The patient data were obtained from the information collected in the RICA registry between March 2008 and September 2011. RICA is a prospective multicenter cohort registry coordinated by the SEMI Working Group on Heart Failure.13,14 Briefly, it includes patients hospitalized for an episode of decompensated HF, consecutively attended by the physicians participating in the study. In the event that a case was not entered in the registry, it was not because of any inherent characteristic of the patient. The data were filled in using a web site12 and anonymized to ensure patient confidentiality. Data were analyzed at the time of inclusion (hospital discharge) and at 3 and 12 months of follow-up. The registry includes data from 52 public health and private hospitals providing different levels of medical care. The study protocol was approved by the Ethics Committee of Hospital Universitario Reina Sofía de Córdoba, and informed consent was obtained from all patients before enrollment in the study.

In addition to providing informed consent, patients had to be older than 50 years and have been admitted for decompensated HF according to the criteria of the European Society of Cardiology15: a) presence of HF symptoms; b) objective evidence of cardiac dysfunction (systolic or diastolic), preferably by echocardiography, and c) in cases of an inconclusive diagnosis, an adequate clinical response to HF treatment.

Patients with HF secondary to pulmonary hypertension and those who refused to participate were not included in the study. In addition, we excluded patients with significant valvular dysfunction (moderate-severe or severe), and analyzed only those patients with at least 1 year of follow-up.

At the time patients were included, we recorded the following data: a) sociodemographic variables (age, sex); b) clinical history and characteristics (hypertension, diabetes mellitus, smoking, acute myocardial infarction, atrial fibrillation, hyperlipidemia, chronic obstructive pulmonary disease, neoplasm, peripheral arterial disease, stroke, Charlson comorbidity index,16 previous functional capacity evaluated with the Barthel index17 before the decompensation that required admission, blood pressure, heart rate, and New York Heart Association [NYHA] functional class); c) analytical data (renal function, hemoglobin and natremia); d) electrocardiographic and echocardiographic parameters (branch block, left ventricular hypertrophy, and left ventricular ejection fraction [LVEF], considered to be preserved at more than 50%), and e) treatment (BB, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, aldosterone antagonists, diuretic, digoxin, statin, dicoumarin, and acetylsalicylic acid).

During follow-up, information was obtained on overall mortality and hospital readmissions for HF. Patients were analyzed in 2 groups, according to whether or not they received BB.

Qualitative variables were analyzed using the chi-square test and quantitative variables with the Student t test. Survival analyses were performed with Kaplan-Meier curves. Comparisons of survival curves were carried out with the log rank test. Multivariate analysis using Cox proportional hazards regression models was carried out, including variables that were statistically significant in the univariate model and those considered to have prognostic value regardless of their statistical significance. The level of significance was established at P<.05. All statistical analyses were performed with IBM SPSS, version 20.0.

The flow chart of the patients evaluated is shown in Figure 1. In September of 2011, there were 1478 patients in the RICA registry, 790 of whom had reached 1 year of follow-up. From this total, 275 were excluded because of significant valvular disease, which left 515 patients for the analysis. Mean age was 77.1 (8.7) years, and 53.6% were women. At 3 and 12 months following discharge, 50 and 70 patients, respectively, had died, and information was missing in 17 (7 with BB) and 4 (2 with BB) cases, respectively.

Figure 1.

Flow chart of patients assessed during 1 year of follow-up. BB, beta-blockers.

(0.17MB).

The baseline clinical characteristics at the time of hospitalization and the treatment at discharge in the overall series and in the subgroups with and without BB treatment are shown in Table 1.

At the time of discharge, 62.1% of patients were prescribed BB. With regard to comorbid conditions, BB were used more frequently in those with myocardial infarction and less frequently in those with chronic obstructive pulmonary disease. In patients who received BB, functional class was better, and diastolic arterial pressure and heart rate were higher. As to concomitant treatment, diuretics, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, acetylsalicylic acid, and statins were prescribed more often in patients treated with BB than in those without.

Although the LVEF was significantly lower in the BB group, most of the patients included (56%) had HF with preserved ejection fraction. Only 15.7% of our patients had LVEF lower than 35%.

The BB prescription rates at 3 months (65.6%) and 1 year (67.9%) were slightly higher than at the time of discharge (62.1%), although the differences were not significant (P>.05).

With regard to the mortality rate, there were 120 deaths during the year of follow-up (Figure 1), the majority (58.5%) due to HF. The remaining deaths were caused by other cardiovascular conditions (19%) and by noncardiovascular causes (22.5%). As can be seen in Table 2 and Figure 2, both the total mortality (Figure 2A) and the composite outcome, total mortality plus readmittance for HF (Figure 2B), were significantly lower in patients treated with BB than in those without (P<.001 and P<.003, respectively). On comparative analysis of BB prescription at discharge and at 1-year follow-up, there were no significant differences in percentages of prescription between the years 2008, 2009, 2010, and 2011 (59.9%, 64.3%, 62.2%, and 60.0%, respectively; P>.05).

Figure 2.

A: Kaplan-Meier curve for all-cause mortality in patients who received beta-blockers and those who did not. B: Kaplan-Meier curve for all-cause mortality or readmission in patients who received beta-blockers and those who did not. BB, beta-blockers.

(0.14MB).

The results of multivariate analysis showed a significant association between BB administration and a reduction in total mortality (relative risk=0.59; 95% confidence interval, 0.41-0.84) and the composite variable, mortality and hospital readmittance (relative risk=0.64; 95% confidence interval, 0.49-0.83) (Tables 3 and 4). In addition to BB use, systolic arterial pressure, glomerular filtration rate, and the use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers were directly related to a reduction in mortality, whereas glomerular filtration and the use of spironolactone were associated with a reduction in the composite event.

In contrast, age, male sex, and functional class were predictive of both clinical events (Tables 3 and 4). We should mention that there was no association with other variables, such as LVEF, diabetes mellitus, acute myocardial infarction, or stroke.

Possibly the main strength of the RICA registry is that it is representative of the elderly HF population, who are usually seen in hospital internal medicine services: advanced age, marked comorbidity, and predominance of women with HF and preserved ejection fraction.28 However, our study also has some limitations. Because sequential analysis of the data from each patient in the various phases of follow-up was not performed, we cannot precisely know how many patients underwent a change in BB treatment (discontinuation or start after discharge). Nor was it the aim of the study to determine the reasons why BB treatment was not prescribed or was discontinued, or what doses were used. However, the clinical differences between patients treated or not with BB were those expected: higher acute myocardial infarction rate, functional class III-IV, lower rate of chronic obstructive pulmonary disease, and lower LVEF in treated patients. In any case, the results clearly demonstrate more frequent use of this treatment than in previous observational evaluations. There were relatively few cases with missing information during follow-up (21 cases), with no differences between those taking BB or not. Therefore, we do not believe that this factor would cause a relevant change in the data. Finally, the results regarding the influence of BB use on the prognosis should be confirmed in a randomized controlled study.