In this document we describe various conventions used in the processing and storage of data within the OHDSI architecture. Data within the OHDSI data architecture falls into one of four categories including Source Data, Standardized Data, Derived Data and Administrative Data.
There is a life-cycle of data within the OHDSI data architecture. Data primarily originates from a person level data source. These data sources include electronic health records, administrative claims records, clinical trial data and billing data among others. These person level data sources are typically received in their own proprietary format which we refer to as the “Native Format”. The approach used within OHDSI is to standardize these sources of person level data by performing ETL processes and converting them from their Native format to the OMOP Common Data Model (CDM) format. Once this conversion is complete the intent is for the CDM to be a read only data source.
Once a data source has been converted from the native schema to the CDM schema the OHDSI tools provide features to perform many different analyses. These analyses include high level population level descriptive statistics and characterization, definition and generation of cohorts, study specification and population level estimation among others. The processes that implement these features generate new data that has been derived from the CDM data schema. This derived data is then stored in the results schema. The OHDSI Platform Metadata that describes the concepts and processes that were used to derive these results from the CDM schema are stored in the OHDSI schema.