Predicting Cardiovascular Disease Rates in England Using Novel Mathematical Models (microPRIME)
Funder: British Heart Foundation
Project dates: Jan 2016 – Dec 2019
The aim of the project is to develop a novel microsimulation model that will predict future heart attack incidence rates in the UK. Although there are many scenario models that predict future population rates of non-communicable diseases (NCDs) they are all based on the extrapolation of previous trends in the disease outcomes, for example extrapolating previous trends in stroke mortality. As a result, such models cannot predict future changes that may result from adverse trends in risk factors (e.g. possible increases in the incidence of heart attacks that may result from increases in obesity and diabetes).
Predictions of future heart attack rates in the microPRIME model are entirely endogenous and are emergent properties of the model; that is, the prediction of the incidence of heart attacks in year n is independent of the prediction in year n-1. In microPRIME, the incidence of heart attacks in each year is a result of predicted risk factor and treatment levels in the population, which could lead to changes of trends in disease incidence in projection scenarios as the trends in risk factors and treatment compete with each other. The methods borrow from pioneering work in infectious disease modelling and use an approach called ‘history matching’ (example of the history matching approach). This approach involves building a theoretically plausible mechanistic model with unknown parameters and then identifying the values of the unknown parameters via calibration of model outcomes with an external measured dataset. It is anticipated that the microPRIME model will be completed by mid-2018 and then be used for scenario analyses and developed for inclusion of more NCD outcomes. The microPRIME project is a British Heart Foundation Intermediate Basic Science Research Fellowship, awarded to Dr Peter Scarborough in 2015.