Predictors of mortality and morbidity in patients with chronic heart failure

Area of Science:

• Cardiology
• Clinical Research
• Predictive Modeling

Background:

• Chronic heart failure (CHF) affects millions globally, necessitating accurate risk stratification.
• Existing prognostic models may not fully capture risk across the spectrum of left ventricular systolic function.

Purpose of the Study:

• To develop and validate robust prognostic models for predicting mortality and cardiovascular events in a large chronic heart failure population.
• To identify key clinical predictors of adverse outcomes in patients with CHF, irrespective of ejection fraction.

Main Methods:

• Utilized data from 7599 patients in the CHARM program with varying degrees of left ventricular systolic function.
• Employed multi-variable Cox regression to build models predicting all-cause mortality and the composite of cardiovascular death/heart failure hospitalization.
• Identified 21 independent predictor variables for the final prognostic models.

Main Results:

• Older age (>60 years), diabetes, and lower left ventricular ejection fraction (<45%) were the strongest predictors of adverse outcomes.
• Other significant risk factors included higher NYHA class, cardiomegaly, prior heart failure hospitalization, male sex, lower BMI, and lower diastolic blood pressure.
• The developed models accurately stratified 2-year mortality risk across deciles, from 2.5% to 44%.

Conclusions:

• Routine clinical variables effectively stratify risk in chronic heart failure patients, regardless of ejection fraction.
• Diabetes emerged as a potent independent predictor of adverse outcomes in this CHF cohort.
• These validated models provide valuable tools for risk assessment and understanding the clinical course of chronic heart failure.