Predicting Renal Failure
Renal disease can be asymptomatic for many years, but early detection and treatment can slow the rate of progression towards renal failure. Analysis of routinely collected biomarkers of kidney function can assist early detection. Current UK guidelines use the estimated glomerular filtration rate (eGFR) as an overall measure of kidney function and recommend that a patient who is losing kidney function at a relative rate of at least 5% per year, as measured by their eGFR, should be referred to a specialist treatment centre. In this study, we consider use of dynamic linear modelling to obtain the predictive distribution of the underlying rate of change in kidney function. Our overall aim is to incorporate model-based predictions into a real-time surveillance system that can alert general practitioners to the possible need for the referral of their patient to a specialist treatment centre.
This work is in collaboration with John New from Salford Royal Hospital