Abstract 16503: Longitudinal Changes in the Intermountain Risk Score--a Metric Composed of the Complete Blood Count and Basic Metabolic Profile--are Prognostic for Mortality
Background: The Intermountain Risk Score (IMRS) is a new metric created in a general medical population from the complete blood count (CBC) and basic metabolic profile (BMP) and validated in NHANES III. IMRS predicts mortality and conditions that lead to it, including heart failure (HF) and myocardial infarction (MI). This study tested the predictive ability of a 1-year change in IMRS.
Methods: Females (F: N=5,698) and males (M: N=5,437) with two CBC and BMP measurements (1999-2009) had baseline inpatient and follow-up IMRS computed (mean time to follow-up: 1.0 years, range: 0.5-2.0). Change in IMRS (ΔIMRS) was evaluated for the prediction of mortality (F: n=1,255 deaths; M: n=1,164 deaths) and incident MI (F: n=133 events; M: n=147 events), HF (F: n=168; M: n=196), and stroke (F: n=176; M: n=148). MI, HF, and stroke were evaluated only for patients with no history of those events. ΔIMRS was calculated as follow-up IMRS minus baseline IMRS. After the follow-up IMRS measurement, outcomes were determined over the succeeding 4.0±2.5 years (maximum: 10 years) and analyzed by Cox regression.
Results: ΔIMRS as a continuous variable predicted mortality (F: p=0.003; M: p<0.001) and for increased (ΔIMRS > 3) vs. decreased (ΔIMRS < -3) IMRS change (F: HR=1.35, 95% CI=1.11, 1.64, p=0.003; M: HR=1.64, 95% CI=1.34, 2.01, p<0.001). Also, baseline and follow-up IMRS were independently predictive in models entering both (F: all p<0.001; M: all p<0.001). ΔIMRS did not predict MI, HF, or stroke, and among males a trend to lower risk existed (p=0.020). In contrast, baseline IMRS predicted higher risk of MI (F: p<0.001; M: p=0.26), HF (F: p=0.004; M: p=0.009), and stroke (F: p<0.001; M: p=0.005), and follow-up IMRS predicted HF (F: p<0.001; M: p<0.001), and stroke (F: p=0.006; M: p=0.13), but not MI.
Conclusions: Change in IMRS over an inter-measurement average of one year was prognostic for mortality among initially hospitalized patients who were followed up to 10 years. Baseline and follow-up IMRS measurements also predicted mortality and incident MI, HF, and stroke events. This suggests that modification of CBC and BMP values during follow-up may reduce the risk of mortality, but increases and decreases in IMRS over time are less valuable for determining the incidence of non-fatal events.
- © 2010 by American Heart Association, Inc.