Research Papers
Data integration and knowledge discovery in biomedical databases. Reliable information from unreliable sources
Authors:
A Mitnitski ,
Dept of Medicine, Dalhousie University, Halifax, NS B3H 2Y9
Dept of Computer Science, Dalhousie University, Halifax, NS B3H 2Y9
A Mogilner,
Dept of Medicine, Dalhousie University, Halifax, NS B3H 2Y9
C MacKnight,
Dept of Medicine, Dalhousie University, Halifax, NS B3H 2Y9
K Rockwood
Dept of Medicine, Dalhousie University, Halifax, NS B3H 2Y9
Abstract
To better understand information about human health from databases we analyzed three datasets collected for different purposes in Canada: a biomedical database of older adults, a large population survey across all adult ages, and vital statistics. Redundancy in the variables was established, and this led us to derive a generalized (macroscopic state) variable, being a fitness/frailty index that reflects both individual and group health status. Evaluation of the relationship between fitness/frailty and the mortality rate revealed that the latter could be expressed in terms of variables generally available from any cross-sectional database. In practical terms, this means that the risk of mortality might readily be assessed from standard biomedical appraisals collected for other purposes.
How to Cite:
Mitnitski, A., Mogilner, A., MacKnight, C. and Rockwood, K., 2003. Data integration and knowledge discovery in biomedical databases. Reliable information from unreliable sources. Data Science Journal, 2, pp.25–34. DOI: http://doi.org/10.2481/dsj.2.25
Published on
12 Jan 2003.
Peer Reviewed
Downloads
