Privacy Methodology Reduced the Risk of Re-identification and Improved Data Utility
Featured Solution: “Clinical Data Sharing: A Proposed Methodology to Enable Data Privacy While Improving Secondary Use” (Privacy Methodology)
Secondary use of clinical trial data honors a patient’s contribution to clinical development by using data to the fullest extent possible while respecting the wishes of the individual patient, enhancing innovative quality advancements across drug development and trial design. One of the challenges for clinical data reuse is retaining data utility to support complex cross-study analysis when applying anonymization techniques to prevent the re-identification of patents. TransCelerate member companies assembled a team of privacy subject matter experts to develop a proposed Privacy Methodology that would improve data utility and promote data reuse. The proposed approach was improved by a round of public comment open to all interested stakeholders.
A sponsor company asked its anonymization vendor to apply the TransCelerate Privacy Methodology to the same clinical trial dataset previously anonymized using the sponsor company’s rule-based approach. The dataset consisted of 69 participants across 32 study sites.
According to the vendor’s analysis, applying the Privacy Methodology’s risk-based approach resulted in an average risk of re-identification score of 0.004. This score of 0.004 demonstrates that the risk-based Privacy Methodology reduces the risk of re-identification compared to the company’s current methodology. Furthermore, according to the vendor, the Privacy Methodology risk-based approach avoided “over” anonymizing key data variables, such as age and race, and resulted in improved data utility over the sponsor’s normal approach. The sponsor’s confidence in applying this approach was boosted by the fact that the updated version of the risk-based privacy methodology reflected a detailed review by global privacy, legal, and regulatory data-sharing experts.
In conclusion, according to the vendor analysis, the Privacy Methodology’s risk-based approach avoided over anonymization while reducing the risk of re-identification and improving data utility. Greater data utility today, may in the future, support data reuse for cross-study analysis, deepen insights in cross-population disease profiles, and reduce the patient burden.
[Sponsor Company]
