Grant: Operating Grant

Chronic kidney disease is common in older populations. The major health consequences are initiation of renal replacement therapy (dialysis or transplant), hospitalization for heart disease and death. Currently, different levels of kidney dysfunction are classified using an estimated glomerular filtration rate (eGFR) which is calculated based on age, gender, serum creatinine and race. Recently, a new system has been proposed incorporating a measure of protein in the urine (proteinuria) into the classification system. The relationship between proteinuria and kidney dysfunction is not known in the older population. This work proposes to study the relationship between proteinuria and kidney dysfunction, to develop and test, specifically for the older population, a classification system incorporating both proteinuria and eGFR and to derive and test a model to predict death, hospitalization for heart disease and worsening of kidney dysfunction. Established databases will be used to select patients with kidney dysfunction who will then be followed forward in time to see if they develop major health consequences from their kidney dysfunction. We will divide our population into age groups based on a statistical model indicating ages at which the relationship between age and health consequences significantly changes. We will then calculate the age-adjusted rates for each outcome, for both the new and old classification system. We will assess the model's ability to distinguish different risk levels in patients in our dataset as well as an external dataset. It is important to understand if the relationship between proteinuria and kidney dysfunction differs for the older population. Current a tool to predict risk of poor outcomes among older adults with kidney disease does not exist. This would provide a valuable tool that doctors could use to improve patient care by increased recognition and communication of patient risk.