Risk scores versus natriuretic peptides for identifying prevalent stage B heart failure.

ABSTRACT

BACKGROUND: Identifying asymptomatic individuals with American Heart Association/American College of Cardiology stage B heart failure (HF) in the population is an important step to prevent the development of symptomatic HF. The comparative utility of 2 screening strategies (biomarkers vs risk scores) in identifying prevalent stage B HF is unknown. METHODS: Participants 30 to 65 years old without symptomatic HF in the Dallas Heart Study who had a cardiac magnetic resonance imaging were included (n = 2,277). Stage B HF (n = 284) was defined by left ventricular (LV) hypertrophy, reduced LV ejection fraction, or prior myocardial infarction. We compared the utility of 2 risk scores (Health Aging and Body Composition HF risk score and the Framingham Heart Failure risk score) with B-type natriuretic peptide (BNP) and N-terminal pro-BNP in identifying stage B HF using logistic regression. RESULTS: Depending upon the method of indexing LV mass (body surface area, fat-free mass, or height(2.7)), the c-statistic for the Health Aging and Body Composition HF risk score (0.73, 0.75, and 0.64, respectively) was greater than that for BNP (0.62, 0.70, and 0.57, respectively) and N-terminal pro-BNP (0.62, 0.69, and 0.56, respectively) (P < .01 for all). These findings were similar for the Framingham Heart Failure risk score except when LV mass was indexed to fat-free mass. Addition of natriuretic peptide levels to the risk scores resulted in a modest but significant improvement in discrimination of stage B HF (Δ c-statistic, 0.01-0.03, P < .05 for all). CONCLUSIONS: Screening for stage B HF in the population is enhanced when natriuretic peptides are measured in addition to, rather than in place of, traditional risk scores.