Probabilistic linkage without personal information successfully linked national clinical datasets

Area of Science:

• Health Informatics
• Data Linkage
• Public Health Research

Background:

• Linking patient data across clinical databases is crucial for public health but often hindered by privacy concerns.
• Accurate data linkage can accelerate research and improve understanding of clinical and public health issues.
• Probabilistic linkage offers a method to connect datasets without compromising personal information.

Purpose of the Study:

• To develop and validate a probabilistic linkage method for national clinical and administrative datasets.
• To assess the effectiveness of probabilistic linkage compared to deterministic linkage.
• To evaluate the impact of linkage methodology on patient data analysis.

Main Methods:

• A step-by-step probabilistic linkage process was developed.
• The method was validated against deterministic linkage using patient identifiers.
• Electronic health records from the National Bowel Cancer Audit and Hospital Episode Statistics were used for 10,566 patients.

Main Results:

• Probabilistic linkage achieved an 81.4% match rate between datasets.
• Deterministic linkage achieved an 82.8% match rate.
• No systematic differences were observed between linked and unlinked patients, and analyses were not sensitive to the linkage method.

Conclusions:

• Probabilistic linkage is a successful method for connecting national clinical and administrative datasets for surgical patients.
• This approach enables linkage outside secure environments, reducing costs and delays.
• Probabilistic linkage effectively protects data security while maintaining high linkage quality.