Atlas data model

Atlas uses the Observational Medical Outcomes Partnership (OMOP) common relational data model (CDM). This data model supports both standard and non-standard medical codes.

The OMOP-CDM relational data model used by FinnGen is explained in detail in the OMOP-CDM v.5.4 specification. OMOP-CDP data model conventions are described here.

Database tables

The data model consists of the following tables. The Standardized clinical data tables and the Result tables are briefly described in the following sections:

Standardized clinical data tables

Standardized clinical data tables contain all individuals and medical events. Individuals are stored in the PERSON table. In FinnGen, each individual has one observation period stored in the OBSERVATION_PERIOD table. It starts at birth or from the first entry to the registers and ends at death or at the end of the follow-up period. The other tables contain medical events.

OHDSI standardized vocabularies are strictly limited to one OMOP-CDM Domain and all medical event tables belong to only one OMOP-CDM Domain. The Domain has an effect on how cohorts are defined in Atlas.

Table	Domain	Description
PERSON		Uniquely identifies each individual and contains demographic information.
OBSERVATION_PERIOD		Defines the span(s) of time during which events happen to an individual. In FinnGen, each individual has only one observation period. It starts at birth or from the first entry to the registers. It ends at death or at the end of the follow-up period.
VISIT_OCCURENCE	Visit	Events where an individual interacts with the healthcare system for a duration of time. For example, an event could be a visit to a hospital, a diagnosis, a procedure, or a drug intake.
CONDITION_OCCURRENCE	Condition	Medical diagnosis events.
PROCEDURE_OCCURRENCE	Procedure	Medical operation events.
DRUG_EXPOSURE	Drug	Drug exposure events.
MEASUREMENT	Measurement	Laboratory test events.
DEATH	Death	Death events.
OBSERVATION	Observation	Events not captured by other domains.

FinnGen uses both standard and non-standard medical codes. Standard and non-standard codes are stored in the medical event tables using the following naming convention:

Column	Descriptions
_CONCEPT_ID	Foreign key to the medical code used in the OHDSI standard vocabulary. Standard codes are stored in this column.
_SOURCE_CONCEPT_ID	Foreign key to the medical code used in the source data. Non-standard codes are stored in this column.

Results tables

The result tables contain cohort definitions as well as cohorts created using them.

Table	Description
COHORT_DEFINITION	Contains information to define a cohort. Cohorts are a set of individuals that satisfy inclusion criteria for a duration of time. The information in the COHORT_DEFINITION table is used to instantiate a cohort and store it in the COHORT table.
COHORT	Contains the individuals that satisfy the definition of a cohort stored in the COHORT_DEFINITION table. The period of time an individual is part of the cohort is captured in the COHORT_START_DATE and COHORT_END_DATE columns. Please note than an individual may enter and exit the cohort more than once.