SMITH’s three university hospitals – in Leipzig, Jena and Aachen – unite medical informatics, clinical, systems-medical, computer linguistics and epidemiological knowledge and expertise. They work hand-in-hand with external partners to develop IT architecture for the interoperable use of data from healthcare and patient-oriented research – data shared across the boundaries of individual institutions and geographical locations. Consortium partners will be able to access and use the results via a special marketplace.
SMITH is funded by BMBF during the development and networking phase since January 2018.
Data integration centres
At each of its three locations, SMITH will establish data integration centres (DICs) that coordinate and cooperate with each other. The DICs will act as data brokers and trust centres to process and organise data, amongst other tasks. They will also operate integrated databases with metadata directories.
The three DICs will all be established in the same way – i.e. embedded in the university hospitals, with access to patients’ electronic medical records (EMRs). As a result, data analyses can be directly informed by patient care delivered under Germany’s statutory healthcare system.
Phenotype pipeline (PheP)
In a methodological use case, SMITH will develop a phenotyping platform for selected indications using data from electronic medical records. The consortium will employ innovative data-analytics methods to make this data suitable for clinical research and improved patient care.
The clinical ASIC use case is aimed at the model-based algorithmic surveillance of critically ill patients in intensive care units (ICUs) – by means of continuous analyses from patient data management systems. This forms the basis for faster diagnostic and therapeutic interventions.
The Help use case focuses on antibiotic stewardship – i.e. the targeted and appropriate use of antibiotics to combat bacterial infections at an early stage. The goal is to leverage IT for infection control in general wards and intensive care units (ICUs).
Data use and access concept
Data use is local, via the data integration centres – which have access to hospital information systems (HIS) and, consequently, enable the use of patient data. This information is analysed and enriched at the individual hospitals. Only authorised DIC employees will have access to these data via the local operative HIS system.
Research using patient data (with locally made changes and notes, etc.) is only possible with the express consent of the patient. The patient is actively involved and able to give consent project-by-project – and in this respect, the DICs act as trust centres.
Measures to strengthen medical informatics in Germany
- Design of joint, coordinated modules for advanced training and continued education
- Proposals for curricula for a master’s of science (MSc) and post-graduate work in medical informatics
- Establishment of a professorship for medical informatics at University Hospital RWTH Aachen
- Establishment of a professorship for medical informatics at University Hospital Jena
- Establishment of a professorship for medical data science at Leipzig University
The SMITH consortium was managed by a steering committee comprising three scientific coordinators – Prof. Markus Löffler, Prof. André Scherag and Prof. Gernot Marx – and three CIOs – Stephan Smers, Andreas Henkel and Volker Lowitsch.
The current draft concept foresees the establishment of data integration centres (DICs) at all three locations. These centres will cooperate closely. Acting as data brokers and trustees, their tasks include preparing and organising data for publication. To this end, the DICs are to operate integrative databases with a metadata directory.
In addition, a concept for a phenotyping methods platform was developed. For selected indications and topics, it will create a new data set from health-record data for patients with certain phenotypes. This could be beneficial for clinical-epidemiological and health-economics issues, for example (focus of work in Leipzig).
Furthermore, the consortium developed concepts for various use cases.
The three sites also designed a shared training and skills/professional development module that offers modular curricula for a Master of Science (MSc) degree and postgraduate studies.