Distributed regression analysis (DRA) is a privacy-protecting analytical method that performs multivariable regression analysis in multi-center studies with sharing only summary-level data. DRA has been demonstrated to obtain precise results as compared to the results obtained from the pooled individual-level data analysis. We have implemented new functionalities in PopMedNetTM, an open-source file-sharing software, to facilitate the execution of vertical DRA. This presentation will describe the workflow, user interfaces, and challenges of vertical DRA.

Learning Objective: After participation in this session, the learner should be able to:
1. Describe horizontal and vertically partitioned data
2. Describe the challenges of distributed regression analysis with vertically partitioned data
3. Describe distributed regression analysis with PopMedNet


Qoua Her (Presenter)
Harvard Pilgrim Health Care Institute

Yuji Samizo, Pennsylvania State University
Thomas Kent, Pennsylvania State University
Aleksandra Slavkovic, Pennsylvania State University
Mia Gallagher, Harvard Pilgrim Health Care Institute
Darren Toh, Harvard Pilgrim Health Care Institute

Presentation Materials: