Cardiovascular diseases (CVDs) are among the leading causes of death worldwide and a growing global CVD morbidity poses a significant challenge to clinical teams. Given that 70% of CVD cases and deaths are linked to modifiable risk factors, such as smoking, poor physical activity and alcohol intake, we need to focus on primary prevention of CVD rather than treatment alone to reduce the number of avoidable deaths.
To help identify individuals who are at high risk of developing a CVD in the next 10 years, we created an accessible 10-year digital cardiovascular risk model, which we named DiCAVA. DiCAVA uses both traditional statistical and machine learning (ML) approaches to identify individuals at highest risk of CVD based on health data that can all be collected remotely. During the study, we also aimed to determine new patient-centric variables that could be incorporated into future CVD risk models to improve their accuracy. Our research, based on UK Biobank data collected from 466,052 participants recruited between 2006 and 2010, is published in European Heart Journal Digital Health (EHJ-DH).
As part of our research we wanted to test the impact of removing the need to assess cholesterol and blood pressure, two of the strongest markers of CVD risk, as these variables cannot easily be collected outside of a clinical setting. We found eliminating these variables did not significantly decrease the risk model’s ability to predict which individuals were at high risk of CVD. This shows that high-risk individuals can be successfully detected based on easily-accessible lifestyle data, such as level of physical activity, waist to height ratio, family history and hypertensive medication. This data could all be reported by potential patients via a smartphone app to clinical teams who would be able to easily view the information on a web-based dashboard and quickly identify those at high risk of CVD.
The development of DiCAVA is an encouraging step towards enabling more predictive and preventative care. It shows that clinical teams can monitor and accurately predict an individual’s risk of developing a CVD with minimal disruption to the patient and while conserving hospital resources. We believe risk scores like this can be applied across a wide range of conditions and will play a vital role in helping people live longer, fuller lives.
To learn more about our remote patient monitoring tool and how it works, click here.
¹ Modifiable risk factors, cardiovascular disease, and mortality in 155722 individuals from 21 high-income, middle-income, and low-income countries (PURE): a prospective cohort study. Lancet 2020;395:795–808.