Federated Query of Electronic Medical Records
The Shared Health Research Information Network (SHRINE) is a web-based query tool that allows investigators to determine the aggregate number of patients seen at participating hospitals who meet clinical criteria of interest (currently demographics, diagnoses, medications, and selected laboratory values).
SHRINE brings value to investigators interested in:
- Identifying potential clinical trial cohorts for recruitment
- Generating new research hypotheses
- Planning research requiring large sample sizes
- Preparing grant applications.
The Harvard-affiliated institutions include Beth Israel Deaconess Medical Center, Boston Children’s Hospital, Brigham and Women’s Hospital, Dana-Farber Cancer Institute and Massachusetts General Hospital.
A national pilot composed of academic health centers from across the United States was successfully deployed in 2013. The SHRINE team works in close collaboration with the i2b2 and SMART teams. Other key collaborations include ACT, SCILHS and UC ReX. For more information, see the Collaborations section below.
Register for the 2015 Annual i2b2/SHRINE conference.
Please contact with any questions about this project.
Past PeopleBill Simons Director of Engineering and Technology Andy McMurry Informatics Team Lead and Architect Clint Gilbert Software Engineer Joanna Brownstein Director of Clinical Research Informatics Katia Zilber-Izhar Project Coordinator Keith Dwyer Software Engineer Michael Bushe Consulting Software Architect Seth Paine UI/UX Designer Shaun Kelly Senior Software Quality Assurance Analyst Tenille Johnson Lead Data Curator, The eagle-i Network Ty Coghlan Software Development Co-op Utkarsh Akhouri Software Quality Assurance Engineer Vijay Raghavan Senior Program Manager
CARRAnet (Childhood Arthritis and Rhematological Research Alliance), funded principally by NIAMS, is employing SHRINE to assemble the largest to date registry of pediatric patients with rheumatic disease across a 60 site network. In addition, CARRAnet is piloting the IRBrely reciprocal IRB reliance system for large multi-site clinical studies in collaboration with the Web Portal team. The CARRA Registry is an observational retrospective and prospective cohort study that enrolls children and adolescents with major rheumatic diseases followed at participating study sites.
The Accrual to Clinical Trials (ACT) project goal is to create a federated network of National Clinical and Translational Science Award (CTSA) Consortium institutions to significantly increase participant accrual to the nation’s highest priority clinical trials. SHRINE is the technical platform underpinning the ACT network.
i2b2 and HMS partnered to pilot a web-based software network that would allow participating Harvard hospitals to link their respective i2b2 instances for the sharing of obfuscated, aggregated counts of patients meeting selected criteria. The full up SHRINE operational network is now an integral part of the Harvard CTSA (Catalyst), went live to qualified users. The goal of i2b2 is to develop the science and the engineering required to enable the clinical investigators of academic medical centers to conduct clinical research that is informed by state-of-the-art genomics and biomedical informatics.
SHRINE is working across a number of healthcare centers throughout the U.S. to develop the SCILHS (pronounced “skills”) open-source platform. The SCHILHS platform will cover more than 8 million patients and enable clinician and patient participation in research.
The UC ReX Data Explorer enables search of 13.9 million de-identified patient records from the five UC biomedical centers with one query. Data Explorer includes de-identified data-sets extracted from each UC institution’s clinical data warehouse. Integrating Biology and the Bedside (i2b2) technology allows clinical records to be transformed into a common data representation and stored in dedicated data repositories at each institution. The data repository leverages SHRINE to send a query to the repository at each participating site and then display the count results returned from each.