Impact of outcome definitions on cardiovascular risk prediction in a contemporary primary prevention population

Background — Estimation of an individual's cardiovascular disease (CVD) risk may enhance risk discussion and treatment decisions. Yet, common cardiovascular outcomes such as heart failure or coronary revascularization are not included in the estimation of atherosclerotic cardiovascular disease (ASCVD) risk. Our objective was to determine the incidence of ASCVD in a contemporary primary prevention population with more than 10 years of follow-up, and how incidence estimates change when incorporating additional cardiovascular endpoints.

Methods — We used the population-level CANHEART (Cardiovascular Health in Ambulatory Care Research Team) database of all Ontario residents alive January 1, 2008, aged 30–99 years, and with no prior history of cardiovascular disease. Individuals were followed to December 31, 2019 for incident first and recurrent cardiovascular events. ASCVD outcomes were defined by hospitalizations for myocardial infarction, stroke and circulatory death, while global CVD outcomes also included hospitalizations for unstable angina, transient ischemic attacks, peripheral arterial disease, out-of-hospital cardiac arrests, heart failure and coronary revascularization.

Results — Among 7496 165 individuals free of cardiovascular disease, their mean age was 50 years (SD: 13.9 years) and 52.3% were women. After 11 years of follow-up, the rate of an incident ASCVD event was 3.95 per 1000 person-years while the rate of a global CVD event was almost doubled at 6.67 per 1000 person-years. The most common additional first manifestations of CVD were heart failure which accounted for 12.0% of additional events and coronary revascularization which accounted for 12.7%. When considering first and recurrent events, the rate of ASCVD was 5.20 per 1000 person-years while the rate of all global CVD events was more than double at 10.90 per 1000 person-years. This was mainly due to a higher proportion of recurrent heart failure (13.7%) and coronary revascularization (23.2%) events.

Conclusions — ASCVD accounts for just over half of all preventable first cardiovascular events and even fewer first and recurrent cardiovascular events in contemporary practice. Estimating broader CVD endpoints may enhance risk-discussions with patients and improve informed decision-making.