Background
In the UK, around 10% of adults have chronic kidney disease, which is becoming more common. Some people go on to develop end stage kidney disease or kidney failure (ESKD). This affects the quality of life of those affected and increases the risk of death. To stay alive, those with ESKD need either dialysis or a kidney transplant. People from non-White ethnic backgrounds are at least twice as likely as those from a White background to require dialysis or a kidney transplant. The Kidney Failure Risk Equation (KFRE) takes into consideration your age, gender, protein levels in urine, and how well your kidneys are functioning in order to calculate your risk of kidney failure, but it does not work as well for some ethnic groups. Recognizing any factors that might make it difficult for people to get tested, and how people view their kidney health, could help them get treatment sooner.
Novelty & Importance
This project will uniquely attempt to determine the number of people who have advanced kidney failure who have opted against dialysis using UKRR data by using machine learning (ML) methods. To inform the ML approach we will use local databases linked to the UK Renal Registry, UKRR dataset and compare outcomes in primary care data. We aim to apply data mining over various data sources and address missing data.
Aims
To develop novel risk prediction algorithms for differentiating individuals preparing for dialysis from those receiving comprehensive conservative care.
Objectives
Year 1: Develop risk prediction algorithms in 1ry and 2ry care and data items to enhance accuracy.
Year 2: To internally validate these algorithms using records split from the original datasets.
Year 3: To pilot and validate use of personalised risk prediction algorithms for identifying patient treatment plans, comparing traditional and ML methods.