With the European Health Data Space (EHDS) expanding access to health data, linkage will become even more important.
The choice of linkage method depends on the available data and applicable governance arrangements, and every approach carries a risk of linkage errors, including missed and incorrect matches. However, correct linkage is necessary, not sufficient, for valid study conclusions. Even when records are correctly linked, decisions about study eligibility, the availability of linked data and the consistency of outcome ascertainment across sources can introduce selection bias or misclassification. These biases can affect the validity and interpretation of study findings.
Three linkage-related challenges include:
- Defining patient eligibility when linkage is required
- Declining eligible sample size as linkage requirements increase
- Ascertaining and comparing outcomes across sources
This is the first blog post in a five-part series on lessons from linked real-world data through the PHARMO Data Network.
Fill out the form to the right to read the full blog.
Fill out the form to read the full blog
"*" indicates required fields