Introduction
Heart failure (HF) still remains a significant clinical, economic, and social problem. Although the incidence of HF decreases, its prevalence increases (partly due to improved survival), and the number of deaths attributable to HF rises.1 Furthermore, in recent decades, cardiovascular disease has been recognized as the leading cause of death at the population level,2 and HF is one of the main contributors to this trend. In a previous publication, covering the years 1980–2010, an increasing number of deaths from HF was documented in a standardized population.3 The same trend has been observed in an analysis of the Polish national database in subsequent years.4 Furthermore, Pikala et al5 analyzed the years 2000–2014, and found out a decreasing trend in the standard expected years of life lost per living person from cardiovascular diseases, for both men and women. This is also true for the main cardiovascular diseases, such as ischemic heart disease, cerebrovascular disease, and artery disease in general, but not HF. In this case, the numbers are higher by about 5%. In recent decades, the characteristics of patients with HF and their survival have been provided for HF in general, without reference to the HF phenotypes.3,4,6 Despite recent advances in modern HF therapy, the prognosis is serious.7 Mortality remains high, up to 20% after 1 year, reaching 50% after 5 years.8-10 Until recently, the phenotyping of ejection fraction (EF) has not been strictly defined, recognizing the mild categorization of 2 subtypes, HF with reduced EF (HFrEF) and HF with preserved EF (HFpEF). Therefore, the data on HF phenotyping were relative and did not show a uniform picture of the problem. New strict definitions appeared in the European guidelines in 2016.11 Based on the literature, there is a general agreement that the characteristics of patients with HFrEF differ significantly from those of individuals with HFpEF, while patients with HF with mildly reduced EF (HFmrEF) are in between these 2 categories.10,12,13 There are numerous data regarding survival of HF patients stratified by the new classification of phenotype, however, they are often contradictory.8,10,13-20 At the same time, information about the Polish population in this respect is scarce. In some other populations, the prognosis of patients with HFrEF was similar to that of patients with HFpEF, while in others, survival was worse among individuals with HFpEF. Optimized treatment has been well established in HFrEF, but only recently recommendations have been published for patients with HFmrEF and HFpEF.21 Therefore, more studies are needed to analyze survival in different HF phenotypes. Furthermore, there is still a knowledge gap in the prognosis in all subtypes of HF, depending on the population studied.
This study aimed to analyze survival and to identify predictive factors of death depending on the HF phenotypes in hospitalized patients with HF from our referral center.
Patients and methods
The exact methodology is described in our previous manuscript.22 In summary, this was a single-center retrospective study of patients hospitalized at the National Institute of Cardiology in Warsaw, Poland, and all analyses were perfomed in collaboration with the Agency for Health Technology Assessment and Tariff System. The hospitalized patients with a code for HF, either emergency or elective, were included, according to the HF billing codes of the health care payer (a homogeneous group of patients). The enrollment period included hospitalizations between January 2014 and May 2019. In each case, the diagnosis of HF was verified by a designated physician. Only the first hospitalization for HF during this period was used for the analysis. A total of 81 individuals were excluded from the analysis for various reasons (congenital heart disease, prior left ventricular assist device [LVAD], prior orthotopic heart transplantation [OHT], and lack of information concerning EF) (Supplementary material, Figure S1). Medical history included data available before hospitalization (available as per January 2014) or obtained during index hospitalization. Pharmacotherapy included information from discharge charts. Regarding optimal pharmacotherapy with angiotensin-converting enzyme inhibitors (ACEIs), mineralocorticoid antagonists (MRAs), and β-blockers (BBs), we set a limit of at least 50% of the recommended dose in HFrEF. Epinephrine, norepinephrine, dobutamine, or dopamine in higher doses were coded as inotropes. Chronic HF was coded when the period from diagnosis was longer than 6 months. The patients who underwent LVAD or OHT during hospitalization were censored as alive (4 patients with HFrEF) with follow-up time for the event. All patients were phenotyped, taking into account EF from echocardiographic assessment. Those with EF below 40% were designated as HFrEF, with EF in the range of 40%–49% as HFmrEF, and with EF equal to or greater than 50% as HFpEF, in accordance with the recommendations in force at the time of the study.11 The final study population included 2601 patients. Data on their survival status and cause of death were obtained from the National Health Fund. The cause of death was verified by an experienced physician based on death certificates and classified as either cardiovascular or noncardiovascular. The end of follow-up was December 31, 2020, which represents the latest available data on the survival status. The study protocol was approved by the Biomedical Ethics Committee of the National Institute of Cardiology (IK-NPIA-0021-77/1799/2019).
Statistical analysis
The distribution of quantitative data was verified using the Shapiro–Wilk test. In all cases, when the normal distribution was not confirmed, the data are presented as medians with interquartile ranges (IQRs). Qualitative variables are presented as percentages of the total sample and by HF subtype. Patient characteristics were compared by the survival status in the total sample and in subsamples by HF phenotype, using the Mann–Whitney test for quantitative data and the χ2 test for qualitative data. Comparisons of the distribution of quantitative data among the 3 HF phenotypes were made using the Kruskal–Wallis test, while the χ2 test was used to compare the frequency of qualitative data. Survival probability was assessed and presented graphically using the Kaplan–Meier survival curves. The median follow-up time was calculated, followed by the proportion of deaths, including in-hospital deaths.
To identify predictors of all-cause mortality for each HF phenotype, the Cox proportional hazard regression was used. Analyses were performed first as univariable and then as multivariable, using the stepwise approach with forward selection (P <0.05) and backward elimination (P <0.1), including variables that were significantly related to the risk of death in the univariable analysis. No separate survival analysis of hospital deaths was performed due to a low number of endpoint cases. Hazard ratios with 95% CIs were calculated. The significance of the final model was determined. All data analyses were performed with Stata Statistical Software: Release 17 (College Station, TX: StataCorp LLC 2021). The significance level was established at a P value below 0.05.
Results
The survival data are based on the patients with HF hospitalized in a tertiary cardiology center. A total of 2601 patients were included, 1608 with HFrEF (61.8%), 331 with HFmrEF (12.7%), and 662 with HFpEF (25.5%). Men constituted 70.1% of all participants with different distributions according to HF phenotypes, that is, 81.3%, 68.3%, and 44% for HFrEF, HFmrEF, and HFpEF, respectively (P <0.001). Of the total group, 1188 patients (45.7%) were admitted to the hospital as emergencies with the following numbers according to the HF phenotype: HFrEF, 656 (40.8%); HFmrEF, 150 (45.3%); and HFpEF, 382 (57.7%) (P <0.001).
Patient characteristics
In Table 1, we show the baseline patient characteristics of different phenotypes according to the survival status. Sex was not associated with the survival status, while older age was associated with an increased risk of death in all HF phenotypes. Regarding the use of inotropes during hospitalization, an increased use of these drugs, almost 2-fold for HFrEF and over 3-fold for HFmrEF, was observed in the patients who died, as compared with those who survived. Chronic HF was significantly more common in nonsurvivors vs survivors of HFrEF and HFmrEF, and insignificantly more common in HFpEF. The ischemic etiology was more common in the patients who succumbed with HFrEF and HFpEF, but not with HFmrEF.
Parameter | Whole population (n = 2601) | P value | HFrEF (n = 1608) | P value | HFmrEF (n = 331) | P value | HFpEF (n = 662) | P value | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Survivors (n = 1644) | Nonsurvivors (n = 957) | Survivors (n = 957) | Nonsurvivors (n = 651) | Survivors (n = 235) | Nonsurvivors (n = 96) | Survivors (n = 452) | Nonsurvivors (n = 210) | |||||
Age, y, median (IQR) | 62.2 (52–70.4) | 66.5 (59.3–76.7) | <0.001 | 59.8 (50.2–66.7) | 64 (57.4–71.1) | <0.001 | 63.9 (52.1–72.4) | 70.6 (62.4–79.7) | <0.001 | 68.2 (57.5–78.3) | 76.1 (67.2–83.1) | <0.001 |
Male sex | 1135 (69) | 689 (72) | 0.11 | 775 (81) | 532 (81.7) | 0.71 | 163 (69.4) | 63 (65.6) | 0.51 | 197 (43.6) | 94 (44.8) | 0.78 |
Chronic HF | 1365 (83.1) | 869 (90.9) | <0.001 | 808 (84.4) | 601 (92.3) | <0.001 | 187 (79.6) | 85 (88.5) | 0.05 | 370 (82.2) | 183 (87.6) | 0.08 |
Ischemic etiology | 581 (35.3) | 462 (48.3) | <0.001 | 433 (45.3) | 368 (56.5) | <0.001 | 79 (33.6) | 35 (36.5) | 0.62 | 69 (15.3) | 59 (28.1) | <0.001 |
Inotropes | 116 (7.1) | 231 (24.1) | <0.001 | 75 (7.8) | 159 (24.4) | <0.001 | 7 (3) | 25 (26) | <0.001 | 34 (7.5) | 47 (22.4) | <0.001 |
Emergency admission | 611 (37.2) | 577 (60.3) | <0.001 | 311 (32.5) | 345 (53) | <0.001 | 81 (34.5) | 69 (71.9) | <0.001 | 219 (48.5) | 163 (77.6) | <0.001 |
Comorbidities and procedures | ||||||||||||
Significant MR | 625 (38.5) | 505 (53.8) | <0.001 | 446 (46.9) | 397 (62) | <0.001 | 66 (28.2) | 33 (35.5) | 0.2 | 113 (25.8) | 75 (36.4) | 0.006 |
Significant TR | 436 (27.2) | 469 (50.4) | <0.001 | 263 (28) | 323 (50.9) | <0.001 | 48 (21) | 41 (44.6) | <0.001 | 125 (28.7) | 105 (51.5) | <0.001 |
Aortic stenosis | 114 (7) | 118 (12.4) | <0.001 | 34 (3.6) | 50 (7.7) | <0.001 | 15 (6.4) | 17 (17.9) | 0.001 | 65 (14.5) | 51 (24.4) | 0.002 |
PAD | 136 (8.3) | 154 (16.1) | <0.001 | 85 (8.9) | 107 (16.4) | <0.001 | 17 (7.3) | 16 (16.7) | 0.01 | 34 (7.5) | 31 (14.8) | 0.004 |
Hypertension | 1308 (79.6) | 810 (84.6) | 0.001 | 726 (75.9) | 532 (81.7) | 0.005 | 198 (84.3) | 84 (87.5) | 0.45 | 384 (85) | 194 (92.4) | 0.008 |
CAD | 1132 (68.9) | 710 (74.2) | 0.004 | 695 (72.6) | 501 (77) | 0.05 | 163 (69.4) | 65 (67.7) | 0.77 | 274 (60.6) | 144 (68.6) | 0.048 |
Myocardial infarction | 549 (33.4) | 450 (47) | <0.001 | 422 (44.1) | 372 (57.1) | <0.001 | 72 (30.6) | 30 (31.3) | 0.91 | 55 (12.2) | 48 (22.9) | <0.001 |
AF | 883 (53.7) | 635 (66.3) | <0.001 | 478 (49.9) | 416 (63.9) | <0.001 | 131 (55.7) | 66 (68.7) | 0.03 | 274 (60.6) | 153 (72.9) | 0.002 |
Stroke / TIA | 171 (10.4) | 144 (15.1) | <0.001 | 100 (10.5) | 104 (16) | 0.001 | 17 (7.2) | 17 (17.7) | 0.004 | 54 (12) | 23 (11) | 0.71 |
Dementia | 38 (2.3) | 37 (3.9) | 0.02 | 17 (1.8) | 23 (3.5) | 0.03 | 5 (2.1) | 5 (5.2) | 0.14 | 16 (3.5) | 9 (4.3) | 0.64 |
Diabetes mellitus | 557 (33.9) | 407 (42.5) | <0.001 | 328 (34.3) | 294 (45.2) | <0.001 | 83 (35.3) | 37 (38.5) | 0.58 | 146 (32.3) | 76 (36.2) | 0.32 |
Anemia | 227 (13.8) | 266 (27.8) | <0.001 | 108 (11.3) | 148 (22.7) | <0.001 | 32 (13.6) | 36 (37.5) | <0.001 | 87 (19.3) | 82 (39.1) | <0.001 |
Renal dysfunction | 412 (25.1) | 469 (49) | <0.001 | 224 (23.4) | 316 (48.5) | <0.001 | 54 (23) | 43 (44.8) | <0.001 | 134 (29.7) | 110 (52.4) | <0.001 |
COPD | 355 (21.6) | 294 (30.7) | <0.001 | 208 (21.7) | 186 (28.6) | 0.002 | 40 (17) | 28 (29.2) | 0.01 | 107 (23.7) | 80 (38.1) | <0.001 |
Liver dysfunction | 47 (2.9) | 78 (8.2) | <0.001 | 30 (3.1) | 60 (9.2) | <0.001 | 9 (3.8) | 5 (5.2) | 0.57 | 8 (1.8) | 13 (6.2) | 0.003 |
Thyroid disease | 473 (28.8) | 291 (30.4) | 0.38 | 243 (25.4) | 199 (30.6) | 0.02 | 64 (27.2) | 28 (29.2) | 0.72 | 166 (36.7) | 64 (30.5) | 0.12 |
Cancer | 401 (24.4) | 242 (25.3) | 0.61 | 173 (18.1) | 142 (21.8) | 0.06 | 66 (28.1) | 29 (30.2) | 0.70 | 162 (35.8) | 71 (33.8) | 0.61 |
At least 1 comorbidity | 1598 (97.2) | 949 (99.2) | 0.001 | 929 (97.1) | 645 (99.1) | 0.006 | 227 (96.6) | 96 (100) | 0.06 | 442 (97.8) | 208 (99) | 0.21 |
PCI | 441 (26.8) | 344 (36) | <0.001 | 320 (33.4) | 268 (41.2) | 0.002 | 62 (26.4) | 28 (29.2) | 0.61 | 59 (13.1) | 48 (22.9) | 0.001 |
CABG | 151 (9.2) | 156 (16.3) | <0.001 | 96 (10) | 111 (17.1) | <0.001 | 29 (12.3) | 13 (13.5) | 0.77 | 26 (5.8) | 32 (15.2) | <0.001 |
Any ablation | 181 (11) | 109 (11.4) | 0.77 | 110 (11.5) | 90 (13.8) | 0.16 | 29 (12.3) | 11 (11.5) | 0.82 | 42 (9.3) | 8 (3.8) | 0.01 |
Valvular surgery | 194 (11.8) | 157 (16.4) | 0.001 | 89 (9.3) | 86 (13.2) | 0.013 | 31 (13.2) | 15 (15.6) | 0.56 | 74 (16.4) | 56 (26.7) | 0.002 |
ICD/ICD-CRT | 640 (38.9) | 468 (48.9) | <0.001 | 492 (51.4) | 417 (64.1) | <0.001 | 52 (22.1) | 26 (27.1) | 0.34 | 96 (21.2) | 25 (11.9) | 0.004 |
Pharmacotherapy | ||||||||||||
ACEIs – optimal dose | 1048 (63.9) | 446 (47.9) | <0.001 | 662 (69.3) | 325 (51.3) | <0.001 | 162 (68.9) | 41 (45.6) | <0.001 | 224 (49.7) | 80 (38.3) | 0.006 |
BBs – optimal dose | 1057 (64.4) | 527 (56.5) | <0.001 | 662 (69.3) | 381 (60.2) | <0.001 | 147 (62.5) | 51 (56.7) | 0.33 | 248 (55) | 95 (45.5) | 0.02 |
MRAs – optimal dose | 1146 (69.8) | 620 (66.5) | 0.08 | 782 (81.9) | 473 (74.7) | 0.001 | 152 (64.7) | 42 (46.7) | 0.003 | 212 (47) | 105 (50.2) | 0.44 |
Loop diuretics | 1353 (82.4) | 891 (95.5) | <0.001 | 843 (88.3) | 612 (96.7) | <0.001 | 179 (76.2) | 85 (93.4) | <0.001 | 331 (73.4) | 194 (92.8) | <0.001 |
VKA or NOAC | 908 (55.3) | 612 (65.6) | <0.001 | 523 (54.8) | 410 (64.8) | <0.001 | 120 (51.1) | 59 (64.8) | 0.03 | 265 (58.8) | 143 (68.4) | 0.02 |
Amiodarone | 252 (15.4) | 197 (21.1) | <0.001 | 167 (17.5) | 171 (27.1) | <0.001 | 36 (15.3) | 10 (11) | 0.31 | 49 (10.9) | 16 (7.7) | 0.2 |
Digoxin | 236 (14.4) | 163 (17.5) | 0.04 | 158 (16.5) | 112 (17.7) | 0.54 | 28 (11.9) | 12 (13.2) | 0.75 | 50 (11.1) | 39 (18.7) | 0.008 |
Data are presented as number (percentage) of patients unless indicated otherwise. Abbreviations: ACEIs, angiotensin-converting enzyme inhibitors; AF, atrial fibrillation; BBs, β-blockers; CABG, coronary artery bypass graft; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; CRT, cardiac resynchronization therapy; HF, heart failure; HFmrEF, heart failure with mildly reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; ICD, implantable cardioverter-defibrillator; IQR, interquartile range; MR, mitral regurgitation; MRAs, mineralocorticoid receptor antagonists; NOAC, non–vitamin K antagonist oral anticoagulant; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; TIA, transient ischemic attack; TR, tricuspid regurgitation; VKA, vitamin K antagonist | ||||||||||||
Comorbidities
The median number of comorbidities was significantly higher in the patients with an adverse prognosis (Table 1). A significantly higher proportion of nonsurvivors was identified in all phenotypes for atrial fibrillation, anemia, renal dysfunction, chronic obstructive pulmonary disease, peripheral artery disease, aortic stenosis, and significant tricuspid regurgitation (TR). However, the remaining comorbidities showed different effects in different HF phenotypes. Of the known risk factors of an adverse prognosis in HF, a higher percentage of nonsurvivors was observed in HFrEF and HFpEF with respect to liver dysfunction, coronary artery disease, postmyocardial infarction, and hypertension. Diabetes significantly affected patient survival only in HFrEF.
Pharmacotherapy and procedures
Standard HF pharmacotherapy with optimal ACEI dosages was taken by a high proportion of survivors in all subgroups of HF. MRAs were significantly associated with a better prognosis in HFrEF and HFmrEF, while BBs were associated with a more favorable prognosis in HFrEF and HFpEF (Table 1). In contrast, the use of loop diuretics (DIURs) was greater in the patients with an adverse prognosis in all subgroups of HF. Out of other drugs, anticoagulation was more frequent in the patients who died vs the survivors in all subtypes of HF, while amiodarone and digoxin were more often taken only in HFrEF and HFpEF, respectively. In our study group, the patients who underwent invasive procedures more often belonged to the deceased group with HFrEF and HFpEF, but not with HFmrEF.
Mortality and prognosis
The median follow-up time was 2.43 years (IQR, 1.56–3.49), and was comparable in all groups (Table 2). The highest mortality rate during follow-up was observed in the HFrEF group, and it differed significantly from that in the HFmrEF and HFpEF groups, which were similar. The Kaplan–Meier survival curves (Figure 1) showed that survival in the HFrEF patients was worse as compared with the HFpEF individuals, while the prognosis in HFmrEF and HFpEF patients was similar. Preliminary comparisons of the death risk for the HF phenotypes, after adjustment for sex and age, revealed by 61% (P <0.001) higher risk of death of HFrEF than of HFpEF, while there was no difference between HFmrEF and HFpEF. The survival rate at 1 and 5 years for HFrEF, HFmrEF and HFpEF was 81%, 84%, and 84%, and 47%, 61%, and 59%, respectively.
Parameter | Whole population (n = 2601) | HFrEF (n = 1608) | HFmrEF (n = 331) | HFpEF (n = 662) | P value |
|---|---|---|---|---|---|
In-hospital death | 113 (4.3) | 77 (4.8) | 13 (3.9) | 23 (3.5) | 0.35 |
Death | 957 (36.8) | 651 (40.5) | 96 (29) | 210 (31.7) | <0.001a,b,c |
Cardiovascular deathd | 652 (70.6) | 481 (77) | 53 (57.6) | 118 (57.3) | <0.001a,b,c |
Identified causes of death, % | 96.5 | 96 | 95.8 | 98.1 | 0.34 |
Follow-up, y, median (IQR) | 2.43 (1.56–3.49) | 2.41(1.31–3.58) | 2.53 (1.81–3.33) | 2.43 (1.78–3.36) | 0.68 |
Data are presented as number (percentage) of patients unless indicated otherwise. a P <0.001 for HFrEF vs HFmrEF b P <0.001 for HFrEF vs HFpEF c P nonsignificant for HFmrEF vs HFpEF d Presented as a percentage of all known causes of death Abbreviations: see Table 1 | |||||
Figure 1. Kaplan–Meier survival curves by different heart failure phenotypes
Abbreviations: see Table 1
In the patients with HFrEF, we also observed the highest percentage of cardiovascular death, and this differed significantly from the individuals with HFmrEF and HFpEF, both displaying a similar death rate (Table 2). The frequency of death from unknown causes was comparable in all the groups (P = 0.34).
Independent predictors of survival
Based on the univariable predictors of overall mortality (Supplementary material, Table S1), we created 3 different models for each of the HF phenotypes. According to the results of the Cox proportional risk analysis (Figure 2), in all subgroups of HF, the use of inotropes during hospitalization was associated with an adverse outcome, while ACEIs reduced the mortality risk. In HFrEF, all other variables were negative predictors of survival. In addition to the use of inotropes during hospitalization, DIUR at discharge, liver dysfunction, a history of HF, and significant TR had the strongest negative predictive value on survival in HFrEF. In HFmrEF and HFrEF, inotropes were the strongest predictors of death, followed by DIURs, aortic stenosis, and stroke, while MRA use was beneficial.
Figure 2. Predictors of all-cause death in the multivariable Cox proportional hazard models according to the heart failure phenotype
Abbreviations: see Table 1
In HFpEF, apart from inotropes, liver dysfunction was the strongest predictor of death, followed by significant TR and ischemic etiology, while ablation procedures were associated with a positive prognosis.
Discussion
To our knowledge, this is the first long-term survival study in the Polish HF population based on a cohort of hospitalized patients with HF stratified by different phenotypes. We have shown that survival was significantly worse in HFrEF than HFmrEF and HFpEF patients, in whom it was similar. Survival analysis revealed that all phenotypes differed in terms of significant predictors of adverse prognosis. It should be emphasized that our study covers a relatively long time, with new guidelines and therapies emerging during this period.
Patient characteristics
Analysis of patient characteristics stratified by the HF phenotype revealed that in terms of the survival status, approximately one-third of the variables had a similar distribution in all the groups. Most of the variables were associated with more advanced HF, as nearly 50% of the hospitalizations were recorded as urgent admissions. As previously reported,22 inotropes were used more frequently in HFrEF, which is consistent with previous publications.14,15,23 Interestingly, we showed that the percentage of inotrope use in HFrEF, HFmrEF, and HFpEF nonsurvivors was comparable and associated with similarly unfavorable prognosis. On the other hand, ACEIs, which are recommended in HFrEF and HFmrEF, showed an association with better survival, regardless of the HF phenotype. Furthermore, BBs (which have recently been questioned in HFpEF)24 were also more frequently taken in the survivor group with HFrEF and HFpEF. Contrary to the current evidence,24 in our study, digoxin use was not associated with an unfavorable prognosis in HFrEF, showing only a deleterious effect in HFpEF.
Most variables, significantly more frequent in the nonsurvivor group, have been reported previously as predictors of death in different HF phenotypes. An interesting finding of our study was that TR was almost twice as frequent in the deceased than in the survivors, regardless of the HF phenotype. In the literature, significant association of TR with a worse prognosis was also reported in patients with HF (without phenotype stratification),25 and in patients with EF below 50%.26
Prognosis by heart failure phenotype
According to some previous reports, hospital mortality did not differ between HFrEF, HFmrEF, and HFpEF,15,23 although differences were observed in some other publications.13,14 Our reported death rates were comparable to most studies on acute HF (AHF),13,15,23 while in some of them they were much higher, reaching 15%–20%.15,27 The relatively high mortality rate in our sample, including not only AHF, may be explained by the character of our center that is a highly-specialized cardiology hospital accepting the most complicated cases.
In general, HF mortality remains high and is estimated to be 15%–30% at 1 year, 30%–50% at 3 years, and 50%–75% at 5 years.9 Thus, our results are comparable with previous data. However, depending on the population and the period studied, there are different data on the survival rates and differences between the HF phenotypes.8,10,13-20 In our study, as in the long-term ESC registry, the patients with HFrEF had worse survival rates than the other participants, while the HFmrEF and HFpEF patients had similar survival rates.13 Some studies reported similar survival rates regardless of the HF phenotype, based on hospitalized patients8,14,15,28 and stable ambulatory patients.20 In other works, the individuals with stable HFrEF showed the highest mortality,19 while in hospitalized patients in a study from Finland, HFpEF was associated with the poorest survival.17 A meta-analysis of 12 studies found that mortality in HFmrEF was significantly lower than in HFpEF and HFrEF.16 An interesting aspect concerning the discrepancies described involves differences in the selection of the study populations and the years in which the studies were conducted. On the other hand, EF is a dynamic state, which may potentially influence the reported results. Patients with improvement in EF had better survival, regardless of the initial HF phenotype.29 Furthermore, comorbidity profile can substantially influence the prognosis.30
Comorbidities are frequently reported in all subtypes of HF. Therefore, they are often recognized as important factors driving the prognosis.19,21 Our findings reporting a higher rate of noncardiovascular death in HFpEF and HFmrEF are similar to other studies.10,17,19 However, there are also publications showing higher30 or similar15 rates of noncardiovascular death in HFpEF, as compared with the other 2 HF phenoptypes. It must be understood that in HFmrEF the predominant subgroup is usually represented by patients with dynamic changes in EF, as compared with stable EF during longitudinal observations.28,32 As we did not evaluate changes in EF during a longer observation, we may only speculate that it could be the explanation behind those findings, showing a closer relationship of HFmrEF with noncardiovascular death. We can conclude that HF-related factors have a minor prognostic impact on both HFmrEF and HFpEF, and finally that comorbidities determine the prognosis over extended time periods.19
Predictors of survival over extended time periods
Since the population stratified by survival status according to the HF phenotypes differs significantly in baseline characteristics, it is not surprising that it affects independent predictors of prognosis. It should be noted that our population represents hospitalized patients with HF (both admitted as emergency and elective) from a specialized center, which may bias these results. In general, prognostic factors, regardless of the HF phenotype, are easily accessible in clinical practice, including demographic factors, valvular disease, comorbidities, and disease progression. Only inotrope use during hospitalization and prescription of ACEIs at higher doses significantly influenced the prognosis, regardless of the HF phenotype. We did not find sex to be a predictor of mortality, which could be due to the specificity of the population studied. However, in numerous studies sex did not show influence on prognosis after adjustment for covariates or was not reported.19,20,33 Age is the main determinant of prognosis in most studies,12,14,17 but some reports failed to find its uniform effect on prognosis.19 In the present study, we found age to be a predictive factor only in HFrEF and HFpEF. At this point, it should be noted that our population is generally younger than in some previous reports, which may affect our results.9,15,17,19 TR, liver and kidney dysfunction, along with DIUR use were found to be associated with a poor prognosis in HFrEF. This may be indicative of more advanced HF, with the progression of left ventricular failure and coexisting right ventricular failure leading to fluid retention, congestion, and finally organ failure.26 The negative association of implantable cardioverter defibrillator / cardiac resynchronization therapy (ICD-CRT) with prognosis appears to confirm the progression to more advanced HF, and a longer period of HF morbidity when electrotherapy may be less effective. This notion seems to be supported by the significant effect of a longer history of HF on survivability. Despite the prevalent literature reports,9,14,34 we did not find a significant role of ischemic etiology, other than a history of MI modifying the prognosis in HFrEF. Similar findings were also observed in other studies.12,17,19,33 In our opinion, this is due to adjustment for many other covariates that weaken the influence of a particular etiology. However, it should not be ruled out that patients in cardiac centers with a history of ischemia can be closely monitored.
In terms of HFmrEF, we observed a mixture of predictive factors that were also documented for HFrEF or HFpEF. The optimal dose of MRAs was found to be protective in this phenotype, as in some other clinical studies.35 These findings confirm the hypothesis that HFmrEF may represent a mixture of dynamic changes in the HF phenotypes and stable EF.32
An interesting observation comes from the analysis of predictors of HFpEF prognosis. Interestingly, contrary to HFrEF, ICD/ICD-CRT had a beneficial effect on survival in this group. As the current recommendation21 for ICD qualification is clear, it may support the notion that patients receiving electrotherapy must have been primarily in the HFrEF subtype, and represented a group with significant improvement over a long period of time. Interestingly, ablation procedures, which were not found to be predictive in other subtypes, beneficially modify prognosis in HFpEF. As tachyarrhythmias are especially detrimental in HFpEF, this may well be the explanation behind our findings.36
Limitations
Our study has several limitations that should be mentioned. This analysis involved retrospective data based on the coding of hospitalizations for HF for administrative purposes. Therefore, some patients with HF who were never recorded as hospitalized for HF were not included. It should also be acknowledged that our overall population was generally younger than in many other reports examining patients with HF. In addition, we analyzed only the patients from a tertiary cardiac center in Poland who were referred to that center for various reasons. Therefore, the population selection error cannot be excluded. The EF evaluation that justified the stratification according to the HF phenotype was not standardized, which could have led to possible differences and misclassification of some patients. Distribution of the HF phenotypes was uneven, and the number of patients with HFmrEF was substantially lower than in the 2 other groups. Furthermore, we did not have information on changes in EF before the study, which did not allow us to observe the history of possible phenotypic changes before hospitalization. However, despite the possible variability between inter- and intraobservers and possible confounders, such as heart rate, load status, and incidental treatment, echocardiography remains the most widely used tool in the evaluation of left ventricular function and stratification of HF. The presence of comorbidities was determined only based on information from medical records, without the possibility of verifying the diagnosis. Furthermore, information on classic cardiovascular risk factors was not available in medical records. We did not analyze the effects of in-hospital invasive interventions. Finally, we were unable to objectively assess clinical status, allowing categorization by the New York Heart Association class on admission to the hospital, or obtain information on pharmacotherapy prior to the start of the study.
Conclusions
Survival probabilities are the lowest in patients with HFrEF, as compared with HFmrEF and HFpEF, where they are similar. This variable did not change significantly after adjustment for age and sex. We found that most of the parameters analyzed, including comorbidities and demographic factors, distinguished survivors from nonsurvivors, and the prevalence of these parameters differed according to the HF phenotype. However, after adjustment, most of them lost their impact on the prognosis, indicating that each phenotype should be analyzed separately, considering other multiple possible prognostic factors. In addition, it is important to realize that our study included patients recruited from among those hospitalized in a specialized cardiac center, so although the results are mostly consistent with previous reports, they cannot be easily generalized.
Tomasz M. Rywik, MD, PhD, Department of Heart Failure and Transplantology, National Institute of Cardiology, ul. Alpejska 42, 04-628 Warszawa, Poland, phone: +48 22 343 44 83, email: trywik@ikard.pl
February 20, 2023.
June 26, 2023.
June 30, 2023.
The authors thank the Heads of Departments of the National Institute of Cardiology for consenting to the use of data from patients treated in these clinics. Furthermore, we acknowledge the role of the physicians caring for this group of patients by listing just some of them: M. Banaszewski, M. Grabowski, M. Kordybach, K. Kryczka, B. Pregowska-Chwała, A. Sioma, J. Pregowski, K. Przybyłowska, P. Syska, J. Zakrzewska-Koperska. We are grateful to all the doctors who performed echocardiograms at the National Institute of Cardiology for the patients included in the study.
The study was supported by a research grant from the National Institute of Cardiology (2.37/VII/2019; to TMR).
Conceptualization: TMR, AD, AgD, and RT-M; data curation: TMR, AgD, HL, and PP; formal analysis: AgD, AD, TMR, and HL; funding acquisition: TMR; investigation: PP, AD, TZ, UC, AW, and HL; methodology: TMR, AgD, AD, and TZ; project administration: TMR, AgD, and AD; resources: HL, AW, UC, and RT-M; software: AgD; supervision: TZ, TMR; validation: AgD, PP, TMR; visualization: AgD, TMR, and AD; writing – original draft: TMR and AGD; writing – review and editing: AgD, AD, RT-M, TZ, TMR, HL, UC, AW, and PP. All authors have made substantial contributions to all of the following: 1) the conception and design of the study, or the acquisition of data, or analysis and interpretation of data; 2) writing of the article or its critical review for important intellectual content; and 3) final approval of the version to be submitted.
None declared.
Rywik TM, Wiśniewska A, Cegłowska U, et al. Heart failure with reduced, mildly reduced, and preserved ejection fraction: outcomes and predictors of prognosis. Pol Arch Intern Med. 2023; 133: 16522. doi:10.20452/pamw.16522
- 1.
- Tsao CW, Aday AW, Almarzooq ZI, et al. Heart disease and stroke statistics - 2022 update: a report from the American Heart Association. Circulation. 2022; 145: e153-e639.
- 2.
- Nichols M, Townsend N, Scarborough P, et al. Cardiovascular disease in Europe 2014: epidemiological update. Eur Heart J. 2014; 35: 2929.Crossref
- 3.
- Rywik TM, Koziarek J, Piotrowski W, et al. Trends in heart failure mortality in Poland between 1980 and 2010. Pol Arch Med Wewn. 2013; 123: 664-671.Crossref
- 4.
- Leszek P, Zaleska-Kociecka M, Was D, et al. Real world heart failure epidemiology and outcome: a population-based analysis of 1 990 162 heart failure patients. Paper presented at ESC Congress 2021, August 27-30, 2021; online.Crossref
- 5.
- Pikala M, Maniecka-Bryła I. Fifteen-year mortality trends due to cardiovascular diseases in Poland using standard expected years of life lost, 2000-2014. Kardiol Pol. 2017; 75: 1033-1040.Crossref
- 6.
- Davidkovova H, Kysely J, Kriz B, et al. Trends in cardiovascular mortality and hospitalisations, and potential contribution of inhospital case-fatality rates to changes in national mortality in the Czech Republic 1994-2009. Heart (British Cardiac Society). 2013; 99: 409.Crossref
- 7.
- Polaska P, Jerzak-Wodzynska G, Smigielski W, et al. Long term outcome of heart failure patients disqualified from heart transplantation. Acta Cardiol. 2021; 76: 525-533.Crossref
- 8.
- Shah KS, Xu H, Matsouaka RA, et al. Heart failure with preserved, borderline, and reduced ejection fraction: 5-year outcomes. J Am Coll Cardiol. 2017; 70: 2476-2486.Crossref
- 9.
- Savarese G, Becher PM, Lund LH, et al. Global burden of heart failure: a comprehensive and updated review of epidemiology. Cardiovasc Res. 2023; 118: 3272-3287.Crossref
- 10.
- Barge-Caballero E, Montero-Fole B, Barge-Caballero G, et al. Clinical profile and causes of death according to ejection fraction in patients with heart failure cared for in a specialized cardiology unit. Rev Clin Esp (Barc). 2022; 222: 152-160.Crossref
- 11.
- Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail. 2016; 18: 891-975.Crossref
- 12.
- Chioncel O, Lainscak M, Seferovic PM, et al. Epidemiology and one-year outcomes in patients with chronic heart failure and preserved, mid-range and reduced ejection fraction: an analysis of the ESC Heart Failure Long-Term Registry. Eur J Heart Fail. 2017; 19: 1574-1585.Crossref
- 13.
- Kapłon-Cieślicka A, Benson L, Chioncel O, et al. A comprehensive characterization of acute heart failure with preserved versus mildly reduced versus reduced ejection fraction - insights from the ESC-HFA EORP Heart Failure Long-Term Registry. Eur J Heart Fail. 2022; 24: 335-350.Crossref
- 14.
- Cho JH, Choe WS, Cho HJ, et al. Comparison of characteristics and 3-year outcomes in patients with acute heart failure with preserved, mid-range, and reduced ejection fraction. Circ J. 2019; 83: 347-356.Crossref
- 15.
- Shiga T, Suzuki A, Haruta S, et al. Clinical characteristics of hospitalized heart failure patients with preserved, mid-range, and reduced ejection fractions in Japan. ESC Heart Fail. 2019; 6: 475-486.Crossref
- 16.
- Lauritsen J, Gustafsson F, Abdulla J. Characteristics and long-term prognosis of patients with heart failure and mid-range ejection fraction compared with reduced and preserved ejection fraction: a systematic review and meta-analysis. ESC Heart Fail. 2018; 5: 685-694.Crossref
- 17.
- Huusko J, Tuominen S, Studer R, et al. Recurrent hospitalizations are associated with increased mortality across the ejection fraction range in heart failure. ESC Heart Fail. 2020; 7: 2406-2417.Crossref
- 18.
- Bhambhani V, Kizer JR, Lima JAC, et al. Predictors and outcomes of heart failure with mid-range ejection fraction. Eur J Heart Fail. 2018; 20: 651-659.Crossref
- 19.
- Vergaro G, Ghionzoli N, Innocenti L, et al. Noncardiac versus cardiac mortality in heart failure with preserved, midrange, and reduced ejection fraction. J Am Heart Assoc. 2019; 8: e013441.Crossref
- 20.
- Fröhlich H, Rosenfeld N, Täger T, et al. Epidemiology and long-term outcome in outpatients with chronic heart failure in Northwestern Europe. Heart. 2019; 105: 1252-1259.Crossref
- 21.
- McDonagh TA, Metra M, Adamo M, et al. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021; 42: 3599-3726.
- 22.
- Rywik TM, Doryńska A, Wiśniewska A, et al. Epidemiology and clinical characteristics of hospitalized patients with heart failure with reduced, mildly reduced, and preserved ejection fraction. Pol Arch Intern Med. 2022; 132: 16227.Crossref
- 23.
- Sweitzer NK, Lopatin M, Yancy CW, et al. Comparison of clinical features and outcomes of patients hospitalized with heart failure and normal ejection fraction (> or =55%) versus those with mildly reduced (40% to 55%) and moderately to severely reduced (<40%) fractions. Am J Cardiol. 2008; 101: 1151-1156.Crossref
- 24.
- Zhou J, Cao J, Jin X, et al. Digoxin is associated with worse outcomes in patients with heart failure with reduced ejection fraction. ESC Heart Fail. 2020; 7: 138-146.Crossref
- 25.
- Messika-Zeitoun D, Verta P, Gregson J, et al. Impact of tricuspid regurgitation on survival in patients with heart failure: a large electronic health record patient-level database analysis. Eur J Heart Fail. 2020; 22: 1803-1813.Crossref
- 26.
- Benfari G, Antoine C, Miller WL, et al. Excess mortality associated with functional tricuspid regurgitation complicating heart failure with reduced ejection fraction. Circulation. 2019; 140: 196-206.Crossref
- 27.
- Sierpiński R, Sokolska JM, Suchocki T, et al. 10 Year trends in hospitalization rates due to heart failure and related in-hospital mortality in Poland (2010-2019). ESC Heart Fail. 2020; 7: 3365-3373.Crossref
- 28.
- Dunlay SM, Roger VL, Killian JM, et al. Advanced heart failure epidemiology and outcomes: a population-based study. JACC Heart Fail. 2021; 9: 722-732.Crossref
- 29.
- Savarese G, Vedin O, D’Amario D, et al. Prevalence and prognostic implications of longitudinal ejection fraction change in heart failure. JACC Heart Fail. 2019; 7: 306-317.Crossref
- 30.
- Correale M, Paolillo S, Mercurio V, et al. Non-cardiovascular comorbidities in heart failure patients and their impact on prognosis. Kardiol Pol. 2021; 79: 493-502.Crossref
- 31.
- Savarese G, Settergren C, Schrage B, et al. Comorbidities and cause-specific outcomes in heart failure across the ejection fraction spectrum: a blueprint for clinical trial design. Int J Cardiol. 2020; 313: 76-82.Crossref
- 32.
- Savarese G, Stolfo D, Sinagra G, et al. Heart failure with mid-range or mildly reduced ejection fraction. Nat Rev Cardiol. 2022; 19: 100-116.Crossref
- 33.
- Szczurek-Wasilewicz W, Gąsior M, Skrzypek M, et al. Predictors of 1-year mortality in ambulatory patients with advanced heart failure awaiting heart transplant. Pol Arch Intern Med. 2022; 132: 16151.Crossref
- 34.
- Vedin O, Lam CSP, Koh AS, et al. Significance of ischemic heart disease in patients with heart failure and preserved, midrange, and reduced ejection fraction: a nationwide cohort study. Circ Heart Fail. 2017; 10: e003875.Crossref
- 35.
- Pfeffer MA, Claggett B, Assmann SF, et al. Regional variation in patients and outcomes in the Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist (TOPCAT) trial. Circulation. 2015; 131: 34-42.Crossref
- 36.
- Machino-Ohtsuka T, Seo Y, Ishizu T, et al. Relationships between maintenance of sinus rhythm and clinical outcomes in patients with heart failure with preserved ejection fraction and atrial fibrillation. J Cardiol. 2019; 74: 235-244.Crossref