Skip to content

Data Export De identification and Formatting Options

Data Export — De-identification & Formatting Options

Article ID RC-EXPRT-04 — Data Export: De-identification & Formatting Options
Domain Exports, Reports & Stats
Applies To All REDCap project types; availability varies by Data Export Rights level
Prerequisite RC-EXPRT-01 — Data Export: Overview & Workflow; RC-EXPRT-03 — Data Export: User Rights & Export Access
Version 1.0
Last Updated 2025
Author See KB-SOURCE-ATTESTATION.md
Related Topics RC-EXPRT-03 — Data Export: User Rights & Export Access; RC-EXPRT-02 — Data Export: Export Formats

1. Overview

This article covers the optional de-identification settings and data formatting options available in REDCap's export dialog. De-identification options allow you to remove or transform potentially identifying information before downloading your data. Formatting options control how values are represented in the exported file. The availability of these options depends on the user's Data Export Rights level — see RC-EXPRT-03 — Data Export: User Rights & Export Access for details on rights levels.

2. Key Concepts & Definitions

De-identification

The process of removing or transforming data elements that could be used to identify an individual participant. REDCap's export de-identification options target four categories: identifier-flagged fields, free-form text fields, the Record ID field, and date/datetime fields.

Identifier Field

A variable that has been flagged by the project designer as containing potentially identifying information. Examples include name, date of birth, and medical record number. The identifier flag is set in the Online Designer or via the Check for Identifiers tool in Project Setup.

Date Shifting

A de-identification technique that adds a random number of days (between 0 and 364) to all date and datetime fields in a record. The same random offset is applied consistently within a single record, preserving the relative time relationships between dates while making the actual dates less identifiable. The offset differs between records.

Hash (Record ID Hashing)

A one-way transformation applied to the Record ID field. Hashing scrambles the Record ID value in a deterministic but irreversible way. Two exports hashed with the same algorithm will produce the same hashed values, allowing records to be linked across exports without revealing the original identifier. Only relevant when the Record ID contains an actual identifier such as a medical record number.

CSV Delimiter

The character used to separate values in a CSV file. The standard delimiter is a comma, but some regional configurations (common in Europe) use a semicolon or tab instead. Mismatched delimiters cause CSV files to display incorrectly in spreadsheet applications.

3. De-identification Options

De-identification options appear in the export dialog under a dedicated section. Options available to a user depend on their Data Export Rights level — users with De-identified rights have these options pre-checked and locked; users with Full Data Set rights have them optional.

3.1 Known Identifier Options

Option What it does
Remove all Identifier Fields Removes all variables flagged as identifiers from the export. This is the simplest and most reliable way to de-identify an export, provided the project designer has correctly flagged all sensitive fields.
Hash the Record ID Field Replaces the Record ID with a hashed (scrambled) value. Use this only if the Record ID itself contains an identifier such as a medical record number, phone number, or social security number. If the Record ID is an auto-assigned integer, hashing is unnecessary.

3.2 Free-Form Text Options

Option What it does
Remove unvalidated Text fields Removes all text box fields that have no validation rule applied. Validated text fields (e.g., those with date, number, or email validation) are still included. Use this when unvalidated text fields may contain participant-entered narrative content.
Remove Notes/Essay box fields Removes all Notes box fields from the export. Notes fields can contain large amounts of free-form text that may include identifying information. This option errs on the side of caution by removing all notes fields regardless of content.

3.3 Date and Datetime Options

For date de-identification, only one of the two options can be selected.

Option What it does
Remove all date and datetime fields Removes all variables with a date or datetime validation from the export entirely. This is the most aggressive date de-identification option.
Shift all dates by value between 0 and 364 days Adds a random offset (0--364 days) to all date and datetime values in the export. The same offset is applied to all dates within a single record, preserving relative date relationships. Different records receive different offsets. Use when date values have research utility but exact dates are a privacy concern.

Important: Date shifting preserves relative time relationships within a record (e.g., days between enrollment and first visit) but makes absolute dates less identifiable. It does not prevent re-identification if the shifted dates are combined with other identifying information.

4. Advanced Data Formatting Options

These options control how the exported data is structured and formatted. They appear below the de-identification section in the export dialog. The defaults are appropriate for most exports.

Option What it does
Export blank values for gray Form Status? Controls whether an instrument with no data saved (gray dot / status 0) exports as '0' or as a blank value. The default exports '0'. Change this only if you are using the export as a re-import file and need blank to mean 'no data' rather than 'incomplete'.
Set CSV delimiter character Changes the character used to separate values in CSV exports. The default is a comma. If your analysis software or regional settings expect a semicolon or tab delimiter, set this accordingly. Mismatched delimiters cause CSV files to open as a single column in spreadsheet applications.
Force all numbers into a specified decimal format? Standardizes the decimal separator used in numeric values. Use this when your analysis software or regional settings expect a period (.) or comma (,) as the decimal character. Mismatched decimal formats cause numeric fields to be read as text in some applications.

4.1 Additional Options (Feature-Dependent)

When certain project features are enabled, additional options appear in the export dialog automatically. For example, if one or more Data Access Groups are configured in the project, an option appears to export an additional variable containing each record's DAG assignment. These options are contextual and only appear when relevant.

5. Common Questions

My de-identification options are grayed out and pre-checked. Why can't I change them? Your Data Export Rights level for one or more instruments is set to De-identified or Remove All Identifier Fields. At these levels, certain de-identification options are mandatory and cannot be unchecked. Contact your project administrator if you need a higher access level.
Should I use date shifting or remove all dates? Use date shifting when date values have research utility and you want to preserve relative time relationships between events. Use remove all dates when the analysis does not require date values at all and you want the simplest approach to de-identification.
When should I use the Hash Record ID option? Only when the Record ID field contains an actual identifier, such as a medical record number, phone number, or social security number. If the Record ID is an auto-assigned integer (the REDCap default), there is no reason to hash it.
My exported CSV opens as a single column in Excel. What is wrong? The CSV delimiter does not match what Excel expects for your regional settings. Open the export dialog and change the delimiter to match your system — typically a semicolon for European regional settings. Alternatively, use Excel's Text Import Wizard to specify the delimiter manually.
Does removing unvalidated text fields also remove validated text fields? No. The Remove unvalidated Text fields option only removes text boxes that have no validation rule applied. Text boxes with date, number, email, or other validation are still included in the export.
What does the gray Form Status export option actually affect? It only affects instruments with no data saved (gray dot). Changing this from '0' to blank is only relevant if you are using the export as a data import file and your import logic distinguishes between 'incomplete' (0) and 'no data entered' (blank). For most analysis purposes, the default is fine.

6. Common Mistakes & Gotchas

Hashing the Record ID when it is an auto-assigned integer. Record ID hashing is only meaningful when the Record ID contains a real identifier. Hashing an auto-assigned integer (1, 2, 3...) provides no privacy benefit and may complicate downstream data linkage.

Assuming de-identification is complete when identifier flagging is incomplete. De-identification options act on fields flagged as identifiers. If the project designer did not flag all sensitive variables during project design, those variables will still appear in de-identified exports. Verify the project's identifier flags before relying on export de-identification as a privacy safeguard.

Selecting both date de-identification options. REDCap only allows one date de-identification option — remove all dates or shift all dates. Selecting both is not possible; choosing one disables the other.

Ignoring the delimiter setting when exporting for international collaborators. Regional CSV delimiter differences are a common source of data formatting problems. When sharing exports internationally, confirm the expected delimiter with the recipient before exporting.