RTU logo

Referent Tracking Literature

Introductory readings

[Ceusters & Smith 2005a] is the first paper that introduced the idea of Referent Tracking (RT) as a new paradigm for entry and retrieval of data in the Electronic Health Record (EHR). It contains an easy to read introduction to the sort of problems that arise when general terms taken from terminologies or ontologies (such as 'fracture' and 'left femur') are used in the EHR to refer to particular entities on the side of a patient such as the particular fracture that John suffered from in 2005 in his left femur. It explains how all ambiguities can be avoided by referring to such entities by means of identifiers instead of codes. A more thorough and formal discussion of the types of statements that are required as well as the technical infrastructure for real implementations can be found in [Ceusters & Smith 2005b].

RT follows a set of rigorous principles that are based on philosophical realism, a branch of ontology of which the foundations for applying it to biomedicine are discussed in [Smith et al. 2005]. Also important to read in order to understand what is wrong with the prevailing concept-based approach and how RT can contribute to a solution, are [Smith & Ceusters 2006a] which covers biomedical terminologies and ontologies in general, and [Smith & Ceusters 2005b] which focuses on the EHR. Key in all this is to keep constantly aware of the distinction between three levels: reality, our understanding of reality, and our representations of reality of which ontologies are artifacts dealing with what is general in reality, and patient records with what is specific [Smith et al. 2006]. RT is primarily involved with the latter.

Challenges being solved

Of course, RT presents some challenges of its own. One specific problem is how to represent phenomena commonly expressed by statements such as: "no history of diabetes", "hypertension ruled out", "absence of metastases in the lung", and "abortion was prevented". Such statements seem at first sight to present a problem for RT, since there are here no entities on the side of the patient to which unique identifiers can be assigned. For concept-based systems, this is not an issue since entities such as "diabetes" or "no diabetes" are both happily classified as "concepts", no further questions being asked.  We solved this problem by introducing the 'lacks' relation which, as the 'instance' relation, holds between particulars and universals, thereby remaining faithful to the principles of unqualified realism within an EHR regime based on the idea of faithfulness to clinical reality [Ceusters et al. 2006a].
We expanded this idea in [Ceusters et al. 2007a] by describing the lacks-relation at the level of universals as well.

Another challenge is keeping track of the different kinds of changes, reflecting for example: (1) changes in the underlying reality, either in a specific patient's condition or the world in general; (2) changes in our understanding; (3) reassessments of what is considered to be relevant for inclusion in a referent tracking database, or (4) encoding mistakes introduced during data entry. In [Ceusters & Smith 2006a] these issues are addressed from the perspective of versioning in ontologies and repositories. The method developed is then used in [Ceusters et al 2007b] to assess the history mechanism of SNOMED CT. We found that this mechanism would benefit from (1) an explicit representation of the provenance of a class; (2) the separation of the time-period during which a component is stated valid in SNOMED CT from the period it is (or has been) valid in reality, and (3) redesign of the historical relationships table to give users better assistance for recovery in case of introduced mistakes.
In [Ceusters 2006] the technique is used to assess the quality of ontologies involved in mapping efforts.

Since June 2007, we are tackling the issue of mistakes in a RTS. [Ceusters 2007] is the first publication in this area. It gives an introduction to the sorts of mistakes that may arise, and proposes a solution for how to deal with them.

Applications

RT has only recently been introduced and concrete applications are under way.  An overview of the efforts that have been conducted in 2006 to demonstrate the usefulness of referent tracking in different domains is given in [Ceusters & Smith 2007b]. The details can be found in papers that have been published earlier or are forthcoming. In [Ceusters & Smith 2006b] it is discussed how the paradigm can be used to make flow-chart types of decision support applications ready for the Semantic Web.  In [Ceusters & Smith 2006c] it is suggested to apply the RT principles to the Digital Object Identifier (DOI) system in the context of digital rights management. In [Ceusters & Smith 2007a], it is described how the principles of referent tracking are used to assess the quality of enterpirse ontologies and how they can lead to building better corporate memories.

Implementations

[Manzoor et al. 2007a] contains a description of how the Referent Tracking System (RTS) that is developed in the RTU implements the Referent Tracking paradigm. To put this system in practice, we propose an architecture based on middleware technology using web services. A preliminary analysis in the context of electronic health records has been caried out in collaboration with Medtuity Inc. In [Rudnicki et al. 2007a] we describe how data from an EHR application need to be decomposed in order to make them accord with the tenets of RT. We outline the ontological principles on which this decomposition is based.
In [Manzoor et al. 2007b], we describe the functional and technical requirements of such an approach and document our experiences with MedtuityEMR, an EHR system that stores patient data in XML

List of relevant papers

Background material

  • [Smith et al. 2005a] Smith B, Ceusters W, Klagges B, Koehler J, Kumar A, Lomax J, Mungall C, Neuhaus F, Rector A, Rosse C. Relations in biomedical ontologies, Genome Biology 2005, 6:R46.
  • [Smith & Ceusters 2005b] Smith B, Ceusters W. An Ontology-Based Methodology for the Migration of Biomedical Terminologies to Electronic Health Records. AMIA 2005, October 22-26, Washington DC;:669-673. (draft).
  • [Smith & Ceusters 2006a] Smith B, Ceusters W. Ontology as the Core Discipline of Biomedical Informatics: Legacies of the Past and Recommendations for the Future Direction of Research, forthcoming in Gordana Dodig Crnkovic and Susan Stuart (eds.) Computing, Philosophy, And Cognitive Science, Cambridge: Cambridge Scholars Press, 2006. (full paper).
  • [Smith et al. 2006] Smith B, Kusnierczyk W, Schober D, Ceusters W. Towards a Reference Terminology for Ontology Research and Development in the Biomedical Domain. Proceedings of KR-MED 2006, Biomedical Ontology in Action, November 8, 2006, Baltimore MD, USA (draft)

Theoretical aspects of referent tracking

Applying referent tracking in specific contexts

Implementation

  • [Rudnicki et al. 2007a] Rudnicki R, Ceusters W, Manzoor S, Smith B. What Particulars are Referred to in EHR Data? A Case Study in Integrating Referent Tracking into an Electronic Health Record Application. In Teich JM, Suermondt J, Hripcsak C. (eds.), American Medical Informatics Association 2007 Annual Symposium Proceedings, Biomedical and Health Informatics: From Foundations to Applications to Policy, Chicago IL, 2007;:630-634. (abstract, draft)
  • [Manzoor et al. 2007b] Manzoor S, Ceusters W, Rudnicki R. A Middleware Approach to Integrate Referent Tracking in EHR Systems. In Teich JM, Suermondt J, Hripcsak C. (eds.), American Medical Informatics Association 2007 Annual Symposium Proceedings, Biomedical and Health Informatics: From Foundations to Applications to Policy, Chicago IL, 2007;:503-507.(abstract, draft)
  • [Manzoor et al. 2007a] Manzoor S, Ceusters W, Rudnicki R. Implementation of a Referent Tracking System. International Journal of Healthcare Information Systems and Informatics 2007;2(4):41-58. (summary, full paper).


701 Ellicott Street ~ Buffalo, New York 14203 ~ 716.898.8630privacy policy | contact us