A Component Based Solution for Managing
Cancer Care and Clinical Research
Udo Altmann, Thomas Bürkle, Ralf Schweiger, Ali Gerhard Tafazzoli, Frank Rüdiger Katz, Joachim Dudeck
Institute of Medical Informatics, Justus-Liebig-University, Heinrich -Buff- Ring 44, 35392 Giessen, Germany
Introduction
Gießen University Hospital is a hospital with 1334 regular beds and supplies medical care for appr. 40,000 inpatient and 300,000 outpatient cases per year.
At Gießen University Hospital several developments took place in the last years that will be integrated to a common environment for the management of patient care and research in cancer patients. The components that form that solution are:
GTDS (Giessen Tumor Documentation System)
The co-operation of the components is realized as follows. Cancer data can be entered into the registry system (GTDS) and displayed by various ways. The traditional one that is still necessary to be supported is the generic user interface of GTDS and is used mainly by registrars who collect the data from the patient files or from paper forms coming from inside the hospital as well as from outside (e.g., GPs). Other ways are communication interfaces, and mainly for displaying information up to now, an HTML interface, that in a first step, implements the following services
Arden Module
By means of database triggers potentially eligible patients can be detected based on the information covered by the minimum basic data set. The necessary database triggers and procedures are generated in an Arden Syntax environment. GTDS has a flexible scheduling system where patient contacts, that can take place anywhere, can be planned (as events of various types that are reflected as documents) and monitored. Another GTDS module stores administrative data of studies and clinical trials (addresses, eligibility criteria, enrolled patients). For demonstration purposes, automatic detection of potential eligibility has been realized for three NCI trials.
XML Based Browser Interface
The XML based browser interface is a general component that allows to describe a document's structure by an XML-file and store it as an XML- file on server side. GTDS has the capability to store and manage that files in relation to the events defined in the GTDS (diagnostic data, therapy data, follow-up, ... ). Data for which structures exist in the GTDS can be extracted from that XML-file and are available for registry purposes. When a user requests information of a cancer patient in the intranet, he first gets an overview of the patient's past and planned events / documents. He can select a document or add an additional document for viewing or entering data. The document is either a standard document if the patient is not enrolled to a study or an enhanced document adapted to the required additional data. GTDS unloads the document and, if it is a new document, fills it with data that are already known in the registry. The user completes the required data and sends it back to GTDS where it is loaded and analyzed. The advantages of this method is that the data remain stored in the way how the user saw and entered them but are also available for the registry. In addition, the user has a unique, non redundant way to enter data and does not have to take care of different purposes for which they are used. The work on this component is still in progress.
Dictionary Based Information Server
The third component, the Information Server GDDS, is a component which has been designed to supply heterogenous clinical applications with context sensitive information. Based on a medical data dictionary it is invoked with an infobutton inside the application which transmits a search term, e.g. a drug name or a cancer term. The Information Server links directly to related information sources and enables navigation between those different sources.
Since GTDS works with highly structured, coded data, it can provide well-defined terms for the search in such a dictionary. The addressed documents can be documents from CANCERNET as well as locally, regionally or nationally implemented guidelines or trial information.
We implemented the following services
An alternative way to access such sources in places, where such a server is not available is the use of the internal dictionary of GTDS.
Information on the Web
WWW:
http://www.med.uni-giessen.de/imiReferences:
U. Altmann, F. R. Katz, A. Tafazzoli, V. Haeberlin, J. Dudeck. GTDS - a Tool for Tumor Registries to Support Shared Patient Care. Proc AMIA Annu Fall Symp: 512-516 (1996)
A.G. Tafazzoli, U. Altmann, F.R. Kratz, J. Dudeck, S. Hölzer. Integrated Knowledge-Based Functions in a Clinical Cancer Registry - Specific Requirements for Routine Applicability. Proceedings of AMIA 1999 Fall Symposium
R. Schweiger, T. Bürkle, J. Dudeck, A.G. Tafazzoli, S. Hölzer. Migration from legacy data to XML - Experience with drug information sources at Giessen University. Proceedings of AMIA 1999 Fall Symposium
Schweiger R, Bürkle T, Ruan W, Dudeck J (1998) XML: Evolution towards a structured electronic patient record. In Moorman P W, van der Lei J, Musen M A: Preceedings of IMIA Working Group 17 EPRIMP conference Rotterdam ISBN 90-9012128-5, 264-267.
Bürkle T, Ruan W, Michel A, Dudeck J (1998) On the way to a web based hospital information system: Concepts for the use of a medical Data Dictionary to present context sensitive information in an intranet environment. In: Cesnik B. Mc Cray A.T. Scherrer J.-R. (eds) Medinfo 98 Proceedings of the 9th World Congress on Medical Informatics, IOS Press Amsterdam 917-921.
Ruan W., Bürkle, T., Dudeck J. (2000). An Object-Oriented Design for Automated Navigation of Semantic Networks inside A Medical Data Dictionary. Accepted for Artificial Intelligence in Medicine