Atlas

This document helps you to get started using OHDSI Atlas with FinnGen register data.

Atlas allows you to create cohorts using a web interface.

Quick start

The Atlas is installed in the FinnGen sandbox.

  • Log into the sandbox:

https://sandbox.finngen.fi/

  • Select the following menu option:

Applications > FinnGen > Atlas

  • This opens the Atlas application in a browser window

A quick start for defining cohorts using Atlas is available here.

Introduction

Atlas allows you to create cohorts using a web interface. Cohorts are groups of individuals who share common characteristics of interest, such as disease classifications, prescription medications or hospital procedures. Cohorts can be used to study associations between these characteristics and other data to reveal which factors increase or decrease the likelihood of developing a certain condition.

You can create cohorts in Atlas using both standard and non-standard medical codes. They are loaded into the OMOP-CDM data model used by Atlas from the service sector data. The loading process is explained in full detail in FinnGen ETL to OPOP CDM. In addition to mapping non-standard codes to standard codes, the ETL process resolves many register data errors, inconsistencies and anomalies.

Atlas user interface

Data Sources

This functionality allows you to create FinnGen data release-level descriptive summary statistics.

An OHDSI tutorial for viewing Data Sources is available here.

How to view a report:

  • Select the Data Sources option in the left menu

  • Select the FinnGen data release using the first drop-down menu

  • Select the report using the second drop-down menu

In the example below, we have selected the FinnGen data release 11 and the Person report. This report starts with a distribution of people based on their year of birth:

This functionality allows you to search the vocabularies (codes) loaded into the Atlas and to show the number of matching records.

Make sure to select the correct FinnGen data release for both codes and the record counts when using the Search.

The Search is made against the vocabularies (codes) from one FinnGen data release while the number of matching records is shown for a potentially different FinnGen data release. These FinnGen data releases can be defined on the Configuration page using the Vocabulary Version and Record Counts columns. The FinnGen data release against which the codes are searched can also be defined on the Search page using the View record count drop-down list.

An OHDSI tutorial for using Search is available here.

How to search codes:

  • Select the Search option in the left menu

  • Type in the search term in the text box at the top

You can change the FinnGen data release that is being searched against using the View record counts drop-down list. However, this does not affect the FinnGen data release from where the codes are selected for the search. This FinnGen data release is defined on the Configuration page.

In the above example, we used the 'Asthma' search term. Standard, non-standard and classification codes are shown in different colours, for example, the standard codes are in blue. More information about these codes is available here.

Concept Sets

The cohort definition starts by creating a Concept Set. A Concept Set contains a set of medical codes that are used when defining a cohort in Cohort Definitions. A cohort is defined using one or more Concept Sets. For example, one Concept Set could contain clinical diagnostics codes while another could contain administered drugs.

OHDSI has tutorials for viewing and creating Concept Sets.

How to view Concept Sets:

  • Select the Concept Sets option in the left menu

  • Click the Name to see which concepts (codes)are included in the Concept Set.

How to create Concept Sets:

More information about viewing and creating Concept Sets is available here.

Cohort Definitions

A cohort is defined in Cohort Definitions using one or more Concept Sets. For example, one Concept Set could contain clinical diagnostics codes while another could contain administered drugs. Cohorts may be created in Cohort Definitions by defining additional criteria such as age inclusion criteria or cohort entry and exit criteria.

In Atlas, people are always included in the cohort for a duration. By default, they are included for the period they belong to a Concept Set.

OHDSI has a tutorial for creating a Cohort Definition.

How to view Cohort Definitions:

  • Select the Cohort Definitions option in the left menu

  • Click the Name to see more information about the Cohort Definition

How to create Cohort Definitions:

More information about viewing and creating Cohort Definitions is available here.

Characterizations

This functionality allows you to create cohort-level descriptive summary statistics.

OHDSI has a tutorial on Characterizations.

In the example below, we have created gender summary statistics in FinnGen data release 11:

More information about Characterizations is available here.

Cohort Pathways

This functionality helps you to understand the sequence of events within a cohort.

OHDSI has a tutorial on Cohort Pathways here.

More information about Cohort Pathways is available here.

Incidence Rates

This functionality allows you to estimate incidence rates.

OHDSI has a tutorial on Incidence Rates.

Configuration

Atlas contains several FinnGen data releases. The Search is made against the vocabularies (codes) from one FinnGen data release while the number of matching records is shown for a potentially different FinnGen data release. These FinnGen data releases can be defined on the Configuration page using the Vocabulary Version and Record Counts columns.

Please follow these instructions to define the FinnGen data release for Search and for defining Concept Sets using the Search functionality**.**

In the example below, we have selected the FinnGen data release 10 in both Vocabulary Version and Record Counts columns:

The FinnGen data release is also selected when generating:

  • Cohort Definitions

  • Characterisations

  • Cohort Pathways

  • Incidence Rates

Because codes may change between FInnGen data releases, it is safest to use the same FinnGen data release when defining Concept Sets and when generating Cohort Definitions or Characterisations.

Atlas data model

You can define better cohorts if you understand the different characteristics of FinnGen register data sources and how the data is made available in the Atlas data model.

For example, Finngen does not have bone density measurements. However, you can use bisphosphonate treatments as a proxy for bone mineral density. The reason for this is that the treatment reimbursement requires a bone mineral density test and FinnGen makes the reimbursement data available from the Kela drug reimbursement register.

Topics in the section

Further information

Last updated