Abstract 14909: The Addition of B-Type Natriuretic Peptide to the Intermountain Risk Score Improves the Predictive Ability for Mortality Among Patients Undergoing Angiography
Background: The Intermountain risk score (IMRS) is an exceptional predictor of mortality that was created in a general medical population and validated in NHANESIII and CV patients. B-type natriuretic peptide (BNP) is a marker of left ventricular dysfunction. This study evaluated whether the addition of BNP to IMRS improves risk stratification for death, and also tested the newly derived coefficients for association with heart failure (HF) incidence among HF-free patients undergoing angiography.
Methods: IMRS was calculated among 4811 males and 3035 females of the Intermountain Heart Collaborative Study from sex-specific weights of factors from the complete blood count, basic metabolic profile, and age. To be included, patients had to have complete IMRS information and a BNP measurement obtained within 3 months of angiography. A BNP-modified IMRS was newly computed using the adjusted regression weights from multivariable models for mortality. The c-statistic evaluated model predictive ability.
Results: Age averaged 66.2±14.4 yrs for females and 64.7±13.1 years for males. C-statistics for 30 days and 1 year death (Table) showed marked improvement in both females and males when BNP was added to IMRS, although the statistical significance was borderline. Quintile 5 vs. 1 at 30 days and 1 year for women were 5.66, p=0.13 and 5.68, p<0.0001 and for men 7.10, p=0.03 and 6.01, p<0.0001, respectively. When applied to those without HF (females: n=2244, males: n=3545), the BNP-modified IMRS substantially improved the prediction of 1 year HF incidence, although 30-day results were less affected (Table).
Conclusions: The addition of BNP to IMRS improved risk prediction for short and intermediate-term mortality. It also was beneficial for HF risk prediction at 1 year. These results reinforce BNP as an independent predictor of risk among cardiovascular patients and suggest that it provides useful additional information when considered with standard clinical laboratory tests.
- © 2013 by American Heart Association, Inc.