Digital Infection Medicine: Individualised Patient Care
Key area 3
Coordination: Hannover Medical School, MHH
The challenges of the pandemic include not yet conclusively researched disease courses and still unknown or not yet established prognostic markers. The identification of such markers and the understanding of the pathogenesis are based on high-quality data on individual courses of the disease as well as on a comprehensive characterisation of both the patients and the virus.
Clinical course data include not only classical clinical data from inpatient care (e.g., laboratory data, medication, vital signs), but also detailed, high-resolution data from the field of intensive care (e.g., ventilation data, extracorporeal membrane oxygenation), imaging and molecular characterisation (omics). Moreover, syndromic data from the non-clinical setting (patient-generated data, e.g., via apps with symptom recording) should also be taken into account, especially in light of the relapsing courses with intermittent improvement followed by rapid deterioration observed in COVID-19 diseases.
Research aim
On the basis of a vast amount of preliminary methodological work and ongoing work to collect SARS-CoV-2-associated heterogeneous data in both basic infection research and clinical care, the focus of this key area is on the development and practical testing of new informatics methodologies for individualised patient care.
Diagnostic and prognostic models
In this area, work is carried out to identify new diagnostic and prognostic markers and marker combinations from heterogeneous datasets (among others, clinical course data including intensive care, image data, omics, and patient-generated data from mobile apps and sensors) that reveal previously unknown relationships in the pathogenesis and course of coronavirus diseases, thereby enabling rapid and accurate diagnosis as well as prognosis of the course and expected resource requirements for individual patients. For this purpose, methods from the field of machine learning and artificial intelligence are used.
Clinical decision support
In this area, diagnostic and prognostic models, together with the therapeutic strategies developed in other focus areas, will be translated into clinically usable, informatics prototype tools for machine decision support. In addition to the practical testing and evaluation of these tools, the avoidance of bias due to unbalanced data (therefore preferential use of data from multiple sites) and the provision of explanatory components are of particular importance (avoidance of the "black box" problem, traceability of decision recommendations). This work shall be carried out with industry participation, as it concerns medical devices according to MDR (Medical Device Regulation).
Essential preliminary work of the partners
The partners of the COFONI network already did essential preliminary work in the area of data collection and representation for the identification of new biomarkers for COVID-19. MHH and UMG, as core sites in the HiGHmed Consortium of the BMBF Medical Informatics Initiative, each have local data integration centres where a common technical platform has been established.
All clinical data are stored there on the basis of an open, common representation standard and are available for analysis with standardised tools. The implementation of the GECCO data model (German Corona Consensus) defined at national level has taken place.
Hannover Unified Biobank of the MHH
With funding from the MWK, a longitudinal COVID-19 cohort with broad clinical data and diverse biomaterials was established at the Hannover Unified Biobank of the MHH under the leadership of Professor Thomas Illig. In addition, data and biosamples from the Göttingen site (PD Dr. Sara Nußbeck, Central Biobank UMG) are included. Together, data and several biosamples from more than 400 patients are already available. The cohorts will be continued at the two sites by existing study personnel depending on the number of infected patients and will be constantly expanded.
Clinical decision support, especially in the emergency room, is addressed by the UMG in the ALINA and OPTINOFA projects; here, an extension to COVID-19-specific instructions for action is possible.
The Department of Medical Informatics at the UMG is significantly involved in the European registry on COVID-19 in multiple sclerosis and has helped to develop the data models there in particular. In addition, UMG is involved in the LEOSS.core cohort (Dr. Seidler).
Our projects in Key area 3 - Digital Infection Medicine
Contact
Prof. Dr. Thomas Illig
Stellvertretender Direktor Institut für Humangenetik, Leiter der Biobank
Medizinische Hochschule Hannover
Carl-Neuberg-Str. 1
30625 Hannover
Illig.Thomas(at)mh-hannover.de
Prof. Dr. Dr. Michael Marschollek
Leiter Peter L. Reichertz Institut für Medizinische Informatik
Medizinische Hochschule Hannover
Carl-Neuberg-Str. 1
30625 Hannover
Marschollek.Michael(at)mh-hannover.de
