Heart failure (HF) is a complex clinical syndrome resulting from alterations to cardiac structure or function. Symptom progression in HF results in frequent hospitalizations, poor quality of life for patients, and a high mortality rate.1–3

HF can be classified according to left ventricular ejection fraction (LVEF) into three basic phenotypes: a) HF with reduced LVEF (HFrEF, ≤ 40%); b) HF with preserved LVEF (HFpEF, ≥ 50%); and c) HF with mid-range LVEF (HFmrEF, 41-49%).3,4 HFrEF accounts for between 40% and 50% of HF patients.2,4,5 Classification of HF patients according to LVEF is important because LVEF phenotype is linked to differences in etiology, comorbidities, and especially responses to treatment.1–3

Several studies have shown that the prognosis of patients with HFrEF and—to a lesser degree—those with HFmrEF can be improved by neurohormonal modulation with angiotensin-converting enzyme inhibitors (ACEI) or angiotensin II receptor blockers (ARB) in combination with beta-blockers and aldosterone antagonists (AA); moreover, prognosis is also improved by treatments that potentiate the natriuretic peptide system and other endogenous vasodilatory systems.6,7

HF prevalence in developed countries is around 2% overall and exceeds 10% in people older than 70 years.8 In Spain, overall prevalence has been estimated at >5%, perhaps reflecting a lack of national population-based studies that would provide more precise estimates of the impact of this disease.5,9 Furthermore, current knowledge of the medical treatment received by patients is derived from registries compiled by specialist HF units and may not accurately represent the population at large.10 To allow appropriate planning of present and future interventions, there is therefore an urgent need for up-to-date population data that accurately describe HF prevalence and the use of prognosis-modifying treatments. In the present study, we report the epidemiology (prevalence and incidence), baseline characteristics, and treatment of the HF population, especially HFrEF patients, in standard clinical practice in Spain during the period 2017-2019. The study is called PATHWAYS-HF to reflect the goal to provide information about the current state of HF and treatment trajectories in these patients.

To the best of our knowledge, this study is the largest population-based study of HF performed in Spain to date. The cohort is large, contemporary, and representative of the Spanish population, thus providing a good estimate of the real-world situation. The main study findings are that HF prevalence in Spain in 2019 was 1.89% of the population older than 18 years, with an incidence of 2.78/1000 persons/y, and that both figures were stable over the preceding 3 years. The results also clearly reflect the association of HF with aging, with the prevalence reaching 9% in the population older than 80 years. The most frequent HF type was HFrEF (51.7%). In contrast with the conclusions of other national and international studies, the present results show that there is wide room for improvement in the treatment optimization of HFrEF patients in accordance with current guidelines.1,3

During the period analyzed, the epidemiological data remained stable, showing a slight upward trend that did not reach statistical significance. The incidence rate remained steady at around 2.8 cases/1000 persons/y. The few previous epidemiological studies of HF in Spain have reported widely differing prevalence values depending on the geographical area and study population examined. For example, the PRICE study reported an HF prevalence of 6.8% in a randomized sample of patients aged ≥ 45 years from 15 hospitals and 55 health care centers in 9 autonomous communities.13 The EPISERVE study reported prevalence values of 2% in primary care, 17% in cardiology, and 12% in internal medicine.14 In a small sample (n=391) in Asturias, Cortina et al. reported a prevalence of 4.9%.15 These values contrast with the considerably lower prevalence detected in population-based studies. For example, Farre et al. reported prevalence values in Catalonia of 1.2% for the population older than 15 years and 2.7% among people older than 44 years.16 These findings are in line with those reported by Carmona et al. in the Community of Madrid.17 The prevalence values reported in non–population-based studies in Spain are thus much higher than the true values; in contrast, other population-based studies have published prevalence values similar to those reported here.1–4,18,19

In the present study, 51.7% of HF patients had HFrEF and 40.2% had HFpEF. These figures are somewhat higher than those published in previous reports.4,5,10 However, the rates for the most prominent comorbidities in our study (hypertension, dyslipidemia, atrial fibrillation, diabetes, and kidney failure) are comparable to those reported in other studies.4,15,16 Conde et al.5 reported that HF-associated comorbidities are a frequent cause of hospital readmissions, and the same study also showed that hospitalized HF patients tend to be elderly and have multiple comorbidities that hinder diagnosis and treatment, interfere with patient progress, and are associated with a worse prognosis. The published literature reveals some variability in comorbidity rates in HF2,5,13,15,16,20; however, the reported comorbidity rate is high in all studies, and the proportion of HF patients with HFrEF is consistently reported at between 40% and 60%.1,3,7

The analysis of medical treatment for HFrEF in the most recent study year (2019) reveals that 35.7% of patients were not prescribed beta-blockers, 19.5% were not prescribed ACEI/ARBs or sacubitril-valsartan, and 70.2% were not prescribed AAs. Current European Society of Cardiology treatment guidelines for HFrEF recommend an ACEI or ARB, a beta-blocker, and an AA (indication class I, level of evidence A).3 The high nonprescription rates found here thus demonstrate that treatment optimization should be prioritized as an essential step toward reducing symptoms, hospitalization, and mortality and increasing functional capacity and quality of life.1–4,9 Growing evidence for the benefits of sacubitril-valsartan recently led to this treatment being recommended as a replacement for ACEI/ARB therapy.21 However, examination of published studies shows that this recommendation is not being followed in clinical practice.22–25 Our data differ from those for the 2834 Spanish patients included in the European Heart Failure registry, which reveal the use of ACEI/ARBs, beta-blockers, and AAs by 92%, 93.3%, and 74.5% of patients, respectively.10 However, while these are real-world data, they reflect the specific experience of HF units; large population-based cohorts not linked to specialized HF units have produced data very similar to ours. For example, results from the recent American CHAMP-HF registry show that, in the absence of contraindications, 27% of HFrEF patients were not prescribed ACEI/ARBs or sacubitril-valsartan, 33% were not prescribed beta-blockers, and 67% were not prescribed AAs.26 There is thus clear room for improvement in treatment optimization for HFrEF.

In our sample of 1550 patients diagnosed with HFrEF in 2017, treatment optimization was concentrated in the first 3 to 6 months, and treatments tended to not to change in the 12–24-month period. In our view, this treatment trajectory indicates an important element of therapeutic inertia and therefore suggests that the treatment of these patients could be improved. However, the slowdown in treatment optimization may in part be due to specific factors or patient characteristics.10 It will be important to establish training programs and simple clinical procedures to improve the alignment of clinical practice with guideline recommendations,1,2 together with strategies to improve treatment adherence27 through coordination between different specialties, especially during transition from one area of care to another.

The cause of HF was ischemic (coronary) in 40% of the patients in the study population. Published rates of ischemic HF vary, but the proportion of ischemic HF in our study is below the > 60% reported in some reviewed series.28 This discrepancy could be due to 2 factors. a) In many patients, the presence of 2 or more overlapping causal factors makes it difficult to link HF to a single, specific etiology. For example, coronary atherosclerosis is frequently detected but is not always sufficiently severe to explain a patient's HF. Likewise, in other patients, hypertension can be considered both a risk factor for ischemic heart disease and a direct, intrinsic cause of HF. b) The retrospective nature of our study may have resulted in an under-recording of ischemic etiology. Nevertheless, our results are in line with other reported findings.4,25 Our analysis is at risk of the usual limitations associated with retrospective database reviews. These include disease under-reporting, possible variability among health care professionals and patients, the potential for classification bias in an observational study, and imprecise diagnosis coding. Another potential limitation is missing data, either due to nonsystematic determination of prognostic biomarkers (such as natriuretic peptides) or of variables such as LVEF or NYHA class. There is also a risk that the proportion of patients in treatment may be underestimated due to missing information related to dose changes, treatment adherence, medication prescribed by private health care providers, or paper prescriptions. It would also have been interesting to have information about how the different drug treatments were distributed according to the 3 LVEF phenotypes associated with HF. However, this was not considered as a study objective because HFrEF is the only phenotype whose prognosis can be modified with available medication. Finally, because our analysis did not consider the cause of death of patients who died at home, death due to a cardiovascular cause may have been underestimated.

During the period analyzed, the epidemiological data for HF remained stable, with a prevalence close to 2%. This figure is lower than that reported in some non–population-based studies but is in line with the reported prevalence in comparable countries. A high proportion of HF patients have multiple comorbidities, and there is wide room for improvement in the optimization of medical treatment of HFrEF in line with scientific society recommendations.

This study was funded by AstraZeneca, Spain.

All authors contributed to the study conception and design. A. Sicras-Navarro collected data and performed the statistical analysis. All authors interpreted data, wrote and revised the text, and approved the submitted manuscript.

A. Sicras-Mainar and A. Sicras-Navarro were independent consultants in relation to the preparation of this manuscript. B. Palacios and L. Varela are on the staff of AstraZeneca.