Related Experiment Videos
[Linking of individual data. Methods of linkage]
1Unité 149 de l'INSERM, Paris.
Revue D'Epidemiologie Et De Sante Publique
|June 1, 1997
Summary
This study overviews personal data linking methods for health statistics, detailing record linkage steps and probability matching techniques. It highlights benefits for epidemiological studies while addressing ethical considerations.
Area of Science:
- Data Science
- Biostatistics
- Public Health
Context:
- Linking disparate personal data sources like administrative and medical records is crucial for comprehensive analysis.
- Existing methods for personal data linkage require careful consideration of identifiants and matching procedures.
Purpose:
- To provide an overview of personal data linkage methods.
- To describe the steps involved in record linkage, including identifiant selection and pair comparison.
- To discuss probability matching and optimization techniques for record linkage.
Summary:
- The paper details the process of linking personal data from various files, such as registers and administrative or medical records.
- It examines epidemiological and demographic examples, outlining key steps like identifiant definition, searching, and comparing records.
- Special attention is given to probability matching and methods for optimizing linkage procedures.
Impact:
- Enhanced record linkage improves the utilization of health statistics.
- Enables novel types of epidemiological studies through integrated data.
- Addresses ethical challenges and emphasizes the need for robust evaluation of linkage methods.