The PERSIMUNE data warehouse builds on the experiences from the MATCH collaboration that developed an expert tool providing personalized medicine to transplant recipients, and implemented as a routine clinical tool with documented benefits in terms of reduced morbidity and cost of patient treatment and care. MATCH captured routine data daily from all analytical & pathological labs, from medication systems, the national quality assurance database for rheumatology patients, and imaging data at Rigshospitalet. Outputs are structured monitoring and treatment plans for all transplant patients at Rigshospitalet, and an active alerting of clinicians in case of out-of-range lab values or missed samples.
The PM structure has expanded the data capture to include biochemistry, microbiology and pathology data from all national labs via MedCom and in addition diagnosis and data on causes of disease, vital sign and medication from central Regional databases. Also links to local databases for disease areas or local lab data are part of the PM patient management system that has a user interface where clinicians can see data from their own patients. The database is approved by the Data protection agency (RH-2015-04, I-suite 03605).
Technically: Based on patient-ID driven, timed requests from PM, data from all different sources are stored in basic layer mirrored tables in the data warehouse, reducing pressure on source databases. The second layer is uniformed, standardised storage of data from different sources. Algorithms to issue monitoring and treatment plans, as well as alert alarms, are developed in collaboration with clinicians.
For research purposes frozen datasets for the specific projects will be made available for researchers after relevant internal and external approvals.
In the initial period, the main focus is on establishing stable access to all the data sources and ensure that the data captured reflects the complete data available for the individual patients. Also standardizing the data from the many sources is a focus-point to improve usability and uniformity of the data.