Privacy-Preserving Record Grouping and Consent Management Based on a Public-Private Key Signature Scheme: Theoretical...

Area of Science:

• Medical Informatics
• Cryptography
• Clinical Trials

Background:

• Clinical and social trials generate vital evidence for medical advancement.
• Gathering patient data necessitates stringent security and privacy measures.
• Current pseudonymization methods, while obscuring direct links, present centralized security risks and potential for data-driven re-identification.

Purpose of the Study:

• To propose a novel, decentralized pseudonymization approach utilizing public-private key cryptography.
• To enable patient-driven creation and management of pseudonyms.
• To facilitate one-time pseudonymization per data record and data grouping without revealing the pseudonymization key.

Main Methods:

• Implemented a public-private key cryptography-based signing mechanism for patient data records.
• Detailed workflows for user registration, login, record storage, and record grouping.
• Evaluated performance, cryptographic collision risks, and conducted a threat analysis.

Main Results:

• Workflow execution times ranged from milliseconds to seconds, demonstrating efficiency.
• No realistic risk of cryptographic collision was identified.
• Compromising three distinct server systems was necessary for data access, with full re-identification requiring further breaches.

Conclusions:

• The proposed decentralized approach offers superior security and privacy compared to centralized methods.
• It eliminates the need for a trusted third party.
• A minor drawback is the inability to directly feedback incidental findings to participants due to quasi-anonymous data storage.