gms | German Medical Science

64. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

08. - 11.09.2019, Dortmund

Portal of Medical Data Models – latest developments and future trends

Meeting Abstract

  • Sarah Riepenhausen - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Julian Varghese - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Philipp Neuhaus - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Cornelia Mertens - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Michael Storck - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Alexandra Meidt - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Stefan Hegselmann - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany
  • Martin Dugas - Westfälische Wilhelms-Universität Münster, Institut für Medizinische Informatik, Münster, Germany

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 64. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Dortmund, 08.-11.09.2019. Düsseldorf: German Medical Science GMS Publishing House; 2019. DocAbstr. 137

doi: 10.3205/19gmds162, urn:nbn:de:0183-19gmds1620

Published: September 6, 2019

© 2019 Riepenhausen et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at http://creativecommons.org/licenses/by/4.0/.


Outline

Text

Introduction: The diversity of medical language makes independent creation of compatible medical data models unlikely. Due to the lack of transparency harmonization of medical documentation is challenging. Interoperability restrictions from diverse, partly proprietary documentation systems complicate data exchange. To address those problems the Portal of Medical Data Models (https://medical-data-models.org/) was developed for open-access to medical data models [1], [2]. Through (uniform) semantic annotations medical concepts of data will be maintained in data exchange and re-use of e.g. routine data for research becomes possible. In order to increase the reach of the portal, contents and functions have been revised and extended.

Materials and Methods: CDISC ODM format [3] was chosen as primary data format for the data models. Those are created and annotated with UMLS codes with ODMEdit [4] which supports standardization by enabling re-use of items, item-groups and UMLS codes [5]. MeSH [6] are assigned to models to facilitate search and allow analysis of content coverage.

Database queries were used to analyze numbers of users, forms, items, (topical groups of) keywords and codes [2].

Appearance, functions, contents and database structure of the portal were modified and complemented based on user feedback and maintenance needs.

Results: In March 2019 >20,300 data models were available to the public, including ~1,300 registered users. The data models consisted of >490,000 items of which 88% were UMLS annotated. The most common medical concept code was “Pharmaceutical Preparation” (C0013227, >3,700). The most frequently used keyword was “Clinical Trial” (>15,000) overall and “Breast Neoplasms” (>1,400) disease-specific.

Java and PostgreSQL now build the technical infrastructure for the portal and its user interface is available in 8 languages. 18 download formats are available, including a revised PDF-converter. In cooperation with the ZB MED – Information Centre for Life Sciences, Cologne, Digital Object Identifiers (DOI) can now be assigned for quotable publication of data models. ODM files can be analyzed with two tools directly from the portal based on the UMLS annotation: ODMSummary grants an overview of the files and sorts items into equivalence classes [7]; the CDEGenerator enables advanced semantic analyses of medical concepts for creation of core data sets [7]. In the portal’s MetaData Registry related items, i.e. items often collected together, can be identified.

Items can be manually annotated with codes from other terminologies, e.g. CDISC SDTM [8].

Discussion: Growing numbers of users, forms and functions of the portal indicate a need for interoperability and transparency in medical research and routine documentation (2). The portal’s contents show broad coverage, especially in oncology – one of the main topics of clinical trials. Data models from routine documentation and underrepresented diseases need to be added. Deprecated and redundant UMLS codes shall be replaced automatically to further standardization and quality.

A revision of the FHIR converters is planned for representation of HL7 FHIR resources [9] as is the use of SNOMED CT [10] once there is a German license. To conform to international standards the portal will be adapted to the ISO-standard “Health Informatics - Metadata Repository Requirements” [11] when available.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.


References

1.
WWU Münster. Portal für Medizinische Datenmodelle (MDM-Portal). [Accessed 2018 Sep 19]. Available from: https://medical-data-models.org/ External link
2.
Geßner S, Neuhaus P, Varghese J, Bruland P, Meidt M, Soto-Rey I, et al. The Portal of Medical Data Models. Where Have We Been and Where Are We Going? Studies in Health Technology and Informatics. 2017;858–862.
3.
CDISC. Operational Data Model (ODM)-XML. [Accessed 2019 Jan 9]. Available from: https://www.cdisc.org/standards/data-exchange/odm External link
4.
Dugas M, Meidt A, Neuhaus P, Storck M, Varghese J. ODMedit: uniform semantic annotation for data integration in medicine based on a public metadata repository. BMC Medical Research Methodology. 2016 Jun 1;16(1):65.
5.
UMLS Terminology Services. [Accessed 2018 Sep 24]. Available from: https://uts.nlm.nih.gov/home.html External link
6.
NCBI. MeSH. [Accessed 2019 Mar 26]. Available from: https://www.ncbi.nlm.nih.gov/mesh/ External link
7.
Varghese J, Fujarski M, Hegselmann S, Neuhaus P, Dugas M. CDEGenerator: an online platform to learn from existing data models to build model registries. Clin Epidemiol. 2018 Aug 10;10:961–70.
8.
CDISC. Study Data Tabulation Model (SDTM). [Accessed 2019 Jan 9]. Available from: https://www.cdisc.org/standards/foundational/sdtm External link
9.
HL7. Resourcelist - FHIR v4.0.0. [Accessed 2019 May 29]. Available from: https://www.hl7.org/fhir/resourcelist.html External link
10.
SNOMED. [Accessed 2019 May 15]. Available from: http://www.snomed.org/ External link
11.
ISO/PRF TS 21526. [Accessed 2019 May 15]. Available from: http://www.iso.org/cms/render/live/en/sites/isoorg/contents/data/standard/07/10/71041.html External link
12.
Storck M, Krumm R, Dugas M. ODMSummary: A Tool for Automatic Structured Comparison of Multiple Medical Forms Based on Semantic Annotation with the Unified Medical Language System. PLOS ONE. 2016 Oct 13;11(10):e0164569.