Data linkage in pharmacoepidemiology: A call for rigorous evaluation and reporting.

Area of Science:

• Pharmacoepidemiology
• Health Data Science
• Observational Research

Background:

• Data linkage is crucial for enriching health databases, improving population definitions, and capturing outcomes.
• Evaluating data linkage quality is essential for reliable pharmacoepidemiologic studies.
• Existing guidelines like RECORD-PE provide a foundation for reporting.

Framework:

• A new checklist is developed for evaluating data linkage performance in pharmacoepidemiology.
• The checklist covers key domains: data sources, linkage variables, methods, results, and evaluation.
• It aims to standardize the reporting of data linkage processes.

Implementation:

• Developed by an international working group representing academia, industry, and regulatory sectors.
• The checklist expands upon the RECORD-PE guidelines.
• It provides specific recommendations for clear and transparent assessment.

Implications:

• Enhances the transparency and reproducibility of pharmacoepidemiologic research.
• Facilitates critical assessment of potential bias introduced by data linkage.
• Increases confidence in evidence derived from linked health data.