Determinants of long-term survival in patients hospitalized for heart failure.

BACKGROUND AND OBJECTIVES: Current models used to stratify patients with heart failure are complex, difficult to use, and limited by short follow-up and patient selection. Our aim was to determine predictors of long-term survival in patients hospitalized for heart failure and to develop a simple mortality risk score to estimate long-term mortality.

METHODS: We prospectively followed up, during 10 years after hospitalization, 701 patients with confirmed heart failure from the HOLA (Heart failure: Observation of Local Admissions) registry.

RESULTS: Mean age was 72.4 +/- 11.7 years; 45% were men. During follow-up, 465 patients died and 5 underwent heart transplantation. A total of 231 patients (33%) were alive and transplant-free at the end of follow-up (5.2 +/- 4.2 years). Median survival was 3.2 years. Multivariate analysis showed that six variables (age, previous renal disease, previous stroke, chronic obstructive pulmonary disease, left ventricular ejection fraction and aortic stenosis) were independent predictors of shorter survival time. By dichotomizing these variables, we obtained six factors with similar predictive values (hazard ratio between 1.5 and 2.0). A risk score for mortality was developed using these predictors by assigning 1 point to each and adding the total for each patient. Median survival for patients with 0, 1, 2, and 3 or more points were 6.5, 5.5, 3.3, and 1.7 years, respectively. One-year mortality rates were 15, 20, 28, and 49%, respectively.

CONCLUSION: The prognosis of patients hospitalized with heart failure is highly variable. A simple risk score, based on six variables readily obtainable on admission can effectively stratify patients according to their predicted mortality.