[Predictive variables for mortality in elderly patients hospitalized due to heart failure with preserved ejection fraction].

Background And Objectives: The prevalence of heart failure (HF) increases with age. Even though the mortality of patients ≥ 80 years of age with HF and preserved left ventricle ejection fraction (LVEF) is very high, the predictor variables are not well-known. The main goal of this study was to evaluate the mortality predictor factors in this subgroup of the elderly population.

Patients And Methods: An observational and prospective study of patients hospitalized due to HF with preserved LVEF has been conducted. The demographic, clinical, functional and analytic factors were evaluated when the patients were admitted with special attention to the co-morbidities. The primary endpoint was the total mortality in the subgroup of patients ≥ 80 years of age after a year of follow-up. The predictor variables were studied by means of a multivariate Cox regression model.

Results: From a total of 218 patients with an average age of 75.6 (±8.7) years of age, 75 patients (34.4%) were ≥ 80 years. The mortality rate of patients ≥ 80 years of age totaled 42.7%, in relation to 26.6% for the lower age group (log-rank<.001). After a multivariate analysis using the Cox regression model in patients ≥ 80, the serum urea levels above the average (hazard ratio [HR] 3.93; 95% confidence interval [95% CI] 1.58-9.75; P = .003), the age (HR 1.17; 95% CI 1.07-1.28; P<.001), the hyponatremia (HR 3.19; 95% CI 1.51-6.74; P = .002) and a lower score on the Barthel index (BI) (HR 1.016; 95% CI 1.002-1.031; P = .034) were independent mortality predictors after an one-year follow-up.

Conclusions: Serum urea levels, age, hyponatremia and a low BI score could be proposed as independent mortality predictors in patients ≥ 80 of age hospitalized for HF with preserved LVEF.