Enhancing Consent Policy Privacy in Blockchain-Based Healthcare Data Sharing

Blockchain-based consent management makes healthcare data sharing auditable, but public chain enforcement can reveal the consent policies. Their conditions, such as disease-specific research restrictions, can disclose sensitive information about the data subject. To address this, we propose the Trusted Proof Service (TPS)-assisted Zero-Knowledge Policy Interpreter (TPS-ZKPI), in which the TPS generates zero-knowledge proofs of access decisions that a smart contract verifies against committed policies. We compile each policy into a fixed-length program for a reusable zero-knowledge policy interpreter, so all policies share one verification interface and a fixed-format public chain transcript. We evaluate TPS-ZKPI on Polygon Amoy with supervised and unsupervised attacks by a passive observer inferring hidden policy properties from public chain metadata. The results show no leakage in any of the 44 supervised tasks under a criterion requiring a mean AUC above 0.55 and significance in at least half of 30 splits, and unsupervised clustering recovers no policy structure beyond the chance level. Overall, our empirical results show that TPS-ZKPI resists the modeled metadata attacks and provides auditable consent enforcement on a public blockchain.