Heart failure (HF) is a global public health problem associated with high morbidity and mortality, and significant resource utilizatiaon.1 The incidence of HF increases with age,2 it affects up to 20% of the population aged at least 75 years,3 and it is the leading cause of hospitalization in this age group in developed countries.4,5 Prognosis has improved in recent years due to a better understanding of the pathophysiology of the disease and an improved therapeutic approach.6 Current recommendations emphasize the relevance of providing individualized strategies and treatments and taking into account the clinical situation and comorbidities of each patient.7 However, older patients are usually underrepresented or even excluded from large clinical trials.3 In addition, conditions that are prevalent in older patients with HF, such as frailty, are not assessed in routine clinical practice, despite their marked prognostic impact and even though their identification would allow individualized plans to be tailored to each patient.8-10 In our setting, the prevalence of frailty has been studied in patients admitted for acute HF who, in general, have been attended in geriatrics or internal medicine services. Our study is novel, in that it assessed the presence and prognosis of frailty in older patients with ambulatory HF, who had been exclusively followed up by cardiology services in Spain. Our aim was to assess the impact of frailty on 1-year all-cause mortality in these patients. We also studied its influence on hospital admission rates during follow-up.

A total of 499 patients were included. Median age was 81.4±4.3 years and 38% were women. Prior to inclusion, the median time since diagnosis of HF was 25 [12-60] months. Most patients (85%) were in NYHA class II (only 0.2% were in NYHA class III-IV). Mean LVEF was 43% and just over half of the patients (54%) had an LVEF of less than 40% (in the of the supplementary data shows the subanalysis of the baseline characteristics of the patients included in the registry by LVEF category). The main cause of HF was ischemic heart disease (48%). The FRAIL scale identified 22% of patients as frail and 49% of patients as prefrail. Table 1 shows the main characteristics of the patients by frailty. Frail and prefrail patients were more frequently women and had a higher mean age; the frailer patients had significantly more comorbidity and geriatric syndromes. Frail patients had significantly worse functional status, more congestion on physical examination, and higher LVEF and N-terminal pro-B-type brain natriuretic peptide (NT-proBNP) values. Frail patients also underwent more admissions for HF in the previous year.

During a median follow-up of 371 [361-387] days, 58 patients (11.6%) died, with noncardiovascular mortality being the leading cause of death (34 patients, 58%). Heart failure was the main cause of cardiovascular mortality (13 patients, 22%). Table 2 shows the association between geriatric syndromes and mortality at 1 year of follow-up. Associations were found between frailty, nutritional risk, comorbidity, and cognitive impairment, and mortality.

Table 3 shows the variables associated with mortality at 1-year follow-up in the univariate analysis (variables with P <.01 are shown). Figure 1 and figure 2 show the incidences of mortality and mortality or hospital readmission in patients by frailty, respectively. Frail patients had higher mortality during follow-up and a higher rate of the combined endpoint of death or hospital admission. In the multivariate analysis, one of the independent predictors of mortality was frailty identified using the visual mobility scale (hazard ratio [HR],=2.26; 95% confidence interval [95%CI], 1.16-4.38; P=.015), whereas the association between frailty using the FRAIL scale and mortality was marginal (HR,=2.35; 95%CI, 0.96-5.71; P=.059) (table 4). After adjustment for confounding variables, the visual mobility scale remained an independent predictor of mortality at 1 year of follow-up (HR,=2.13, 95%CI, 1.08-4.20; P=.02) (table 5).

Figure 1.

All-cause mortality according to the degree of frailty of the patients included in the FRAGIC registry.

(0.12MB).

Figure 2.

All-cause mortality or hospital admission according to degree of frailty.

(0.13MB).

In total, 164 patients (32.8%) were admitted during follow-up. The variables associated with all-cause admission in the univariate and multivariate analyses are shown in tables 2 to 5 in the supplementary data. The variables independently associated with hospitalization were crackles, the daily furosemide dose, and frailty (in this case identified using the Clinical Frailty scale).

The main findings of this study are as follows: a) there was a high prevalence of frailty in the population of ambulatory patients aged at least 75 years with HF followed-up by cardiology specialists in Spain; and b) frailty was independently associated with total mortality and all-cause hospitalization at 1 year of follow-up.

Western societies are currently aging and this trend is expected to continue in the coming years. In these countries, there is an increasing prevalence of geriatric syndromes and, in particular, of frailty and thus their assessment is recommended given their prognostic impact in different clinical situations.8,24,25 Frailty is defined as an aging-associated decrease in physiological reserve, which leads individuals to experience increased vulnerability to stressful situations.8 It is essential that clinicians have available reliable and easy-to-use measures to identify frailty, mainly to plan intervention strategies and adapt care to the characteristics of each patient. Frailty is not easy to assess, because there are 2 different approaches. One approach considers frailty to be a phenotype of poor physical functioning and therefore uses objective physical variables to define it, such as gait speed and grip strength. This approach is represented by use of the Fried19 scale and the Short Physical Performance Battery (SPPB).17 The other approach considers frailty to be a multidimensional entity resulting from the accumulation of deficits; under this definition, the most commonly used measures are the Clinical Frailty scale20 and the FRAIL scale.18 The FRAIL scale, due to its simplicity, is easy to use in daily clinical practice.

In Spain, and in the setting of HF, frailty and disability have mainly been studied in patients attended in geriatrics or internal medicine services during hospitalization for acute HF. In the RICA registry,26 Chivite et al. analyzed 2195 patients of at least 75 years of age admitted for acute HF to the internal medicine services of several Spanish hospitals. Most patients (57%) had moderate or severe disability regarding their everyday activities as assessed using the Barthel index, and an association was found between disability and increased all-cause mortality at 1 year of follow-up. Their study did not assess frailty or other geriatric syndromes. Rodríguez-Pascual et al.27 used Fried's definition of frailty to study the prognostic impact of frailty in 497 patients of at least 75 years of age after admission for acute HF and who had been followed up by geriatricians. In total, 57% of the patients were frail and had significantly higher mortality, readmissions, or functional deterioration at 1 year of follow-up. Vidán et al.28 studied 416 patients of at least 70 years of age, who had also been hospitalized for HF and attended in geriatrics, internal medicine, and cardiology services. Using Fried's scale, they identified 76% of the patients as frail (without distinguishing between robust and prefrail patients) and having higher mortality and readmission rates.

On the other hand, very few studies have assessed the prevalence and prognosis of frailty in patients with ambulatory HF. Cacciatore et al.29 studied the association of frailty in 120 ambulatory HF patients (mean age, 74±6.3 years) followed up by the geriatrics department of a hospital. Frailty, as defined by the Frailty Staging System (FSS), was associated with higher mortality at follow-up. In another single-center study, which was conducted in Spain, Lupon et al.30 found higher mortality in patients with geriatric syndromes who had been followed up in an outpatient HF unit. The main limitation of these studies, apart from being single-center studies, is the definition of frailty as a combination of items that are different from each other, sometimes poorly related, and which are closer to the definition of disability than to that of frailty.31

Our study is novel because, unlike previous studies, the FRAGIC registry is the first prospective multicenter study conducted in Spain in which all the patients included were under outpatient follow-up (chronic HF without recent admissions) by cardiology services. This aspect is relevant, given that the assessment of frailty and other geriatric syndromes varies according to the patients’ clinical situation, whether acute or chronic, and that the loss of function on hospital admission is widely documented. Thus, the inclusion of outpatients provides a real-life perspective of the study population, as well as its being the recommended setting for assessing frailty.8 The baseline characteristics of the our study population did not substantially differ from those of previous studies conducted in the setting of geriatrics or internal medicine,26–28 although the percentages of women and patients with HF of nonischemic etiology were lower in our study. Furthermore, in our study, there was lower mortality during follow-up (the main characteristics of the previous studies vs those of our study are shown in table 6 of the supplementary data). Consequently, the characteristics and complexity of the elderly patients with HF currently followed up by cardiology services were similar to those of patients followed up by other specialists. This result is in contrast to those of previous studies,32–34 in which associations were found between better prognosis and better baseline characteristics in patients followed up by cardiology services and possibly between better prognosis and the higher prescription of optimal medical treatment or advanced treatments.35 These results suggest the need for further studies in this regard.

We found an independent association between low hemoglobin levels and higher mortality. This result is in line with those of previous studies, which found associations between greater morbidity and mortality and worse quality of life.36 Previous studies have found that low lymphocyte counts and higher NT-proBNP also identified patients with worse prognosis,6,37–40 although their association with mortality was at the limit of statistical significance when different confounding variables were included in the model (table 5). Frailty, when assessed using the visual mobility scale, was an independent predictor of mortality before and after model adjustment, even though we included many other predictors or confounding factors and despite the sample size. This scale was demonstrated to be a useful and simple measure to identify patients with higher mortality during follow-up. Taken together, hemoglobin, lymphocyte, and NT-proBNP levels, as well as the degree of frailty estimated using the visual mobility scale, can be readily obtained in consulting rooms. These factors can help identify patients with a worse prognosis and who would therefore benefit from a specific approach and close follow-up (figure 3). The relevance of measuring frailty lies in its potential reversibility: it has been shown that frail patients have received clear benefits by their inclusion in multimodal rehabilitation programs.41 Given the observational nature of our study, no specific interventions for frailty were applied despite its identification. It is striking that less than a third of the patients were robust, yet only 1.8% were included in cardiac rehabilitation programs, which highlights current challenges.42-44

Figure 3.

Central illustration. Main prognostic variables in elderly outpatients with heart failure attended in cardiology services. COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; VMS, Visual Mobility Scale; HT, hypertension.

(0.26MB).

Frailty is highly prevalent in older patients with ambulatory HF followed up by cardiology services in Spain, and is independently associated with mortality at 1 year of follow-up. Given its potential reversibility, it is vital to identify frailty. Frailty in these patients should be approached and followed up with particular care. New studies are needed on the prevention and treatment of frailty in these patients.

None declared.

Study design: P. Díez-Villanueva, C. Bonanad, F. Formiga, F. J. Martín-Sánchez, M. T. Vidán, A. Ariza, M. Martínez-Sellés. Data collection: C. Jiménez-Méndez, P. Díez-Villanueva, C. Bonanad, C. Ortiz-Cortés, E. Barge-Caballero, J. Goirigolzarri, A. Esteban-Fernández, Á. Pérez-Rivera, M. Cobo, J. López, C. Guerrero, H. García Pardo, C. Robles, D. Iglesias, J. M. García Pinilla, L. López Rodríguez, A. Ariza, M. Martínez-Sellés. Data review and statistical analysis: P. Díez-Villanueva and A. Sanz-García. Manuscript preparation: C. Jiménez-Méndez, P. Díez-Villanueva, C. Bonanad, A. Sanz-García, F. Formiga, F. J. Martín-Sánchez, M. T. Vidán, A. Ariza, M. Martínez-Sellés and F. Alfonso. Review, editing, and approval of the manuscript: all authors.

None declared.