SP 1 - CKDNapp

Dataset consolidation, generation of clinical input variables, and medical interpretation and validation

Chronic kidney disease (CKD), a leading cause of death, is a common, complex disease with a variable disease course and multiple comorbidities making adverse event and disease prediction, as well as comorbidity/medication management challenging. With the help of clinical decision support (CDS) software these issues can be tackled. CDS can be programmed on the basis of mathematical models derived from actual patient data. Meaningful model calculation requires a vast, comprehensive variable set, preferably longitudinally collected, as provided by prospective cohort studies. Here, physicians carry out variable definition, and weighing, as well as validate, and assess endpoints from collected patient records building the basis for any event prediction.
The German Chronic Kidney Disease (GCKD) study is a multi-center, prospective cohort study with collected and continuously abstracted data of 5,217 patients (begin of study 2010-2012). Thus, it provides a perfect resource for developing CKDNapp. All CKDNapp models will be based on collected and continuously abstracted GCKD data.
This subproject (SP 1) pursues three aims to provide and optimize GCKD data for further use and to develop CKDNapp, enabling physicians to practice personalized CKD care: (1) Variable and dataset definition/abstraction for mathematical modeling and CKDNapp development, (2) medical interpretation of all mathematical models, and (3) assessment and refinement of CKDNapp. This subproject is carried out at the Institute of Genetic Epidemiology at the University Clinic Freiburg.

Dataset consolidation, generation of clinical input variables, medical interpretation of estimated models and CKDNapp validation.