Project Summary

This project seeks to develop a framework of distributed algorithms to efficiently synthesize evidence in large clinical research networks in order to better identify patient subpopulations at high risk of adverse events and use prediction models to better support clinical decision making for patients and stakeholders. Motivated from a collaboration with PEDSnet, a PCORI-funded national pediatric clinical research network, this project seeks to address the analytic needs of data integration with a focus on protecting patient privacy, communication efficiency, and accounting for between-site heterogeneity. Specifically, the project plans to develop a novel distributed learning framework for integrating evidence in clinical research networks, by studying risks of adverse events associated with pediatric Crohn’s disease (PCD).

Stakeholders, including patients, healthcare providers, and researchers, will be engaged to prioritize research questions throughout the investigation. To ensure the broader impact of the proposed research, this project will develop, validate, and evaluate methodology and software using electronic health records (EHR) data from PEDSnet, which contains EHR data from 6.5 million pediatric patients—8 percent of the total US children population—from eight of the national leading pediatric academic health centers that provides services to 3 million children per year. It includes multiple types of longitudinal clinical encounters for more than 15,000 children with Inflammatory Bowel Diseases (IBD) diagnosis from 2009 to 2017, with detailed clinical data, patient-reported outcome (PRO) data, and administrative data. By working closely with PEDSnet, this project will validate the proposed methods and disseminate them to PEDSnet consortium and PCORnet®.

The long-term objective of this research is twofold. One, it seeks to use data from PEDSnet to develop, validate, and evaluate the proposed distributed algorithms as a novel data integration strategy. Two, it aims to develop a concerted effort among patients, nurses, clinicians, care partners, informaticians, and statisticians toward a data integration strategy that is patient-centric and generalizable to different health conditions and distributed research networks. The development of the proposed methods within PEDSnet will not only provide timely information to better support clinical decision making for PCD patients and stakeholders, but also enable more efficient data integration for other research networks.

Project Information

Yong Chen, PhD, MA
University of Pennsylvania

Key Dates

39 months
June 2020
January 2024


Project Status
Award Type
Funding Opportunity Type
Research Priority Area


The state where the project originates, or where the primary institution or organization is located.

View Glossary
Last updated: January 23, 2021