057 A simple prediction score for significant renal artery stenosis in patients with coronary artery disease

Abstract

Renal artery stenosis (RAS) is a strong independent predictor of mortality in patients (pts) with coronary artery disease (CAD). To develop and validate a score predicting RAS in patients with CAD. Three hundred consecutive pts (50 females) with significant CAD underwent abdominal aortography following coronary angiography to screen for significant RAS defined as luminal narrowing of > 50%. Univariate and multivariate analyses were performed comparing pts with and without RAS. Significant factors associated with RAS were included in constructing a score that predicts RAS. The score was internally validated in pts randomly selected from the entire study group (validation group; n = 103), using ROC curves and the Hosmer-Lemeshow goodness-of-fit test. Twenty-seven pts (9%) had a significant RAS. Univariate predictors of significant RAS were: age > 65 years (OR = 4.5, p < 0.0001), hypertension (OR = 3.6, p = 0.001), and female gender (OR = 3.6, p = 0.015). We found a tendency of more prevalent renal insufficiency (37.1% vs. 21.5%; p = 0.05) and the presence of 2 or more significant CAD lesions (70.4% vs. 50.9%; p = 0.05) in pts with RAS. Multivariate analysis showed that age > 65 years (OR = 4.1%, 95% CI = 1.6-10.3, p = 0.003) and hypertension (OR = 3.1, 95% CI = 1.2-7.7, p = 0.015) were independent predictors of RAS. The ranged from 0 to 7: 2 points for age > 65 years and hypertension 1 point for female gender, renal insufficiency, and > 3-vessel disease). Internal validation showed a good performance (ROC curve = 0.79 and Chi2 Lemeshow = 3.45). For a score < 2, the negative predictive value is 98%. Applying this criteria, 48.3% of our population would not require systematic abdominal angiography. The performance of our predictive score was good, and significant reduction in the need to perform systematic abdominal aortography could be expected with the use of this score.