REDCap: Randomization Table Corrections
Modifying a Randomization Table in Production
Randomization tables should be developed and tested while a project is in Development status. All users should be encouraged to test and confirm their table allocations prior to their project being moved into Production.
In the case that a mistake was made and a project needs to upload an entirely new randomization table to their project, the steps below can be used to delete and create a new randomization table in a Production project.
NOTE: Make a backup copy of the project's data and XML file prior to modifying the Randomization set up.
Removing Current Randomization Table
Navigate to Set up Randomization Module in Project Settings. Under this tab, you will see the options under STEP 1 for general randomization settings.
- Document any stratification currently set up and which field/event it is set under.
- Document any randomization by group/site.
- Make a note of which field is currently being used to randomize records in the project.
Erasing the randomization model will delete all settings related to randomization from the project. This includes the previous randomization table in the project. If you would like to keep an copy of the original randomization table, you will need to download the current randomization table from Randomization STEP 3 before erasing the randomization model.
Once these items are documented, select "Erase randomization model"
Adding New Randomization Field
If a project wishes to keep the values from their old randomization set up, then a new field must be created to take the place of the initial field.
This copied field should match the options of the first randomization OR can be changed to match the new randomization table if options need to change. The first randomization field can be set to @HIDDEN or @READONLY to prevent former values from being altered. Adding branching logic to check if the original randomization field has a value can help show the randomization for old records that were randomized under the old randomization, but hide the field from new records that will use the new randomization.
Adding New Randomization Table
After adding a new field to use for Randomization, navigate back to the Randomization Set Up and re-enter the values from the original Randomization. For Part C, instead of using the original randomization field you will need to assign the new field as the randomization variable. Once this is complete, select "Save randomization model" and navigate to STEP 3.
Find the new randomization allocation table needed for the project. Make sure that the header for the first column (randomization field name) matches the new field created).
In STEP 3, select "Upload more allocations (Administrators only)". The current allocation table will be empty from removing the Randomization settings earlier, adding in new allocations will now start the new table at the beginning.
If the upload was successful you will be able to confirm the changes and see a green check mark indicating the table was uploaded.
Navigate to the Randomization Dashboard and you should see a fresh randomization table with no allocations currently being used that will be the distribution of the current randomization schema.