Randomisations can be marked as in error by an administrator if necessary. Doing so excludes the randomisation from reports and, where minimisation is used, excludes the randomisation from the balancing algorithm when future randomisations are performed.
For trials using permuted blocks marking in error does not affect the block that the allocation was taken from. Once an allocation is used in a block, it cannot be undone. This means you should try to minimise mistakes to avoid unduly affecting the treatment balance.
Randomisations should only be marked as in error when a mistake has been made, such as randomising ineligible subjects or randomising the same subject twice. Randomisations marked as errors would not normally be included in an intention to treat analysis, and consequently care should be taken not to introduce bias by inappropriate marking. Some useful discussions of post-randomisation exclusions can be found in these papers:
Fergusson D, Aaron S, Guyatt GH, Hebert P. Post-randomisation exclusions: the intention to treat principle and excluding patients from analysis BMJ. 2002; 325:652-654.
Yelland LN, Sullivan TR, Voysey M et al. Applying the intention-to-treat principle in practice: Guidance on handling randomisation errors Clinical Trials. 2015; 12(4):418-423.
The Yelland paper makes the following recommendations for handling common randomisation errors:
| Randomisation error | Recommendations | How to handle in Red Pill |
|---|---|---|
| Participant randomised using incorrect baseline information | Accept the randomisation but record the correct baseline information | Edit the randomisation form |
| Ineligible participant randomised | Keep the participant in the trial and collect all relevant data, unless an unbiased process for excluding ineligible participants has been pre-specified. Seek clinical input to determine their appropriate management. | - |
|
Participant randomised multiple times:
(a) Only one set of baseline and outcome data will be obtained. (b) Multiple sets of baseline and outcome data will be obtained. |
(a) Retain the initial randomisation and disregard the second randomisation. (b) Retain both randomisations, unless an unbiased process for excluding the second randomisation has been pre-specified. |
(a) Mark the second randomisation as in error. If a withdrawal form is present complete it to prevent other forms in the CRF being completed. |
| Participant received incorrect treatment | Record the treatment the participant received and seek clinical input regarding their ongoing treatment. | Open a query documenting what happened. |
Reproduced under the terms of the Creative Commons CC BY license
To mark a randomisation as in error the appropriate record should be viewed and the Mark as randomised in error link followed in the subject details section. Marking in error cannot be undone, so care should be taken to ensure the correct record is chosen by double checking the subject identifier shown in the heading. The user will be asked to enter a reason and their password to confirm the need for marking as in error.
A randomisation marked in error
After entering a reason and the correct password and clicking the Mark as in error button the record will be marked. The date and time, reason and user who marked the record as in error will be recorded in the details for the randomisation concerned. A red warning triangle will be displayed in the status column of the subject list for those marked in error.