AEGiS-14IAC: A unified approach to data quality and accuracy especially with respect to data mining for the epidemic disease.

14th International AIDS Conference

Barcelona, Spain - July 7-12, 2002

A unified approach to data quality and accuracy especially with respect to data mining for the epidemic disease.

Int Conf AIDS 2002 Jul 7-12; 14:(abstract no. G12541)

Vararuk A
Department of Computation University of Manchester Institute of Science and Technology, Manchester, United Kingdom

ISSUES: The need for a common approach to data capture and analysis.

DESCRIPTION: While viewing data for my current research project I am finding that there are different standards in how data is captured - Locally, Nationally and Internationally. There are common items that are obviously required, and these are captured as a matter of course. Then the details will vary depending on the treatments that are applied to the HIV/AIDS patients and what is required from various reporting agencies above. So some cases will be reported with all the details required for a data-mining project to provide significant benefits. While other cases will have 'key' items omitted, as they are not relevant to the reporting authorities at the time, but within the larger view they would be immensely useful.

ISSUES: A standard approach is needed for the capture and storage of 'key' data, and when captured ensuring that it is always stored electronically. There is no common data interchange format to allow researchers to perform comparisons with other groups or request data from different regions without having to perform some data conversion.

RECOMMENDATIONS: I am proposing two things: 1st A feedback mechanism from data analysis groups to allow newly identified 'key' items to be recorded as the need is determined. 2nd An XML based approach to data interchange that is agreed at an International Level - and can be expanded at a National level as the needs are determined. This will allow the quality and accuracy of data to be expanded greatly whilst allowing researchers to request data (in a suitably controlled manner) whilst not having to 'waste' time performing conversions between database systems / methodologies. Adopting this just within the realms of HIV/AIDS research will provide many benefits, and should be extended to other areas of medical research as this is proven.

Keywords: AEGIS, Research Design, Acquired Immunodeficiency Syndrome, Software, Research, Biomedical Research, Databases, Human, statistics & numerical data, methods

020707
G12541