In this blogpost, we discuss the significance of OHDSI, OMOP, and EHDEN in improving healthcare through data standardization and efficient analysis. We detail the advantages of OMOP CDM data warehouses for hospitals and showcase how these configurations contribute to clinical evidence generation.
OMOP, OHDSI, EHDEN – say what?
What is OHDSI?
The Observational Health Data Sciences and Informatics (or OHDSI, pronounced "Odyssey") program is a multi-stakeholder, interdisciplinary open-science community with the overall goal of improving healthcare for patients. Through its open-source software portfolio and methods for data standardization and analysis (including OMOP CDM), it aims to enable analytics on substantial amounts of health data.
What is OMOP?
The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) is an open community data standard, developed by the OHDSI community. It is designed to standardize the structure and content of observational data and to enable efficient analyses that can produce reliable evidence.
What is EHDEN?
EHDEN is the European Health Data & Evidence Network. It is a federated data network launched to enable international large-scale research to generate real-world evidence (RWE) from real-world data (RWD). From the start, EHDEN has been working hand-in-hand with OHDSI.
Collaboration as an important catalyzer of research and innovation
With most health data being held in an unstructured format (e.g., Electronic Health Records) or in a structured format but in different databases and with different medical coding dictionaries, the data can serve for its primary use within the hospital but becomes difficult to reuse for research.
Because of the growing need for standardized health databases, various initiatives have grown in the past decades, including the OHDSI community. The latter developed the OMOP CDM Model to create interoperable health databases.
In practice, hospital data gets extracted, processed, and structured in the standardized OMOP format, so it can be used to create Findable, Accessible, Interoperable and Reusable (FAIR) data warehouses. These federated clinical data warehouses can serve hospitals as well as researchers to gain insights into real-world populations, on the effectiveness of treatments, the progression of diseases, and much more. This is the concept of a federated approach, as shared by EHDEN.
Data Quality in full transparency
The quality and completeness of the data highly influences the quality of a performed analysis or research. It is therefore of utmost importance that the necessary quality checks are done throughout the data processing cycle, from source data to target data.
MirrorHealth guarantees data quality through various manual and automated checks:
- Data quality reports: Providing background and transparency on quality assurance protocols. For example, computing Recall (true positive/all positive) and precision metrics on extracted events.
- Validation dashboards: Giving a highly detailed quality assessment of all extracted data from both structured and unstructured medical files.
- Source analysis: Providing an overview of data sources and population density.
- Data dictionaries: Indicating which proxies are used for each variable.
In addition, we deploy standard OHDSI tools that connect directly to the CDM (such as ACHILLES, ATLAS, and the Data Quality Dashboard) to ensure hospitals have everything at hand to make optimal use of their data of which they remain owners and in control at all times.
A harmonized data warehouse facilitating federated analytics and research
As a hospital, the advantages of having an OMOP CDM data warehouse with standardized observational health data are numerous:
- A harmonized data warehouse and easy-to-read data visualizations can help to gain valuable clinical insights into the quality of care.
- The costs of healthcare could be reduced through information gained from medical and economic analyses of the hospital's processes.
- Benchmarking becomes significantly easier, as well as investigator-initiated studies and publications in collaboration with other hospitals.
- Patient recruitment for clinical trials gets streamlined and accelerated.
"Our federated AI data model ensures all data remains under the custody of the hospital. Hospitals are in full control of how that data is used, and patients remain their sole proprietors."