Format

Send to

Choose Destination
See comment in PubMed Commons below
Inform Prim Care. 2011;19(3):127-34.

Defining datasets and creating data dictionaries for quality improvement and research in chronic disease using routinely collected data: an ontology-driven approach.

Author information

1
Primary Care and Clinical Informatics, Department of Health Care Management and Policy, University of Surrey, Guildford, UK. s.lusignan@surrey.ac.uk

Abstract

BACKGROUND:

The burden of chronic disease is increasing, and research and quality improvement will be less effective if case finding strategies are suboptimal.

OBJECTIVE:

To describe an ontology-driven approach to case finding in chronic disease and how this approach can be used to create a data dictionary and make the codes used in case finding transparent.

METHOD:

A five-step process: (1) identifying a reference coding system or terminology; (2) using an ontology-driven approach to identify cases; (3) developing metadata that can be used to identify the extracted data; (4) mapping the extracted data to the reference terminology; and (5) creating the data dictionary.

RESULTS:

Hypertension is presented as an exemplar. A patient with hypertension can be represented by a range of codes including diagnostic, history and administrative. Metadata can link the coding system and data extraction queries to the correct data mapping and translation tool, which then maps it to the equivalent code in the reference terminology. The code extracted, the term, its domain and subdomain, and the name of the data extraction query can then be automatically grouped and published online as a readily searchable data dictionary. An exemplar online is: www.clininf.eu/qickd-data-dictionary.html

CONCLUSION:

Adopting an ontology-driven approach to case finding could improve the quality of disease registers and of research based on routine data. It would offer considerable advantages over using limited datasets to define cases. This approach should be considered by those involved in research and quality improvement projects which utilise routine data.

PMID:
22688221
[Indexed for MEDLINE]
Free full text
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for BCS Learning and Development
    Loading ...
    Support Center