OHDSI

OHDSI The Observational Health Data Sciences and Informatics initiative is a multi-stakeholder collaborative network that advances large-scale observational research through standardized data, open-source software, and community-driven methods. It connects clinical research programs, regulatory agencies, academic centers, and industry partners to support reproducible analyses, evidence generation, and comparative effectiveness studies across heterogeneous data sources.

Overview

OHDSI operates as a distributed research network integrating clinical terminologies, data warehouses, and analytic platforms to enable observational research at scale. Major participants include National Institutes of Health, Food and Drug Administration, European Medicines Agency, Centers for Medicare and Medicaid Services, and academic centers such as Harvard University, Stanford University, Johns Hopkins University, Columbia University, and University of Oxford. Industry collaborators range from Pfizer, Roche, Johnson & Johnson, AstraZeneca, to IQVIA and Epic Systems, while nonprofit and philanthropic supporters include Bill & Melinda Gates Foundation, Robert Wood Johnson Foundation, and Wellcome Trust.

History and Development

OHDSI traces its conceptual lineage to earlier efforts in clinical data standardization and distributed research networks, building on projects like Mini-Sentinel, Sentinel Initiative, PCORnet, i2b2, and OMOP workstreams initiated at Columbia University, Erasmus University Medical Center, and University of Washington. Key milestones involved collaborations with regulatory reviews at European Commission panels, methodologic workshops at International Society for Pharmacoeconomics and Outcomes Research, and demonstration studies presented at American Medical Informatics Association conferences, leading to rapid expansion across United States, United Kingdom, Netherlands, Canada, Japan, Australia, and South Korea research sites.

Organizational Structure and Governance

Governance in OHDSI comprises steering committees, working groups, and community councils that include representatives from Massachusetts General Hospital, Mayo Clinic, Vanderbilt University Medical Center, University of California, San Francisco, and industry partners such as GlaxoSmithKline and Novartis. Advisory input has come from stakeholders associated with World Health Organization, European Centre for Disease Prevention and Control, and national health systems like NHS England and Veterans Health Administration. Formal decision-making processes are influenced by consensus-driven working groups and methodological task forces modeled after consortiums like Human Genome Project and Global Alliance for Genomics and Health.

Common Data Model and Standards

A foundational element is the Common Data Model (CDM) adapted from Observational Medical Outcomes Partnership efforts and informed by vocabularies including SNOMED CT, LOINC, RxNorm, ICD-9, ICD-10, and ATC classifications. The CDM enables mapping across electronic health record sources such as Cerner, Epic Systems, claims databases like Medicare and MarketScan, and registries affiliated with European Medicines Agency and national cancer registries. Data governance and privacy practices align with regulations and frameworks including Health Insurance Portability and Accountability Act, General Data Protection Regulation, and institutional review mechanisms at universities like Yale University and University of Pennsylvania.

Tools and Technologies

The software ecosystem includes open-source tools developed in languages and platforms familiar to researchers: analytic packages in R (programming language), database connectors for PostgreSQL, Microsoft SQL Server, and Oracle Database, and visualization tools interoperable with Shiny (software) and Tableau Software. Core components comprise tools that evolved alongside projects like OHDSI Atlas-style platforms, cohort definition utilities inspired by i2b2, and empirical calibration methods reflecting methodological literature from Stanford University and Harvard Medical School. Execution environments often run on cloud providers such as Amazon Web Services, Microsoft Azure, and Google Cloud Platform for scalable cohort generation and distributed analytics.

Research Activities and Key Studies

Research outputs span comparative effectiveness, drug safety, and disease natural history studies performed with collaborators from Food and Drug Administration, European Medicines Agency, National Institutes of Health, and academic consortia including Mayo Clinic and Brigham and Women's Hospital. Notable large-scale studies assessed cardiovascular safety, vaccine safety, and real-world effectiveness of therapeutics, leveraging datasets from Medicare, MarketScan, and international claims registries in South Korea and Japan. Methods research addressed phenotyping, bias quantification, and empirical calibration, citing methodological frameworks from Epidemiologic Methods groups at University of California, Los Angeles and Columbia University.

Collaboration, Training, and Community Outreach

OHDSI emphasizes education through community-driven tutorials, hackathons, and training sessions held at venues like American Medical Informatics Association and International Conference on Pharmacoepidemiology, as well as regional chapters in Latin America, Asia-Pacific, and Europe. Outreach includes partnerships with clinical trial networks such as ClinicalTrials.gov contributors, public health agencies like Centers for Disease Control and Prevention, and standards organizations including Health Level Seven International. The community model mirrors collaborative ecosystems exemplified by Apache Software Foundation, Linux Foundation, and research networks like All of Us Research Program.

Category:Clinical research networks