A clinical score to predict elevated arterial stiffness: Derivation and validation in 3,943 hypertensive patients

Abstract

Purpose/Background/ObjectivesAortic stiffness assessed by carotid-femoral pulse wave velocity (PWV) is an important predictor to gauge the overall risk of hypertensive patients; nonetheless, it is underutilized in everyday practice. We derived a simple scoring system based on clinical variables that can identify patients with a priority for measurement of PWV, i.e. those with elevated PWV (≥ 10 m/sec) and at higher risk for events.MethodsPatient data from three outpatient clinics (n = 3,943) were used to form a derivation, internal and external validation cohort. For derivation, independent predictors of high PWV from a binary logistic regression model were dichotomized and implemented in a clinical prediction scoring system with the acronym SAGE (office systolic blood pressure >160 mmHg: 4 points, age ≥ 60 years: 4 points, glycemia [blood glucose ≥126 mg/dl]: 1 point, eGFR ≤60: 2 points).ResultsIts performance was validated at the internal and external validation cohorts with c-statistics being 0.781 (95% CI: 0.753–0.808) and 0.718 (95% CI: 0.682–0.755) respectively (Figure 1). A cut-off of 5 points to identify patients with high PWV in the external validation cohort yielded a positive predictive value, negative predictive value, sensitivity and specificity of 60.7%, 84.8%, 51.9% and 78.3% respectively.ConclusionsThe SAGE score that takes into account easily measured clinical variables (systolic blood pressure, age, glucose and eGFR) can be used to predict elevated levels of PWV and prioritize its measurement in specific patients. Its use will result in greater acknowledgement of the role of aortic stiffness and aid physicians in implementing it in clinical practice.