Pb1979: data integration between clinical research and patient care: a framework for context-depending data sharing and in silico predictions

Topic: 8. Chronic myeloid leukemia - Clinical Background: The transfer of new insights from basic or clinical research into clinical routine is usually a lengthy and time-consuming process. Conversely, there are still many barriers to directly provide and use routine data in the context of basic and clinical research. In particular, no coherent software solution is available that allows for a convenient and immediate bidirectional transfer of data between treatment situations and research settings. Aims: We aim to develop a generic concept for context-depending data-sharing and integration of computer simulations for both scientific research and routine care in a common hospital setting, specifically in terms of centralised and privacy-compliant management of patient-identifying and medical data. The resulting software is intended for simultaneous use in clinical and research settings. Methods: Our software solution is based on the high-level Python Web framework Django and builds on a recently established generic data management concept. To this end, we designed and implemented a web-based software framework providing the necessary features for data analysis, visualization as well as the integration of computer simulation and model predictions with audit trail functionality and a regulation-compliant pseudonymization service. Results: Our established framework demonstrates a strategy how quality-assured and harmonized health data from multiple decentralized clinical data sources can be provided to multiple users working in different clinical or research contexts and how analytic results such as mathematical model predictions can be returned for direct application in clinical decision making. As a core element of the research data management infrastructure, we integrated an independent Trusted Third Party/pseudonymization service to meet the necessary standards for data protection and ethical requirements. At the user side, the front-end application provides two tailored views: a clinical (i.e., treatment context) perspective focusing on patient-specific data visualization, analysis and outcome prediction and a research perspective focusing on the exploration of pseudonymized data. We illustrate the application of our generic framework by two use-cases from the field of hematology/oncology, specifically relating to Chronic Myeloid Leukemia (CML). Both computational models estimate a recurrence probability for a given patient in the context of treatment reduction or cessation. Summary/Conclusion: Our implementation demonstrates the feasibility of an integrated generation and backward propagation of data analysis results and model predictions at an individual patient level into clinical decision-making processes while enabling seamless integration into a clinical information system or an electronic health record. Keywords: Prediction, Dose intensity, treatment-free remission, Prognosis