Keane K. Lee, MD; Mark A. Hlatky, MD
In healthy men, does the Reynolds Risk Score predict cardiovascular events more accurately than a score based on traditional risk factors only?
Prospective cohort study to validate a risk score in men.
10 724 men 50 to 79 years of age (median age 63 y) from the Physicians Health Study II who were free of cardiovascular disease, diabetes, and cancer.
Reynolds Risk Score, which in addition to traditional cardiovascular risk factors (age, blood pressure, total and high-density lipoprotein cholesterol levels, and smoking), includes high-sensitivity C-reactive protein (hs-CRP) levels and parental history of myocardial infarction (MI) at < 60 years of age, provides an estimate of 10-year cardiovascular risk. Reynolds Risk Score was evaluated for global fit (Bayes information criterion [BIC], lower values indicate better model fit), discrimination (C-index, larger values indicate better discrimination), calibration (chi-square values < 20 suggest that predicted probabilities adequately reflect actual risk), and reclassification improvement, compared with a model based on traditional risk factors only.
Cardiovascular events (MI, stroke, coronary revascularization, or cardiovascular death).
1294 first cardiovascular events occurred over a median follow-up of 11 years. In analyses that controlled for traditional risk factors, the hazard ratios were 1.2 (95% CI 1.1 to 1.4) for hs-CRP levels 1 to 3 mg/L and 1.4 (CI 1.2 to 1.7) for hs-CRP levels > 3 mg/L (compared with hs-CRP levels < 1 mg/L), and 1.7 (CI 1.5 to 2.0) for presence of parental history of MI at < 60 years. Reynolds Risk Score had better global fit (BIC 23 008 vs 23 048, P < 0.001) and discrimination (C-index 0.71 vs 0.70, P < 0.001), but similar calibration (chi-square 13 vs 11) compared with a model based on traditional risk factors. Reynolds Risk Score reclassified 18% of men into higher or lower risk categories with 84% accuracy (20% of those in the 2 middle categories with 100% accuracy) (Table). The net reclassification improvement was 5.3% overall and 14% for men in the 2 middle categories.
Adding high-sensitivity C-reactive protein levels and parental history of coronary disease to traditional risk factors improved cardiovascular risk prediction in healthy men.
Reynolds Risk Score vs traditional risk score to predict 10-year risk for cardiovascular events in healthy men
*Data are numbers of men in each category (actual 10-y event risks). Cells in italics represent reclassification from the traditional risk score category.
Keane K. Lee, Mark A. Hlatky. Reynolds Risk Score for men predicted cardiovascular events more accurately than traditional risk factors only. Ann Intern Med. 2009;150:JC3–14. doi: 10.7326/0003-4819-150-6-200903170-02014
Download citation file:
Published: Ann Intern Med. 2009;150(6):JC3-14.
Cardiology, Coronary Risk Factors, Prevention/Screening.
Results provided by:
Copyright © 2017 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use