|
|
Methods and tools for the development of computer-interpretable guidelines |
|
| HGML |
|
Hypertext Guideline Markup Language |
|
| keywords |
XML, XHTML, markup, clinical practice guidelines,
clinical decision support, knowledge representation, knowledge acquisition
|
| developed by |
Clinical Informatics Research Group, University of Medicine and Dentistry of New Jersey; Department of Computer Science
Rutgers University |
| introduced |
2000 (at AMIA Annual Symposium) |
| status |
Under development |
| support |
|
| in use |
|
| tools |
Information available on request on Guideline Evaluation Engine, Mozilla-based Guideline Editor, MS-based Guideline Editor.
|
|
| description |
HGML - Hypertext Guideline Markup Language - is an XML/XHTML specification for the identification of condition and
recommendation elements within text guidelines.
HGML defines tags for conditional and associated recommendation elements which can be identified in a guideline text using
a markup editor. The correlation of conditional variables, subject to their constraints, to clinical data allow delivery
of recommendations linked to their original context.
A guideline editor - an enhancement of the Mozilla Web Browser's "Composer" feature - has been developed to support guideline markup.
An associated Guideline Evaluation Engine allows
information about a patient (either from an Electronic Medical Record system, or directly from a
clinician) to be collected, and relevant recommendations to be retruned to a web browser.
The output from the system is linked to the related
text in the original guideline, and to supporting evidence including references, related text, and survival curves from
models evaluated by Decision Maker, a companion project developed and maintained by members of the HGML development group.
|
|
| plans |
Implementation of HEDIS guidelines including Hypertension (JNC_VII), Cholesterol,
Stroke, Immunization.
Refinement of the HGML elements for increased flexibility and scope, making it easier it identify critical guideline
components and their relationships
Enhancement of markup editor
Implementation of a guideline server, integrated with a prototype point-of-care delivery system, in which guideline
recommendations relevant to a specific patient can be delivered (integrated with the Logician medical record system)
Exploration of Semi-Automatic means for encoding clinical guidelines, using UMLS/MetaMap to identify clinical terms in
the guidelines and infer possible encodings to assist in the markup process.
Integration with a medical record system using an
"infobutton"/web-based interface for a prototype Continuing Medical
Education application.
|
|
| references |
| Hagerty CG, Pickens D, Kulikowski C, Sonnenberg F.
HGML: a hypertext guideline markup language.
Proc AMIA Symp. 2000;:325-9.
[PubMed]
[AMIA]
|
"
Existing text-based clinical practice guidelines can be difficult to put into practice. While a growing number of such documents have gained acceptance in the medical community and contain a wealth of valuable information, the time required to digest them is substantial. Yet the expressive power, subtlety and flexibility of natural language pose challenges when designing computer tools that will help in their application. At the same time, formal computer languages typically lack such expressiveness and the effort required to translate existing documents into these languages may be costly. We propose a method based on the mark-up concept for converting text-based clinical guidelines into a machine-operable form. This allows existing guidelines to be manipulated by machine, and viewed in different formats at various levels of detail according to the needs of the practitioner, while preserving their originally published form.
" |
|
Hagerty CG, Chang J, Pickens DS, Kulikowski CA, and Sonnenberg FA. Semi-Automated Encoding of Guidelines. Medinfo. 2004;2004(CD):1625.
[]
[]
|
"
The encoding of clinical guidelines can be an arduous task
requiring human expertise that presents a bottleneck on the
road to operability. We have developed a system that enhances
text-based guidelines by annotating them with semantic
tags that identify terms with their associated Unified Medical
Language System. (UMLS.) concepts using the National
Library of Medicine’s MetaMap utility. Additional annotations
are then made, employing information extraction techniques
to infer and identify the specific conditions and recommendations
required to encode the guideline. Feedback
from the user in this framework can allow machine learning
techniques to improve these inferences.
" |
|
|
| contact |
C. Greg Hagerty and/or Frank A. Sonnenberg
University of Medicine and Dentistry of New Jersey
Robert Wood Johnson Medical School
Clinical Academic Building
125 Paterson Street
New Brunswick, New Jersey, 08903 USA
E: cgreg cgreg.com
sonnenbeumdnj.edu
|
| links |
|
|
| acknowledgements |
| C. Greg Hagerty, University of Medicine and Dentistry of New Jersey |
| page history |
Entry on OpenClinical: 06 May 2005
Last main update: 06 May 2005 |
|
|
|
|
HGML at University of Medicine and Dentistry of New Jersey