Background: The Observational Health Data Sciences and Informatics (OHDSI, www.ohdsi.org) programme is a multi-stakeholder, interdisciplinary collaborative creating open-source solutions that enhance the value of observational health data through large-scale data sharing and analytics.
Objectives: OHDSI’s mission is to optimise the value of observational health data through large-scale analytics. Our international research community enables active engagement across multiple disciplines (e.g. clinical medicine, biostatistics, computer science, epidemiology, and life sciences), spanning multiple stakeholders (e.g. researchers, patients, providers, payers, product manufacturers and regulators).
Methods: OHDSI employs rigorous data-standardisation conventions to transform patient-level clinical and claims data to a transparent and reproducible, harmonised information model. The OMOP Common Data Model (CDM) allows for the use of shared analytic tools for data quality assessment, phenotype building, cohort building, and population- and person-level predictions. All tools are open-source and managed by an active international community.
Results: Currently, over 50 international data partners, representing over 140 multidisciplinary research collaborators in over 20 countries have transformed data from more than 660 million patients. The CDM and standardised vocabulary have been downloaded over 2300 times since 2015. There are nearly 600 community forum members with more than 3600 visits to the top 5 posts alone. About 70 studies have been published exploring OHDSI analytic methodology and the impact of observational big-data analytics. Multiple research studies are under way using standardised data from OHDSI to investigate questions including fully characterising oral antibiotics for acne vulgaris treatment and delineating clinical treatment pathways for diabetes mellitus.
Conclusions: OHDSI establishes an international collaboration to achieve high-quality, efficient and transparent observational research. This effort holds promise for improving population-level estimation, comparative effectiveness research, quality improvement and public policy.