Ontology-based Trial Management Application


Overview

ObTiMA is an ontology-based clinical trial management system intended to support clinicians in both designing and conducting clinical trials. The design phase is facilitated by the Trial Builder in which all aspects of a clinical trial can be specified: A trial chairman can define the outline and metadata of a trial in a master protocol to describe, e.g., trial goals or administrative data. He can further setup treatment plans for guiding clinicians through individual patient treatment where events, e.g., surgery or chemotherapy, can be defined with all necessary information. In addition, the particular treatment order can be freely setup on a timeline and so can treatment stratifications and randomizations to be applied for a patient. A Case Report Form (CRF) can be assigned to each treatment step to collect documentation data.

The ontology-based creation of CRFs in the Trial Builder is one of ObTiMA's major functionalities. A graphical user interface allows defining content, navigation, and layout of CRFs to capture all patient data during a trial, e.g., medical findings or diagnostic data. The resulting descriptions are based on concepts from the Health Data Ontology Trunk (HDOT) for each CRF item along with metadata, e.g., data type and measurement unit, and are used to setup the trial database.

The user interface makes the underlying aspects of the ontological metadata transparent to users and tries to close the gap between clinical practice and the actual logical representation of ontological concepts. (Even if natural language descriptions are given for concepts, they rarely reflect the needs of clinical perception of reality. Therefore an application-specific, simplified view of the ontology is given showing only its relevant portions in a clinician-friendly way.) If an item has been created based on a concept, its attributes are determined automatically, e.g., label, data type or answer possibilities, but it can be adopted manually.

Since many trials collect similar or equal data, it is possible to store components of or complete CRFs in a repository as templates. When setting-up a clinical trial, appropriate CRF templates can either be directly reused or can be quickly created by composing them from existing CRF components. This in turn fosters the CRF standardization since CRFs can then readily be compared on the level of single items (through ontological concepts) and also on component level or in their entirety.

The second major functionality is the patient data management system supporting clinicians during a clinical trial. It is automatically setup based on the items defined in the Trial Builder in the design phase. It guides the clinicians through the treatment of the individual patients according to the given treatment plans and provides an easy user interface to fill in the CRFs for a patient (again hiding the actual underlying ontology concepts from the clinician). When the PDMS is set-up, the trial database is automatically derived from the ontology-based CRF definitions. Thus, provided that appropriate rights are given, the database can then also be accessed by other trials or applications through using a semantic mediation service based on the ontology.

ObTiMA itself is composed of different modules. In addition to the above described basic components, a DICOM server and DICOM viewer, a SAE and SUSAR reporting tool and a consultation tool are integrated. These tools are optional to handle images used in clinical trials or to simplify the SAE and SUSAR reporting according to GCP criteria. The consultation tool stores all consultations in a standardized way in the trial database. ObTiMA itself fulfils GCP criteria, including an Audit Trail. Data safety and security are guaranteed as pseudonymization of private data is implemented according to roles and rights assigned to users of ObTiMA. It is of utmost importance that ObTiMA and TOB will be certified for the use within clinical trials.


Publications

Holger Stenzhorn, Gabriele Weiler, Mathias Brochhausen, Fatima Schera, Vangelis Kritsotakis, Manolis Tsiknakis, Stephan Kiefer and Norbert Graf
The ObTiMA System - Ontology-based Managing of Clinical Trials
in Proceedings of the 13th World Congress on Health (Medical) Informatics (Medinfo 2010), Cape Town, South Africa, September 2010

Micke Kuwahara, Yuzuru Tanaka, Jonas Sjöbergh and Norbert Graf
Trial Outline Builder - Versatile Software Tool for Building, Managing and Analyzing Clinical Trials
in Proceedings of the 8th HealthGrid Conference, Paris, France, 2010

Mathias Brochhausen, Gabriele Weiler, Fatima Schera, Jochen Rauch, Norbert Graf and Stephan Kiefer
Ontology-based Trial Management System (ObTiMA)
in Nature Preceedings http://dx.doi.org/10.1038/npre.2009.3753.1, 2009

Mathias Brochhausen, Gabriele Weiler, Luis Martín, Cristian Cocos, Holger Stenzhorn, Norbert Graf and Martin Dörr
Applications of the ACGT Master Ontology on Cancer
in Proceedings of the 4th International IFIP Workshop On Semantic Web & Web Semantics (OTM 2008: Workshops), LNCS 5333, 2008, 1046–1055.

Norbert Graf, Gabriele Weiler, Mathias Brochhausen, Fatima Schera, Alexander Hoppe, Manolis Tsiknakis and Stephan Kiefer
The Importance of an Ontology Based Clinical Data Management System (OCDMS) for Clinico-Genomic Trials in ACGT (Advancing Clinico-Genomic Trials on Cancer)
in Proceedings of the SIOP Conference, Mumbai, India, November 2007

Gabriele Weiler, Mathias Brochhausen, Norbert Graf, Alexander Hoppe, Fatima Schera and Stephan Kiefer
Ontology-based Data Management Systems for Post-genomic Clinical Trials within an European Grid Infrastructure for Cancer Research
in Proceedings of the 29th Annual International Conference of the IEEE EMBS, Lyon, France, August, 2007