Privacy-protecting analytic and data-sharing methods

 

Privacy-Preserving Distributed Analysis of Biomedical Big Data

This project will develop an open-source software tool to perform distributed regression in vertically partitioned data, a data environment where information about an individual is available in two or more data sources. The method does not require data to be combined physically, but produces statistically equivalent results as if the datasets were linked and pooled centrally. Instead of sharing patient-level information, participating sites will only transfer non-identifiable information matrix and other summary statistics needed in the statistical modeling process. This approach offers much greater protection for data privacy while allowing one to perform sophisticated statistical analysis.


Principal Investigator: Darren Toh, ScD

Funder: National Institute of Biomedical Imaging and Bioengineering (U01EB023683)

Related Links: Privacy-Protecting Distributed Analysis of Biomedical Big Data

 

 

Privacy-Preserving Analytic and Data-Sharing Methods for Clinical and Patient-Powered Data Networks

During this project, we will partner with stakeholders from a variety of perspectives, including patient groups, health system leaders, multicenter research governance experts, regulatory, compliance, and confidentiality board members, and researchers. The project will assess stakeholders’ understanding of summary score-based privacy-preserving methods, and the benefits and limitations of using them in multisite patient-centered outcome research (PCOR) studies. We will also develop a suite of analytic tools to perform rigorous PCOR analysis without sharing potentially identifiable data and will create freely available analytic tools and educational materials for these methods.


Principal Investigator: Darren Toh, ScD

Funder: Patient-Centered Outcomes Research Institute (ME-1403-11305)

Related Links: PCORI Privacy-Preserving Methods

 

 

Utilizing Data from Various Data Partners in a Distributed Manner

This project will develop a new capability within PopMedNetTM (PMN) to enable secure, automatable distributed regression analysis in horizontally partitioned data, a data environment where information from different people are held at different institutions. Distributed regression analysis allows sites to maintain control of their patient-level data while generating valid, statistically equivalent regression estimates as if the data are pooled centrally.


Principal Investigator: Darren Toh, ScD

Funder: Office of the Assistant Secretary for Planning and Evaluation (ASPE) and Food and Drug Administration (HHSF22301006T)

Related Links: The OS PCORTF-Funded Projects