Skip to content

Clinical Data Mart Setup and Usage

Clinical Data Mart (CDM): Setup and Usage

Article ID RC-CDIS-03 — Clinical Data Mart (CDM): Setup and Usage
Domain Clinical Data Interoperability Services
Applies To Institutions using bulk clinical data import for registries and retrospective studies
Prerequisite RC-CDIS-01 — Clinical Data Interoperability Services: Overview & Control Center Setup
Version 1.0
Last Updated 2026
Author See KB-SOURCE-ATTESTATION.md
Related Topics RC-CDIS-01 — Clinical Data Interoperability Services: Overview & Control Center Setup; RC-CDIS-02 — Clinical Data Pull (CDP): Setup and Usage; RC-CDIS-04 — CDP vs CDM: Feature Comparison

Prerequisite: CDIS must be configured at the system level before CDM can be used. See RC-CDIS-01 — Clinical Data Interoperability Services: Overview & Control Center Setup.

1. Overview

Clinical Data Mart (CDM) is a REDCap module that imports clinical data in bulk from an EHR for multiple patients at once. Unlike CDP, which pulls one patient at a time with adjudication, CDM uses a pre-defined project structure and imports data directly without a review step. CDM is ideal for registries, retrospective studies, and bulk data retrieval across large patient cohorts.

2. What Is the Clinical Data Mart (CDM)?

The Clinical Data Mart (also called Data Mart or CDM) is a REDCap module that imports clinical data in bulk from an EHR for many patients at once. Unlike CDP, which pulls one patient at a time with an adjudication step, Data Mart is designed for larger-scale, batch data retrieval.

Data Mart is best suited for: - Registries - Prospective or retrospective clinical studies and trials - Searching for specific lab values or diagnosis codes for a cohort of patients over a defined time period

3. Administrator Prerequisites

1. CDIS Must Be Enabled

The CDIS infrastructure must be configured and enabled in the Control Center before Data Mart can function. See RC-CDIS-01 — Clinical Data Interoperability Services: Overview & Control Center Setup.

2. Grant User-Level Data Mart Privileges

Unlike CDP (which is enabled per project), Data Mart access is a user account privilege, not a project-level right. An administrator must grant this privilege to individual user accounts via the Browse Users page in the Control Center.

Once granted, the user will see an option to create a Clinical Data Mart project on the Create New Project page.

Note: There is no optional User Access Web Service for Data Mart (unlike CDP). Access is controlled solely through the user account privilege granted in the Control Center.

4. Creating a Data Mart Project

A user with Data Mart privileges can create a new Data Mart project from the Create New Project page.

Important: The user must have launched REDCap from within the EHR at least once before creating a Data Mart project. If they have not done so, they will be prompted to complete this step first.

During project creation, the user defines the initial data pull configuration, including: - Specific MRNs (patient identifiers) to include - Date range for the data pull - Data fields from the EHR to import

5. Pre-Defined Project Structure

Unlike CDP, no field mapping is required for a Data Mart project. The entire project structure — instruments and fields — is pre-defined when the project is created:

  • Demographics — a single data collection form (non-repeating)
  • Vital Signs, Labs, Allergies, Medications, Problem List — created as repeating instruments

Each individual data value on the repeating instruments is stored as a separate repeating instance of the form. This means a patient with 10 lab results will have 10 repeating instances of the Labs instrument.

Additional instruments or events may be added to a Data Mart project, but modifying or deleting any of the pre-defined fields or instruments may break the data pull.

6. Pulling Data

Once the project is created, data is pulled via the Clinical Data Mart page, accessible from the left-hand project menu.

Fetch Clinical Data

Users click the "Fetch clinical data" button to trigger a data pull. By default: - A user can only pull data one time. - The data pull configuration cannot be modified after initial setup.

These defaults can be changed by a REDCap administrator in the project's Project Setup page (see Project-Level Settings below).

Requesting Configuration Changes

If the project-level setting permits it, users with Project Setup/Design rights in the project can request changes to the data pull configuration (e.g., adding new MRNs, adding new fields, or modifying the date range). Such requests go through an audit process and must be approved by an administrator via the To-Do List in the Control Center.

7. Project-Level Settings (Administrator-Controlled)

On the Project Setup page for a Data Mart project, a REDCap administrator can adjust:

Setting Default Description
Allow multiple data pulls Off (one pull only) When enabled, users can fetch data as many times as they wish
Allow configuration changes Off When enabled, users can request changes to the MRN list, date range, or field selection; changes go through the To-Do List for admin approval
Automatic daily data pull (cron job) Off When enabled, new EHR data is pulled automatically once per day

8. EHR Access Requirement

Users must have access to the EHR and must have launched REDCap from inside the EHR interface at least once. This is required because the EHR launch initiates the OAuth2 authorization needed to pull data.

Before approving a Data Mart project or enabling configuration changes, administrators may want to verify: - IRB approval status - The requesting users' EHR access authorization

This review step is optional but recommended, particularly at institutions with policies around research data access.

10. Clinical Notes in Data Mart

CDM accepts clinical notes in any format (not just HTML), making it more flexible than CDP for institutions that receive notes in varied formats. Clinical notes are imported via the FHIR DocumentReference resource.

11. Common Questions

Q: Can I customize the pre-defined project structure (instruments and fields) in CDM? You can add new instruments and fields to a Data Mart project, but you cannot modify or delete the pre-defined instruments (Demographics, Vital Signs, Labs, Allergies, Medications, Problem List) without risking broken data pulls. If you need a highly customized structure, use CDP instead.

Q: What is the difference between allowing multiple data pulls and allowing configuration changes? "Allow multiple data pulls" lets users fetch data as many times as they want. "Allow configuration changes" lets users request modifications to the MRN list, date range, or field selection; these requests go through the To-Do List for admin review. Both settings are off by default and must be enabled in Project Setup.

Q: Can I pull data retroactively for patients who were added to the project long ago? Yes. The data pull is based on the MRNs and date range you specify at project creation or via a configuration change request. You can include any MRNs and any date range, regardless of when the project was created. Data will be retrieved from the EHR for that date range.

Q: Do I need to approve configuration change requests for Data Mart projects? Only if you enable the "Allow configuration changes" setting in the project. If that setting is off (the default), users cannot request changes and cannot pull data a second time. If it is on, change requests appear in the administrator's To-Do List for review.

Q: What does "automatic daily data pull (cron job)" do in CDM? If this setting is enabled, REDCap will automatically fetch new EHR data once per day without requiring the user to manually click "Fetch clinical data." This is useful for ongoing registries where you want fresh data regularly. The cron job uses the same configuration (MRNs and date range) that was set at project creation.

Q: Can I delete a patient record from a Data Mart project after data has been pulled? Yes, you can delete records from a Data Mart project like any other REDCap project. However, deleting a patient's MRN from the pull configuration and re-pulling will not remove existing data — it will just prevent future data pulls for that patient. To permanently remove a patient, you must manually delete the record(s).

12. Common Mistakes & Gotchas

Modifying or deleting pre-defined instruments or fields. The Demographics, Vital Signs, Labs, Allergies, Medications, and Problem List instruments are created with a specific structure to match FHIR data. If you delete or significantly alter these instruments, future data pulls may fail or insert data into the wrong fields. Leave the pre-defined structure intact.

Setting too narrow a date range at project creation. The date range is used to query the EHR for all relevant clinical data. If you set a range that is too narrow (e.g., only 1 month when you need 2 years), you will miss historical data. You cannot change the date range after project creation unless you enable the "Allow configuration changes" setting and request an admin-approved change.

Forgetting that the initial user must have launched from the EHR first. A common surprise is when a user tries to create a Data Mart project and is blocked because they have not completed an EHR Launch. The system requires this step to establish OAuth2 authorization. Ensure users complete at least one EHR Launch before they attempt to create a Data Mart project.

Approving configuration change requests without reviewing IRB or access status. The To-Do List shows change requests, but it does not automatically check IRB approval or EHR access authorization. It is your responsibility to verify these details before approving. A recommended practice is to review the requesting user's credentials and project IRB status before accepting changes.

Assuming repeating instances will be deduplicated. Each individual EHR value (each lab result, each vital sign, each medication) becomes a separate repeating instance in CDM. If the same lab test is performed twice on the same date, you will get two repeating instances. This can create a large number of instances for patients with many test results. Plan your data analysis accordingly.