transitions to HFpEF from HFrEF but not from HFmrEF were associated with benefit (Figures 1, 4A, and 4D). Compared with stable HFmrEF, both transitions from HFmrEF to HFpEF and to HFrEF were associated with increased risk, whereas transitions to HFmrEF from HFpEF were associated with increased risk and from HFrEF a decreased risk (Figures 1, 4B, and 4E). Compared with stable HFrEF, both transitions from HFrEF to HFmrEF and to HFpEF were associated with a reduced risk, whereas transition to HFrEF from HFpEF but not from HFmrEF was associated with increased risk (Figures 1, 4C, and 4F).
Comparison with previous studies
The features and findings from the present study in relation to those from previous studies of EF changes and EF recoveries are presented in Online Table 3.
In this large, contemporary and unselected HF cohort, EF change over time was a common occurrence across all EF groups. Important factors associated with increasing EF included female sex, indicators of less severe HF, specialized HF follow-up, absence of ischemic heart disease but presence of several other modifiable comorbidities (e.g., anemia and atrial fibrillation), and preserved renal function. Predictors of decreasing EF included concomitant diabetes and ischemic heart disease, lack of specialized HF follow-up, and higher NT-proBNP levels. Use of ACE inhibitors/ARBs was associated with lower likelihood of EF increase, but not with EF decrease, i.e. stable EF. Overall, increased EF was associated with a more favorable outcome, whereas decreased EF portended a poor prognosis. The prognostic differences were most evident for transitions to and from HFrEF.
EF change is a common occurrence
Although an initial EF assessment remains a fundamental tool with which to define, prognosticate, and guide treatment in HF, the implications of longitudinal EF change are becoming increasingly recognized. However, the extent to which EF assessment occurs in real-world HF populations remains insufficiently covered, and hitherto published reports have almost exclusively focused on EF improvement in HFrEF. In the SwedeHF cohort, changes to and from all EF categories were assessed, and it was observed that approximately 25% of patients with HFrEF at baseline transitioned to a higher category and that 10% of patients showed complete recovery. The proportions of patients with EF improvement in HFrEF in the present study are either similar to (7,13) but mostly lower than (6,8–10) those in previous reports, possibly reflecting an older population and more severe HF syndrome and comorbidities, all of which may contribute to a lower potential for improvement. Surprisingly, 39% of HFpEF patients transitioned to lower EF categories, a finding almost identical that from the study by Dunlay et al. (4), where similar to the present study, echocardiograms were performed at the clinician’s discretion, presumably often following episodes of deterioration. However, in a study by Tsuji et al. (6), echocardiography was performed at pre-specified time points, and downward EF transition in HFpEF was consequently observed in approximately 10% of patients. Thus, selection bias likely exaggerated the high proportion of transition from HFpEF, and although it might not have constituted a static HF entity in some, downward transition is in reality probably less common than observed here. The two-thirds of all HFmrEF patients who transitioned either to HFrEF (37% of patients) or to HFpEF (25% of patients) partly supports the hypothesis that HFmrEF represents a transitional state (6,19). Although such a high proportion of transition may partly result from the intermediate and narrow 40% to 49% EF interval and interexamination variability, it could also indicate a multitude of phenotypes with various background factors and disease severity contained in the HFmrEF category, which may in turn warrant more careful assessment and monitoring. Moreover, because HFmrEF resembles HFrEF more than HFpEF, particularly regarding ischemic causes and risk of incident ischemic events (5,20), transition from HFmrEF to HFpEF may reflect recovery after myocardial infarction, whereas downward transition to HFrEF may indicate progressive HF or a new ischemic event, unlike EF deterioration from HFpEF.
Predictors of EF increase and decrease
Predictors of increases in EF included characteristics linked to less severe HF, fewer comorbidities, and shorter HF duration, the last characteristic possibly indicating patients who had HF treatments were recently initiated, with less remodeling and thus a greater potential for recovery. Some of those predictors, including female sex, nonischemic cause, higher blood pressure, and shorter HF duration have been reported previously in smaller populations (21–23). Female sex was associated with both an increase in EF and less likelihood of decrease. This association is concordant with findings from both animal and human studies reporting less remodeling and left ventricular dilation as a result of volume and pressure overload, less severe necrosis and apoptosis following ischemia, and enhanced infarct healing and myocardial recovery in females compared to that in males. Indeed, the observed less-prominent adverse remodeling could constitute an important determinant of the improved prognosis observed among women with HF compared to men with HF (24). Our analysis found concomitant anemia to be a predictor of EF increase, which could be explained by previous data showing that treatment of anemia can lead to EF improvement (25). Atrial fibrillation was also associated with improved EF. One explanation for that may be residual confounding from temporal variations in EF resulting from paroxysmal atrial fibrillation at the baseline measurement or successful rate or rhythm control achieved in patients with persistent atrial fibrillation at the follow-up measurement. Surprisingly, therapy with ACE inhibitors/ARB, beta-blockers, or mineralocorticoid receptor antagonists did not predict EF improvement, in contrast to several previously published analyses (4,26–29). However, understanding the impact of medications in real-world settings is difficult due to confounding by indication and reverse causation. The absent associations may also be explained by the very high use of beta-blockers (≈90%) or the very low use of mineralocorticoid receptor antagonists (≈30%), which may result in insufficient discrimination. Finally, it is of course also plausible that these drugs are ineffective in HFmrEF, although post hoc analysis from randomized trials suggest that at least ARBs and possibly beta-blockers improve outcomes in HFmrEF but not in HFpEF (20,29).
Consistent with the associations observed between factors representing a milder HF syndrome and increasing EF, some of the variables associated with decreasing EF signified a more severe HF state, including higher NT-proBNP levels and being an inpatient versus an outpatient. Notably, an important predictor was history of ischemic heart disease, confirming previous unadjusted findings from the SwedeHF study of more prevalent and incident ischemic heart disease in patients with EF deterioration (5). Surprisingly however, the strongest predictor of EF decrease was the absence of peripheral artery disease, which is difficult to explain other than as a chance finding. The central role of diabetes in HF once again receives confirmation in the present study. Interestingly, although the association between diabetes and HFrEF in previous studies has been determined mainly by coexisting ischemic heart disease, which was adjusted for, the persistent association indicates a more complex relationship. Furthermore, the potential contribution of diabetes itself to HFpEF is becoming increasingly recognized (30).
EF changes and outcomes
EF change was inversely associated with risk of adverse outcome, with the most prominent associations apparent for EF increase, which resulted in a 39% lower risk of death or HF hospitalization. Although the unadjusted event rates in stable HFpEF were similar to those in stable HFrEF rates, patients with HFpEF and decreasing EF had the highest unadjusted event rates, whereas HFrEF patients with increasing EF had the lowest rates, clearly acknowledging the clinical implications of EF change over time and also the prognostic distinction between HF with recovered EF and HFpEF (8,11–13).
Little is known from published studies about the prognostic impact of EF deterioration from HFmrEF and HFpEF. In the study by Dunlay et al. (4), a 7% increased adjusted mortality risk was observed for every 5% EF decrease among HFpEF patients (4), and although EF was categorized in the present study, the findings appear comparable. The most substantial impact of EF change, however, was evident in patients transitioning to and from HFrEF, particularly in those whose EF improved. Partial or complete recovery from HFrEF portended a significantly better prognosis than static HFrEF, HFmrEF, and HFpEF, indicating a more benign HF phenotype with reversible causes amenable to treatment and, in some cases, perhaps alleviated altogether. Similarly, deterioration from HFpEF to HFrEF resulted in worse outcomes than static HFrEF, possibly reflecting a more vulnerable phase in the course of HF and more progressive disease with associated complications.
The prognostic implications of EF change between HFmrEF and HFpEF were less evident. The absent or sometimes even conflicting risk associations observed for EF change between HFmrEF and HFpEF could have resulted from the lower number of patients in these categories and/or by the lack of a linear relationship between EF and risk of outcomes in higher EF segments, as shown in the CHARM (Candesartan in heart failure—assessment of reduction in mortality and morbidity) study (2,20). With particular reference to the dubious prognostic power of EF in the upper EF ranges, other more precise measurement of cardiac function could possibly have more future clinical potential. For instance, the use of global longitudinal strain and global circumferential strain have proven their superiority compared with EF in several settings (31,32), and implications of changes in these parameters over time, particularly relating to prognostic accuracy in chronic HF, warrant further investigation.
The present observational study is subject to confounding. Indeed, although extensive adjustments were performed, the authors cannot rule out potential residual confounding. The follow-up EF measurement was not performed at pre-determined time points. Thus, patients with EF assessments within a short time period might have been less likely to exhibit a change in EF category. Although all multivariate models were adjusted for the time between EF assessments, residual confounding might have been a limitation. Bias by indication due to HF deterioration has been discussed extensively earlier. Finally, inclusion criterion for the SwedeHF study was HF judged by a clinician, thus, the fact that a few patients with HFpEF might not have had HF cannot be ruled out.
In this nationwide HF cohort, EF increase over time occurred in one-fourth of patients with HFrEF and HFmrEF, and a decrease occurred in more than one-third of those with HFpEF and HFmrEF. EF changes were associated with a wide range of important clinical, and organizational factors. Changes in EF were independently associated with outcomes, particularly in patients transitioning to and from HFrEF.
COMPETENCY IN MEDICAL KNOWLEDGE: The findings from this nationwide HF registry population indicate that assessment of change in EF adds important prognostic information, as opposed to single measurements. Moreover, EF change is a common occurrence and is determined by several clinical and organizational factors, some of which are modifiable.
TRANSLATIONAL OUTLOOK: Further studies are warranted to elucidate mechanisms behind EF change, particularly for EF deterioration from HFpEF. Moreover, a better understanding of the efficacy and safety of continued or withdrawn HF therapies is needed in patients with EF improvement.
The authors thank all local investigators and the patients who participated in the SwedeHF study.
↵∗ Drs. Savarese and Vedin are joint first authors.
Supported by grants 2013-23897-104604-23 and 523-2014-2336 to Karolinska Institutet from the Swedish Research Council, grants 20120321 and 20150557 from the Swedish Heart Lung Foundation, and grant 20110120 from the Stockholm County Council. No funding agency had any role in the design or conduct of the study, or collection, management, analysis, or interpretation of the data, or in the preparation or approval of the manuscript. Dr. Savarese has received honoraria from Vifor, AstraZeneca, Roche, and Servier; and has received grants from Boehringer Ingelheim, and Merck Sharp & Dohme. Dr. Lam is a paid consultant for National Heart Centre, National Medical Research Council of Singapore, Abbott Diagnostics, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Boston Scientific, Coarvia, Janssen Research & Development LLC, Medtronics, Menarini, Merck Sharpe & Dohme, Novartis, Stealth BioTherapeutics, and Vifor Pharma; has received grants from National Medical Research Council of Singapore, Boston Scientific, Bayer, Roche Diagnostic, Medtronics, and Vifor Pharma; and is a paid adviser for Roche Diagnostic, AstraZeneca, Novartis, Amgen, Boehringer Ingelheim, and Abbott Diagnostics. Dr. Lund is a consultant for and has received grants from AstraZeneca, Boehringer Ingelheim, Relypsa, Novartis, and Vifor Pharma; a consultant for Sanofi and Bayer; and has received speaker honoraria from Abbott, Novartis, and Vifor Pharma. Dr. Vedin is an employee of Boehringer Ingelheim; and is a paid consultant for and has received speaker fees from Alnylam, Boehringer Ingelheim, Fresenius Medicare, Merck Sharpe & Dohme, Novartis, Orion Pharma, and Servier. Dr. Dahlström has received a research grant from AstraZeneca; and has received consulting and speaker fees from AstraZeneca and Novartis. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- angiotensin-converting enzyme
- angiotensin receptor blocker
- ejection fraction
- heart failure
- heart failure with midrange ejection fraction
- heart failure with preserved ejection fraction
- heart failure with reduced ejection fraction
- Received November 10, 2018.
- Revision received November 17, 2018.
- Accepted November 26, 2018.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- Solomon S.D.,
- Anavekar N.,
- Skali H.,
- et al.
- Lund L.H.,
- Vedin O.,
- Savarese G.
- Dunlay S.M.,
- Roger V.L.,
- Weston S.A.,
- Jiang R.,
- Redfield M.M.
- Vedin O.,
- Lam C.S.P.,
- Koh A.S.,
- et al.
- Tsuji K.,
- Sakata Y.,
- Nochioka K.,
- et al.
- Cioffi G.,
- Stefenelli C.,
- Tarantini L.,
- Opasich C.
- Agra Bermejo R.,
- Gonzalez Babarro E.,
- López Canoa J.N.,
- et al.
- Basuray A.,
- French B.,
- Ky B.,
- et al.
- Lupón J.,
- Díez-López C.,
- de Antonio M.,
- et al.
- Savarese G.,
- Orsini N.,
- Hage C.,
- et al.
- Lund L.H.,
- Carrero J.J.,
- Farahmand B.,
- et al.
- Lam C.S.P.,
- Teng T.-H.K.
- Lund L.H.,
- Claggett B.,
- Liu J.,
- et al.
- Merlo M.,
- Pyxaras S.A.,
- Pinamonti B.,
- Barbati G.,
- Di Lenarda A.,
- Sinagra G.
- Piro M.,
- Della Bona R.,
- Abbate A.,
- Biasucci L.M.,
- Crea F.
- Silverberg D.S.,
- Wexler D.,
- Blum M.,
- et al.
- Solomon S.D.,
- Skali H.,
- Anavekar N.S.,
- et al.
- van Campen L.C.,
- Visser F.C.,
- Visser C.A.
- Cleland J.G.F.,
- Bunting K.V.,
- Flather M.D.,
- et al.
- Cho G.Y.,
- Marwick T.H.,
- Kim H.S.,
- Kim M.K.,
- Hong K.S.,
- Oh D.J.
- Stanton T.,
- Leano R.,
- Marwick T.H.