120 Disparity in cardiovascular risk stratification between a novel non-invasive arterial stiffness method and three conventional risk scores in a high risk population
Background Current international guidelines for predicting cardiovascular disease (CVD) risk include non-invasive assessment of arterial stiffness. We aimed to evaluate how estimates of pulse wave velocity (PWV), calculated using a new, rapid, validated non-invasive device, compare with traditional risk scores in discriminating patients at increased risk of CVD. Subjects and Methods: We used a non-invasive oscillometric device (Arteriograph, Unimedic Ltd) to estimate PWV in patients with and without established CVD (n=80, 79% male; mean age 67±1(SEM)years). Risk factors included hypertension (70%), hypercholesterolaemia (65%), type 2 diabetes mellitus (35%) and current smoking (18%). 59% had clinical ischaemic heart disease. Individuals were categorised as ‘high’ or ‘low’ CVD risk according to published thresholds using 3 established methods (Framingham (JBS2) risk charts, European Society of Cardiology (ESC) HeartScore® and the Pocock risk score). Data were analysed using SPSS (version 14.0). The study was approved by the local research ethics committee.
Results PWV was positively correlated (Spearman) with increasing risk estimated by all three scoring systems (Framingham rs = 0.420, p = 0.001; HeartScore rs = 0.394, p = 0.000; Pocock rs = 0.305, p = 0.006). For all three scoring systems, in at least 25% of subjects assessed as low risk, PWV indicated CVD high risk (Framingham: 27%; HeartScore: 25%; Pocock: 44%). In contrast, around half of our subjects identified as ‘high’ risk by conventional risk scores, were identified as in the low CVD risk category when assessed by PWV (Framingham: 49%; HeartScore: 47%; Pocock: 50%).
Conclusion We observed important disagreement in the stratification of CVD risk in a high risk population by an index of arterial stiffness compared with commonly used risk scores. One interpretation of our data is that combination of non-invasive measurements of arterial stiffness with classical CVD risk scoring systems may improve stratification in high risk populations. However, further prospective work is needed to determine clinical outcomes for patients misclassified by assessment of arterial stiffness compared with classical risk scoring. This will help to test the clinical utility of combined risk scoring and arterial stiffness, in producing finer CVD risk stratification.