Comprehensive analysis of traditional and novel risk factors for stroke in persons with kidney failure in Ireland

Patients with kidney failure (dialysis or renal transplantation) are up to 20 times more likely to suffer from stroke compared to the general population (1). Several risk factors for stroke have been identified in patients receiving dialysis, including traditional shared risk factors such as hypertension, diabetes, atrial fibrillation, and heart failure (2), and hemodialysis-specific risk factors including cerebral hypoperfusion (3) and blood pressure variability (4). The Kidney Disease Clinical Patient Management System (KDCPMS) is a patient management system used in nephrology centres in Ireland since 2007. A multi-university and multi-agency [National Renal Office, Irish Kidney Association, Irish Nephrology Society, National Office of Clinical Audit, and HSE Health Identity Management Services] collaboration has been developed to answer fundamental questions regarding stroke risk, and outcomes, in the Irish kidney failure population. Objective: To identify traditional and novel risk factors for stroke incidence and outcomes in Irish patients with kidney failure requiring dialysis or renal transplant, which will inform Irish stroke prevention strategies. Design: Retrospective data analysis from national administrative databases. Data linkage will be performed using the HSE Individual Health Identifier (IHI). Participants: Persons at 26 renal centres in Ireland. Methods: Data from KDCPMS will be linked to the Irish National Audit of Stroke with the IHI. Natural Language Processing Models will be used to extract structured data regarding vascular access type, atrial fibrillation, and past medical history from free-text dialysis notes. Hypothesis-driven associations between known stroke risk factors (1)traditional risk factors, 2)haemodialysis-specific risk factors, and 3)vascular access-specific risk factors) will be determined and compared to international data. A stroke prediction model using gradient-boosted decision trees will be created to guide stroke prevention strategies. Impact: The outcomes will directly inform national stroke prevention strategies. The permanent integration of the HSE IHI with KDCPMS will also serve as a catalyst for future research.