Carotid intima media thickness measurement in cardiovascular screening programmes

Background Cardiovascular (CV) risk-assessment methods such as the Systematic Coronary Risk Evaluation (SCORE) system divide subjects into risk categories based on age, gender, cholesterol, blood pressure, and smoking: low (<5% 10 year CVD mortality risk); intermediate(5-9%): and high (>10%). Drug treatment recommendations are linked to these categories. However, risk stratification by SCORE leaves much room for improvement. The use of non-invasive measurements of atherosclerosis, such as carotid intima media thickness (CIMT), has been suggested to improve risk assessment based on the consistent findings of increased relative risks with increasing CIMT. Yet, a recent systematic review of the role of CIMT measurements in CVD screening programs concluded that data to support a role for CIMT measurements in risk profiling, i.e., data from studies showing that for an individual a high or low CIMT measurement leads to a shift from one to another risk category and this shift is followed by different treatment consequences, are lacking. Despite the absence of evidence on added value, increasingly commercial initiatives promote to have your CIMT measured on the premise that a CIMT measurement much better assesses an individual's future CV risk than traditional risk factors do. Solid evidence should become available to refute or support the view before efforts and costs are wasted and false feelings of security in people are created. Aim To assess whether a CIMT measurement helps to distinguish a high risk from a low risk subject, above and beyond the contribution of traditional risk factors, at acceptable cost? The general aim is divided into several specific objectives: 1. Does a CIMT measurement help to distinguish a high risk from a low risk subject, on top of the information available from traditional risk factors? 2. Are there subgroups** within the general population, in which a CIMT measurement helps to distinguish a high risk from a low risk subject, on top of the information available from traditional risk factors. ** subgroups are for example, men/women, age groups, subjects with hypertension/hyperlipidemia/obesity; subjects with intermediate risks estimated by Framingham/SCORE prediction models. When either of the above objectives is confirmative: 3. Is it cost-effective to add CIMT measurement to traditional risk assessment with cardiovascular risk factors? Methods Design: Individual participant data from cohort studies will be pooled into one cohort for the current study. This project requires pooling of individual participant data from existing cohorts because of the need for precise estimates of the observed CV risks in subjects that go from high to low risk and vice versa. Study population: These cohorts are general population based, as a CIMT measurement for risk profiling focuses specifically on subjects aged 40-75 yrs free from symptomatic CVD and diabetes mellitus. We aim at pooling of ten to fifteen cohorts, leading to one total cohort of over 55.000 subjects with over 3000 CV events. We have contacted principal investigators of 6 cohorts already, and all were willing to participate. Data from 5 cohorts is already in Utrecht. Measurements: Available data include risk factor levels needed to estimate the SCORE risks (age, sex, systolic pressure, smoking, total and HDL cholesterol). Participants need to have been followed for occurrence of CV events with data on type of CV event, fatal or non-fatal, date of event and date of baseline examination and all cause mortality. CIMT at baseline needs to be available. Data analyses: We start with imputing missing values using multiple imputation. Next, we combine all cohorts. A derivation cohort (75% random sample) and a validation cohort (25% random sample) will be constructed. With the derivation cohort the additional value of a CIMT measurement on top of SCORE will be assessed using regression models, estimates of change in area under the curve, calibration estimates, and estimates of the re-classification of diseased and non diseased for a priori defined probability thresholds and across all probability thresholds. This approach will be repeated for several subgroups. Economic evaluation A full economic evaluation of the lifetime costs, benefits and cost effectiveness in terms of quality adjusted life years gained of switching from usual care to CIMT approach will be performed when the incremental benefit of CIMT in the entire population or in sub-groups is shown. Yield Refute or support the view that CIMT has added value to identify groups at high or low CV risk before efforts and costs are wasted and false feelings of security in people are created.