gms | German Medical Science

GMDS 2014: 59. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

07. - 10.09.2014, Göttingen

Metadata-driven Assisted Creation of Medical Forms

Meeting Abstract

Suche in Medline nach

  • D. Alexandre - Universitätsmedizin Mainz, Mainz
  • M. Lablans - Universitätsmedizin Mainz, Mainz
  • F. Ückert - Universitätsmedizin Mainz, Mainz

GMDS 2014. 59. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Göttingen, 07.-10.09.2014. Düsseldorf: German Medical Science GMS Publishing House; 2014. DocAbstr. 362

doi: 10.3205/14gmds127, urn:nbn:de:0183-14gmds1277

Veröffentlicht: 4. September 2014

© 2014 Alexandre et al.
Dieser Artikel ist ein Open Access-Artikel und steht unter den Creative Commons Lizenzbedingungen (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.de). Er darf vervielfältigt, verbreitet und öffentlich zugänglich gemacht werden, vorausgesetzt dass Autor und Quelle genannt werden.


Gliederung

Text

Introduction and objectives: Software developers working on clinical information systems deal daily with both heterogeneous data structures and the creation of complex user interfaces. Developed software solutions overlap each other’s features and work differently although they deal with equivalent data and processes [1]. A metadata repository enables data harmonization and also the reutilization of software components, as shown in this document.

The project described herein aims at easing the creation of clinical forms. Based on metadata, available remotely on a repository, the developed software library enables the automatic rendering of labels, tooltips, input widgets and data validation. It includes easy-to-set reusable components that software developers can use to render clinical forms, relying on already defined metadata rather than re-inventing data elements.

Materials and methods: Java is the chosen programming language for the project due to the cross-platform capabilities, objected-orientation, use simplicity, scalability, large set of API and security features. JavaServer Faces (JSF), a specification for building component-based web user interfaces is a good example of Java extensibility. JSF enables reusability of web interface components and relieves the user interface building process.

Along with JSF, standard web technologies such as HTML, CSS, Javascript, Bootstrap and jQuery are used. A web-based solution was chosen due to the deployment advantages and modern web browsers potential [2].

A metadata repository, holding clinical data element definitions, types, validations and permissible values was used. The web user interface modules communicate with metadata repositories though REST.

Results: A Java library that includes JSF components has been developed to ease the process of creating clinical forms through automatisation and user interface components reusability. Developers can get started by adding the library to their own JSF project and reusing the pre-built components. Each data element defined in the metadata repository has a unique identifier, which is used to specify the desired visual component - the library automatically renders the appropriate widget for the label and necessary input(s). The input types can range from a text box and a data selector to a complex pre-defined selection of values that can trigger other inputs.

The data structure is defined remotely in a metadata repository from which the component library automatically gets the necessary data to resolve the appropriate widget. As a result, JSF rich clinical forms can be quickly built and the developer does not have to fiddle with the clinical data structure definition, complexity, validation and usability issues. Domain experts are able to focus on the data structure defined in a metadata repository and developers reuse this knowledge simply through the identification of the desired data.

The software system is separated in independent modules, thus enabling the faster development of each part and inciting the reusability the modules according to project specific requirements. As an example, the metadata can be further developed without interfering with the development and release of a web clinical form creation tool connected to it.

In fact, the developed JSF component library reuses another module – a connector to a metadata repository that enables querying for data structure, definition and data validation through REST calls. This connector is useful as an independent module when, for instance, a developer needs to obtain repository data for a project that does not include the JavaServer Faces specification (e.g. to obtain a list of the data elements available). The metadata repository connector supports caching, which enables faster data access. It queries the metadata repository on the fly, but avoids unnecessary calls. Caching is particularly relevant when the JSF component library queries for metadata to resolve the appropriate widget and to validate the inputs of a form with a considerable number of data elements, as the number of REST calls are drastically reduced. Besides, with caching the metadata is still accessible locally even if when metadata repository becomes unreachable.

Both the connector module and the JSF components library are multilingual, providing complete widgets to be rendered in a selected language, considering labels, tooltips, data options and validations. Choosing another language has an immediate automatic effect on the form and it is also supported by the caching mechanism, as multiple languages can be cached. Furthermore, form inputs are frequently grouped in sections according to its semantic relationship in the context of each project. The developed modules embrace this featuring by being able to query metadata repositories groups and subgroups of data elements.

Discussion: The created JSF components library enables the quick development of complex clinical forms while abstracting from the complexity of data structuring, widget design and input validation. Unlike other tools that help creating forms, this one connects to a metadata repository specialised in a domain (in this case clinical data). By simply setting the desired data elements from a metadata repository, a developer can build a form that automatically renders the adequate sections, labels, input widgets and validates the inputs.

Nevertheless, the described system relies on a metadata repository. The completeness, correctness and accessibility of the metadata are required for the quality and usability of the forms. A central clinical metadata repository was developed and used for the development of the library [3].

The software solution can be further developed to offer the possibility for non-computer-science-specialists to create clinical forms, i.e. without requiring code development. A web user interface, including the metadata repository connector, would present a searchable list of the available data elements. By means of a usable selection method, such as drag and drop or input box that suggests values as the user types, the desired data elements could be selected. The JSF component library then enables the immediate visualisation and interaction with the form. The end user would therefore be able to create complex clinical forms through the selection of data items in a web user interface.

Additionally, the potential of the project would increase with the possibility to save, export and import generated forms. There could be a form repository in which each defined form would have the ordered reference of the included data elements.


References

1.
Eder J, Dabringer C, Schicho M, Stark, K. Information Systems for Federated Biobanks. In: Hameurlain A, Küng J, Wagner R, editors. Transactions on Large-Scale Data- and Knowledge-Centered Systems I. Berlin, Heidelberg: Springer; 2009. p. 156-90. (Lecture Notes in Computer Science; 5740). DOI: 10.1007/978-3-642-03722-1_7 Externer Link
2.
Nadkarni PM, Brandt CM, Marenco L. WebEAV: automatic metadata-driven generation of web interfaces to entity-attribute-value databases. J Am Med Inform Assoc. 2000 Jul-Aug;7(4):343-56.
3.
Lablans M. Konzept der CCP-IT des DKTK. Version 1.4.1. 30.01.2014 . Verfügbar unter: https://www.unimedizin-mainz.de/fileadmin/kliniken/imbei/Dokumente/MI/AGVF/Konzept_CCP-IT.pdf Externer Link