Approved research
Comparative analysis of current health risk prediction capabilities
Lay summary
Comparative analysis of current health risk prediction capabilities
As non-communicable diseases strain global health systems, prevention of diseases is more important than ever. Existing risk prediction models often lack holistic approaches, and expert opinion remains the gold standard, particularly in certain health topics. However, expert opinion is a scarce and valuable resource. Furthermore, experts are often only consulted when health problems have already come up. Both these problems have a chance to be addressed by having accessible, digital tools available to the general public. However, these can only benefit patients and the healthcare system if they provide valid and actionable information.
The central problem is that "risk" is not a well-defined term when it comes to health - there is no universally accepted definition. Despite this, there are various risk calculator tools available which e.g. assess a user's risk of developing a cardiovascular disease within the next 10 years or prescribe a "Heart Age". These are usually very specific and not transparently validated.
This research project aims to understand the current state of risk prediction by evaluating existing risk prediction tools, expert opinion, and a novel, general-purpose digital health risk assessment model called Health Check.
This project will generate insights into the validity of tools by determining whether a higher risk output by the tool would in fact correspond to a shorter time to disease. We will perform an equivalent analysis with physicians where we ask them to order a set of patient profiles according to their risk. As evident from a systematic literature review we are currently conducting, very little research has been published on either the validity of tools or the accuracy of experts.
In conclusion, this research endeavours to advance the understanding of medical risk assessment capabilities and the validity across various disease groups. As a secondary output, this project enables the validation of a digital tool that will enable people to predict and prevent their diseases. The overarching goal is to enable cost-efficient prevention of diseases in the general public, reducing the increasing pressure on the healthcare system and improving the quality of life on an individual level.