Predictors of Long-Term Mortality in Older Patients Hospitalized for Acutely Decompensated Heart Failure

Editor's Note: Commentary based on Passantino A, Guida P, Lagioia R, et al. Predictors of Long-Term Mortality in Older Patients Hospitalized for Acutely Decompensated Heart Failure: Clinical Relevance of Natriuretic Peptides. J Am Geriatr Soc 2017;65:822-6.

Rationale for Study/Background: Evaluate prognostic utility of NT-pro-BNP for short-term and 1-year mortality in elderly patients hospitalized with acute decompensated heart failure.

Funding: None.


Design: Retrospective observational study.

Inclusion Criteria: Patients aged 75 years or older admitted for decompensated heart failure in two Italian centers.

Exclusion Criteria: Patients without NT-pro-BNP values, or with acute coronary syndromes, isolated right heart failure, hypertrophic cardiomyopathy or recent surgical or percutaneous procedures.

Exposure: NT-pro-BNP level.

Primary Outcome(s): 2 month and 12 month all-cause mortality.

Secondary Outcomes: N/A.

Statistical Analysis: Association between NT-pro-BNP and mortality after adjustment for other factors. The optimal cut-off point of NT-pro-BNP to predict mortality was selected based on the cut-off that, when used to discriminate high-risk from low-risk patients, yielded the highest model chi-square value from univariate Cox analysis.

Results: This retrospective study evaluated 279 patients ≥75 years of age (mean age 80 years; average survival 2.2 years) admitted for acute decompensated heart failure. Prognostic factors, including NT-pro-BNP levels, associated with short (2 months) and long term (1 year) mortality were assessed. In-hospital and 1-year mortality were high (10% and 36%, respectively). NT-pro-BNP had a univariate C-index for predicting mortality at 2 and 12 months of 0.74 and 0.75, with optimal cutoffs at 2 months of 8444 pg/ml and 1 year of 8275 pg/ml. NT-pro-BNP was an independent predictor of mortality after adjustment for relevant covariates. Additional independent predictors included eGFR (estimated glomerular filtration rate), hemoglobin, diabetes, systolic blood pressure, and moderate to severe tricuspid regurgitation, with a final model C-index of 0.765. NT-pro-BNP was the most powerful predictor of short and long term mortality in this population.

Conclusion: Acute decompensated HF among older patients is associated with very high all-cause mortality during short-term and longer-term follow-up. NT-pro-BNP has a strong independent dose-response association with all-cause mortality among this population, with NT-pro-BNP levels >8000pg/ml being most reflective of high risk.

Limitations of Study: A retrospective study linking baseline evaluation with long-term mortality fails to consider interval events (e.g., myocardial infarction) which may be relevant. Information on intensification of treatment or HF symptoms over time is not available, and symptom improvement is an important outcome not explored in this analysis.

Geriatric Perspective for the Cardiovascular Clinician: Natriuretic peptides support the diagnosis and establish disease severity and prognosis in hospitalized patients with HF. However, natriuretic peptides may be elevated in the setting of co-morbidities such as valvular heart disease, hypertension, pulmonary hypertension, pulmonary embolism, sepsis, renal failure, and ischemic heart disease - more common in older populations. Natriuretic peptides may also be elevated among older patients in the absence of HF, presumably from subclinical changes in heart structure and function and/or decline in renal clearance. This study illustrates that BNP remains a significant predictor of mortality in older adults. It also identifies a higher cut-off value predictive of long-term outcomes in the hospitalized older population with acute heart failure.

This study focused on mortality, not symptomatic improvements or re-hospitalization, but prognostic stratification based on natriuretic peptides assesses a spectrum of risk; ideally, those at the highest risk should be treated more intensely with closer follow-up. In addition, identification of the high risk patient may serve as the springboard for discussing prognosis and advance care planning.

Clinical Topics: Acute Coronary Syndromes, Anticoagulation Management, Geriatric Cardiology, Heart Failure and Cardiomyopathies, Prevention, Pulmonary Hypertension and Venous Thromboembolism, Vascular Medicine, ACS and Cardiac Biomarkers, Anticoagulation Management and ACS, Acute Heart Failure, Heart Failure and Cardiac Biomarkers, Pulmonary Hypertension, Hypertension

Keywords: Aged, 80 and over, Blood Pressure, Acute Coronary Syndrome, Glomerular Filtration Rate, Tricuspid Valve Insufficiency, Natriuretic Peptide, Brain, Heart Failure, Peptide Fragments, Myocardial Infarction, Hypertension, Renal Insufficiency, Diabetes Mellitus, Hypertension, Pulmonary, Cardiomyopathy, Hypertrophic, Sepsis, Hemoglobins, Pulmonary Embolism, Geriatrics

< Back to Listings