Abstract
OBJECTIVES--To report on the development of computer assisted methods for linking medical records and record abstracts. DESIGN--The methods include file blocking, to put records in an order which makes searching efficient; matching, which is the process of comparing records to determine whether they do or do not relate to the same person; linkage, which is the process of assembling correctly matched records into a time sequenced composite record for the individual; and validation checks and corrections, in which any inconsistencies between different records for the same person are identified and corrected. SETTING--The dataset comprising the Oxford record linkage study which includes hospital inpatient records and vital records. RESULTS AND CONCLUSIONS--Probability matching, using an array of identifiers, achieves much higher levels of correct matching than is generally achievable by exact character by character comparisons. The increasing use of information technology to store data about health and health care means that there is increasing scope to link records for research and for patient care. Sophisticated methods to achieve this on a large scale are now available.
Full text
PDF
Selected References
These references are in PubMed. This may not be the complete list of references from this article.
- Baldwin J. A., Gill L. E. The district number: a comparative test of some record matching methods. Community Med. 1982 Nov;4(4):265–275. doi: 10.1007/BF02548597. [DOI] [PubMed] [Google Scholar]
- Dunn H. L. Record Linkage. Am J Public Health Nations Health. 1946 Dec;36(12):1412–1416. [PMC free article] [PubMed] [Google Scholar]
- NEWCOMBE H. B., KENNEDY J. M., AXFORD S. J., JAMES A. P. Automatic linkage of vital records. Science. 1959 Oct 16;130(3381):954–959. doi: 10.1126/science.130.3381.954. [DOI] [PubMed] [Google Scholar]
- Newcombe H. B. Record linking: the design of efficient systems for linking records into individual and family histories. Am J Hum Genet. 1967 May;19(3 Pt 1):335–359. [PMC free article] [PubMed] [Google Scholar]