Causation and Novel Risk Modelling for Person-Centred Prevention and Control of Cardiovascular Diseases
Cardiovascular risk prediction is important to health promotion and to support decisions on primary preventive interventions of cardiovascular disease (CVD).
The invention of new biomarkers and medical technologies and the explosion of big data and its analytical methods create opportunity to identify cardiovascular diseases at early stage. This project answers the call from clinicians for a new and more accurate clinical guideline for diagnosis and treatment of CVD among asymptomatic individuals at medium or high risk, fail to be identified using existing risk models. Linking data generated from a randomised clinical trial and quality national patient and death registers, we will analyse the complex interaction between biological, behavioural, psychological and social determinants and asymptomatic carotid atherosclerosis in a subset of adult population 40+ years old in Västerbotten County in Sweden, and develop individualised risk prediction models for 3-year and 5-year cardiovascular disease morbidity and mortality. This 5-year project brings together seven multidisciplinary researchers from clinical, public health and social sciences. We will employ advanced epidemiological and statistical methods such as multilevel structural equation modelling and latent trajectory model to address the research questions. We will use employ machine learning technique, such as boosted regression tree in developing the risk prediction models. This project will contribute to innovative risk model for clinicians to tailor a person-centred cardiovascular disease prevention and control programme for individuals at risk.