Abstract: SA-PO850
Developing a Clinical Decision Support Software to Monitor and Tailor Treatment of CKD Patients
Session Information
- CKD: Socioeconomic Context and Mobile Apps
November 09, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: CKD (Non-Dialysis)
- 2101 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention
Authors
- Schultheiss, Ulla T., Medical Center - University of Freiburg, Freiburg, Germany
- Altenbuchinger, Michael C., University of Regensburg, Regensburg, Germany
- Eckardt, Kai-Uwe, Charité – Universitätsmedizin Berlin, Berlin, Germany
- Oefner, Peter J., University of Regensburg, Regensburg, Germany
- Gronwald, Wolfram, University of Regensburg, Regensburg, Germany
- Kottgen, Anna, Medical Center - University of Freiburg, Freiburg, Germany
- Raffler, Johannes, Helmholtz Center Munich, Neuherberg, Germany
- Zacharias, Helena U., University Medicine Greifswald, Greifswald, Germany
Group or Team Name
- GCKD investigators
Background
CKD is a complex disease with several therapeutic challenges: silent onset; different: etiopathologies, progression patterns, prognosis and comorbidities; polypharmacy. Clinical decision support (CDS) software may improve CKD management.
Methods
Within the German Chronic Kidney Disease (GCKD) study, a multi-center, prospective, observational CKD stage 3 cohort study, demographic, phenotypic, and clinical parameters of 5,217 Caucasian patients have been collected. We will model these data and variable dependencies by state-of-the-art machine learning methods to predict the risk of adverse endpoints, to interpret the results in the context of current biomedical knowledge, and to use the estimated models as a backbone for a CDS software, which will be provided as a user-friendly app to nephrologists (CKDNapp).
Results
CKDNapp, based on mathematical models and enriched by state-of-the-art literature (Fig 1 (1)), will assess patient parameters in one consistent framework, schematically represented by a network (Fig 1 (2)). Only the mathematical models will enter the app (Fig 1 (3)) ensuring data security. Nephrologists will be able to use CKDNapp as a CDS system (Fig 1 (4)), entering patient data like clinical parameters and disease history. CKDNapp will integrate all provided patient information and return personalized adverse event / disease progression prediction, support medication management, and offer in silico modification of patient parameters exploring effects of future treatment strategies.
Conclusion
The overall goal of CKDNapp is to facilitate personalized CKD patient treatment. First results will be available for presentation at ASN Kidney Week 2019.
Figure 1 Schematic workflow of the development and application of CKDNapp.
Funding
- Government Support - Non-U.S.