The need to include not only diagnostic but also research and other types of data in the assessment of the Big Data future health
Our partner Dimitra Tsaoussis, from the DIAS (Data-Intensive, Applications and Systems) Laboratory at EPFL (École Polytechnique Fédérale de Lausanne) interviewed Prof. Bogdan Draganski, neurologist at Centre hospitalier universitaire vaudois (CHUV) and Mr Nicolas Rosat, CHUV IT Manager.
The CHUV is one out of the five Swiss university hospitals, which collaborates with the Faculty of Medicine and Biology of the University of Lausanne. With a budget of 1.5 billion Swiss francs in round figures (2014), it counts 9,999 collaborators who attended 46,146 patients during last year 2014.
Prof. Draganski, director LREN (Laboratoire de recherche en neuroimagerie) - Department of Clinical Neurosciences, explains that the CHUV core competencies in this field are an accurate diagnosis of brain disorders based on the analysis of behavioural, clinical and brain imaging data. In order to solve the Big Data management problems that CHUV faces in its everyday work, like the provenance of the clinical data, missing data and low-quality and unstructured data, the hospital invests in its own research projects of LREN, DNC (Département des neurosciences cliniques)-CHUV and other European projects like the Human Brain Project.
The IT CHUV department has 150 employees and a budget of 40 million Swiss francs. As Mr. Rosat explains, “during the last 12 years, the CHUV has invested money in building and using a datawarehouse. The current data volume is 100 Gb, which is quite common and there is no performance problem”.
Mr. Rosat notes that “biobanks are not integrated in the datawarehouse but have ad hoc IT applications”.There are more than 30 biobanks in the CHUV.
Prof. Draganski shares with us some of his main concerns. As he explains, the data generated at CHUV for diagnostic and research purposes is always linked to patients’ health. “What we consider as diagnostic data, next to data acquired for scientific projects, can be used to address questions pertinent to research. The constraint to use such data for research is the fact that data acquisition for diagnostic purposes is covered by contributions to the health insurance. The majority of these data are viewed only once to help medical decision-making to end up in the archive systems of the hospital. The potential of this Big Data remains largely unused”. Prof. Draganski believes that it is really important to include all types of data in the assessments related to Big Data. “The tendency in Europe is to use biobanks. Biobanks exist in Lausanne and elsewhere in Europe. This is data donated and the researcher has full access to it. Otherwise, patients’ data falls under the ethics’ legislation”, explains Prof. Draganski.
As part of its strategy for 2014-2018, the CHUV has launched a project around a biobank in Lausanne (BIL) whose purpose is to collect patient data to the benefit of a general consent. To support this initiative CHUV IT has created a datawarehouse for clinical data, image data and genomic data.
As Mr. Rosat explains, “for the sequencing and the genomic data processing, a collaboration between the CHUV and SIB/Vital-IT (Lausanne University) was initiated”.
Still at an early stage, CHUV plans to build a data depot to provide their researchers with clinical data. As noted by Mr. Rosat, “the main issues are going to be:
- The security of data access, issue that becomes bigger with the genomic data processing
- The data reusability (also known as «secondary use»): how to qualify the data quality and how to use nomenclatures and ontologies”
This interview is one of the several that the partners of RETHINK big are doing to the main stakeholders in order to identify the industry coordination points that will maximize European competitiveness in the processing and analysis of Big Data over the next 10 years.
- Bogdan Draganski LinkedIn profile
- Nicolas Rosat LinkedIn profile
- CHUV website
- CHUV Département des neurosciences cliniques