Data sources

The Data Sources entry in the navigation bar opens the Data Sources catalog. It is also the page you see after signing in.

Data Sources catalog

The catalog shows a card for each data source you can see: public data sources and data sources to which you have been granted access. Each card shows:

• The name and a short description.
• Public, if the data source is publicly available.
• Your access: Have access, Pending access or No access.
• Details such as Subject, Published, Data source type and Version.

Use Search to find data sources by name or description. Use Sort by to sort the cards by Access, Name A-Z or Name Z-A.

Select a card to open the data source.

Data source pages

A data source opens with a sidebar of pages:

Group	Pages
Overview	Description, Dashboard, Data Density
Demographics	Person, Observation Period, Death
Clinical Domains	Visit, Condition Occurrence, Condition Era, Procedure, Drug Exposure, Drug Era, Measurement, Observation
Data quality	Data quality dashboard

Description

The Description page shows:

• Description: detailed information about the data source and how to access it.
• Metadata: the dataset ID and further attributes the data provider maintains.
• Files: resource files the data provider added. Select Download to download a file.

Requesting access

If you do not have access to a data source yet, the Description page shows No access and a Request access button. Select Request access to send a request to the administrator. The status changes to Pending access until the request is approved.

If the page shows Restricted, access cannot be requested. Contact your administrator.

Data characterization reports

The Dashboard, Data Density, Demographics, and Clinical Domains pages show the data characterization results of the data source:

• Dashboard: the number of persons, gender split, age at first observation, and observation time.
• Data Density: the number of records per domain over time.
• Person, Observation Period, Death: demographics in more detail.
• Clinical Domains: the most frequent concepts per domain, as a tree map or table. Select a concept to see its prevalence over time and by age and gender.

These reports help you understand what is in a data source before you design a cohort: how many persons, which time span, and how complete each domain is. Chapter 4 "Data sources" of the ATLAS user manual explains each report in detail.

If the reports are empty, the data characterization has not run for this data source or no results are available for the specific report.

Data quality dashboard

The Data quality dashboard page shows the results of the latest data quality check of the data source. The check runs the OHDSI Data Quality Dashboard (DQD), a set of several thousand checks organized according to the Kahn framework:

• Plausibility: Does the data conform with basic logical and medical expectations?

  Example: Does the measurement unit provided for a specific lab test make sense (for example, is it a unit like cm or m for body height)?

• Conformance: Does the data conform to the OMOP Common Data Model?

  Example: Is the person ID given for a diagnosis entry indeed the primary ID of an entry in the PERSON table?

• Completeness: Are all the expected data elements and vocabulary mappings present?

  Example: Does every medication entry have a standard OHDSI code identifying the medication given?

The page shows the Overall pass rate and a table with the number of checks that passed and failed per category, split into Verification and Validation checks.

Your administrator runs the data quality check. If no results are shown yet, the page says so.