CONTENTS

EDITORIAL

IN BRIEF

REVIEWS

TUTORIAL

VIEWS & POSITIONS

EVENTS DIARY

LETTERS

ANNOUNCEMENTS

Issue 2, September 1997

ISSN 1368-1591

A proposal for a global database of cases in clinical medicine

Internet-based, universally accessible databases of clinical cases are emerging. Medical documents on the Internet currently include thousands of digitized clinical photographs, images, and case descriptions. Computer algorithms and programs for searching these databases and hyperlinks between the important elements within them can be accomplished without any technologic impediments. The Internet, which connects distributed medical practitioners, has become sufficiently reliable for us to consider the feasibility of promoting a disciplined global medicine case database.

Asuccessful model for this inititative already exists. The Human Genome Project (via NCBI Entrez) hyperlinks gene sequences, MEDLINE searches, image databases, and peer reviewed journals from diverse institutions and is inclusive of individual and isolated researchers. It has rapidly defined chromosomal DNA sequences, integrating the work of thousands of investigators and insitutions. Many have proposed that it has provided a superior economic and inter-institutional schema for advancing science versus traditional publishing formats.

We have a similar opportunity to create well-structured and well-maintained data repositories and software tools for linking clinical medicine cases - using previously defined nomenclatures and controlled medical vocabularies. A concerted project on the part of the medical profession would allow users to navigate a global case database, built from the contributions of community practitioners as well as academic centers. Refined computer algorithms and programs for accessing abstracted and/or indexed medical literature have evolved, with search engines built into the US National Library of Medicine's MEDLINE service. Linguistic and natural language searching, automated retrieval methods, and medical concept matching can be applied to uncover information from relatively unstructured Internet clinical medicine databases.

This would permit clinicians worldwide to conduct searches that detect permutations of disease, matching on diagnostic parameters, clinical symptoms, and other taxonomies. Similar cases could be detected in a statistically rigorous manner. Case-based analyses would allow users to identify the economic utility of various therapies, and to developing new disease management protocols. Recording and sharing case experiences would allow clinicians to better assess so-called 'probablistic models' for disease outcomes and response to therapy, particularly in the dynamic fields of AIDS and oncology.

I propose that the medical community supports a distributed international case database, archiving publicly available medical case information. Any identifying data would need to be removed to preserve patient confidentiality. An accession number, annotations of case features, and nomenclature tags from controlled vocabularies such as grade of cancer, severity of illness, etc. would be used to identify cases for retrieval.

To implement the project it is proposed that teams of clinicians and librarians post case files to Internet in a disciplined way. Files might contain descriptive fields from a recognized nomenclature such as SNOMED. This nomenclature is used widely in the US for coding patient encounters (other systems are in use internationally). These descriptive fields could then be accessed according to the researchers' needs (such as limiting searches to specific disease presentations).

The potential impact that delivering up-to-date multimedia medical information to healthcare workers might have on medical decision-making is profound. A collaborative, inter-institutional effort to maintain files of identifying information on medical Internet documents is required.

Gary Malet DO
Informatics Fellow
Oregon Health Sciences University, USA

Society for the Internet in Medicine

Comments to: simq@cybertas.demon.co.uk

Copyright © 1997 Society for the Internet in Medicine. All rights reserved.
Date: September 29, 1997
Document URL: http://www.cybertas.demon.co.uk/simq/issue2/editorial.html