FAU medical computer scientists utilise patient data for medical research purposes
When doctors and nurses record the treatment of a patient in hospital medical records, it is not just important for the success of the course of treatment, but also provides research scientists with valuable information for use in medical research. Data known as routine clinical data, such as the age and sex of the patient, pre-existing illnesses/conditions, blood pressure, blood count, medication dosage or OP reports which are documented in the patient’s medical records can, for example, be used for comparative medical studies or for quality assurance in hospitals. Prof. Dr. Hans-Ulrich Prokosch from the Department of Medical Computing Science at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) and his working group have set themselves the goal of using this data more effectively in the future. Over the next three years the research scientists are to receive around €900,000 in funding from the German Federal Ministry of Economics and Technology (BMWi), the German Federal Ministry of Education and Research (BMBF), and the EU’s Framework Programme for Research. This funding will support the three current research projects which will primarily focus on different aspects of analysing, processing and utilising clinical data.
In the first project “Integrated Data Repository Toolkit” (IDRT) the BMBF is supporting the creation of a platform which brings together data from clinical practice and pre-clinical research within a data protection compliant format. One particular challenge facing developers is the task of putting the often very heterogeneous databases from various sources into one analogous form. Professor Prokosch and his team intend to adapt existing, tried-and-tested IT architectures to the requirements of integrated medical research in Germany, and integrate them into one seamless project. As such, using freely available, open source components such as i2b2 (Informatics for Integrating Biology and the Bedside) and Talend Open Studio, should help lay the foundations for sustainable use by an increased number of users. The project involves the participation of FAU medical computer scientists in conjunction with colleagues from Göttingen and Leipzig as well as the TMF (Technology, Methods and Infrastructure for Networked Medical Research) umbrella organisation.
While structured data, for example data from electronic patient records, has already been analysed successfully in numerous different research projects, unstructured information such as medical reports, and radiology and pathology results have remained largely unexploited for research purposes. And this, indeed, is the starting point for a further cloud4health project led by Erlangen medical computer scientists and sponsored by the BMWi. The researchers also hope to make use of these resources by means of textual analysis and interpretation. It is hoped that by using cloud computing technologies and concepts, dynamic resources could be made available for essential CPU and memory intensive processes. Furthermore, services are to be developed for the processing of data as Software-as-a-Service (SaaS), which will in turn enable the development of registers, the identification and assessment of adverse reactions to pharmaceutical products and help carry out financial management audits of medical services. The project will place particular emphasis on data protection considerations – after all, who actually feels comfortable providing sensitive data to a global cloud? Erlangen medical computer scientists will provide expertise for medical processes and data warehousing, and Erlangen University Hospital shall provide support in the form of the evaluation of the processes developed. The Department for Medical Computing Science is also working with a number of other partners: these include, Averbis GmbH – the project’s consortium manager, which is providing long-standing experience and tools for automated text analysis –, the Frauenhofer Institute for Algorithms and Scientific Computing (SCAI, in St. Augustin), which boasts extensive experience in the field of cloud computing and text analysis processes, the TMF with its expertise in data protection issues, and Rhön-Klinikum AG which provides a further evaluation perspective.
The team of Erlangen experts participating in the EU project eHealthMonitor hope to create an application for patients that brings together data from electronic medical records and information from other data sources such as wearable sensors and the Internet. The objective is to create a personal health database which does not only collect data, but also selectively supplies and interprets information based on the individual preferences and pre-existing illnesses/conditions of the user. The greatest challenge lies in the complexity of the data sources which have to be conflated in spite of their different structures, quality and protection requirements.
“Doctors have to spend a great deal of their working day carrying out daily routine documentation in hospital, it is therefore our aim to make this treasure chest of electronically available data readily available for research purposes”, explains Hans-Ulrich Prokosch. “For some time now both the medical research community and the pharmaceutical industry have committed themselves to making it their goal to accelerate medical advancement through new innovative research approaches and the optimised use of data. With these three research projects we are systematically continuing down the path we began as part of the larger EHR4CR research project (borne by the European Federation of Pharmaceutical Industries and Associations within the framework of the Innovative Medicine Initiative). This in turn gives rise to concepts and open source tools, which on the one hand open up new perspectives for the German research community in terms of transnational research projects, and on the other hand could also help identify business models for SMEs”, says Prokosch.
More information for the media:
Prof. Dr. Hans-Ulrich Prokosch
uni | media service | research No. 16/2012 on 20.4.2012