03 · What You Need to Know
How to Organize Enrollment and Intervention Assignment
Why Are Enrollment and Assignment Different Responsibilities?
Enrollment determines whether an individual becomes a participant in the randomized comparison. Assignment determines which intervention condition that enrolled participant receives.
These decisions must occur in an appropriate sequence. Eligibility and required enrollment procedures should be completed before the assigned treatment is disclosed to anyone who could influence participation.
Suppose a recruiter knows that the next allocation is to an experimental intervention. If eligibility requires professional judgment, the recruiter might interpret an ambiguous criterion more favorably for a particular candidate.
Even without deliberate manipulation, expectations about treatment benefits can influence recruitment timing or decisions about borderline eligibility.
Allocation concealment prevents knowledge of upcoming assignments from entering these decisions. It is distinct from blinding, which concerns knowledge of treatment identity during later trial activities.
What Are the Main Responsibilities in the Randomization Process?
| Function |
Principal responsibility |
Essential safeguard |
| Eligibility assessment |
Confirm that the candidate meets prespecified inclusion and exclusion criteria |
Apply criteria without knowledge of upcoming allocation |
| Consent and enrollment |
Complete required consent and formally enter the participant into the trial |
Finalize enrollment before treatment assignment is revealed |
| Sequence generation |
Create the allocation sequence using the approved random mechanism |
Protect future assignments from recruiters |
| Allocation administration |
Release the appropriate assignment after enrollment |
Prevent unauthorized access, prediction, or alteration |
| Intervention delivery |
Implement the treatment corresponding to the released assignment |
Follow the assigned condition and document deviations |
These functions may be performed by separate personnel or through a combination of trained staff and secure electronic systems. Their separation should be operationally meaningful rather than merely reflected in job titles.
Who Should Assess Participant Eligibility?
Eligibility should be assessed by trained personnel authorized under the protocol and applicable ethical or regulatory requirements.
Depending on the study, this may involve a clinician, research nurse, investigator, research assistant, or another appropriately qualified professional.
The assessor should use prespecified criteria and record the information supporting the decision.
In clinical trials, certain eligibility judgments may require licensed professional expertise. In educational studies, eligibility might depend on enrollment status, prior coursework, age, or other documented characteristics.
Where judgment is required, the protocol should clarify who has decision-making authority and how uncertain cases are resolved.
Most importantly, the eligibility decision should not be influenced by knowledge of the next treatment assignment.
Who Should Obtain Consent and Enroll Participants?
Consent should be obtained by personnel appropriately trained and authorized for the study, in accordance with the approved protocol and applicable requirements.
Enrollment may be performed by the same person who obtains consent, but these activities should not be confused with treatment assignment.
The research team should define precisely when enrollment becomes final. For example, the protocol may require confirmation of eligibility, completion of informed consent, and entry of mandatory baseline information before randomization.
Not every baseline assessment must necessarily precede randomization, but information used to determine eligibility or stratification should be available at the required allocation stage.
Once the participant is irreversibly enrolled according to the protocol, the assigned intervention can be released.
The enrollment record should identify who completed the process, when it occurred, and whether any exceptions were authorized.
Can the Same Person Enroll Participants and Assign Interventions?
Yes, provided the allocation mechanism prevents that person from knowing or predicting future assignments before enrollment is completed.
For example, a research assistant may screen a participant, obtain consent, enter the required information into a secure randomization system, and receive the assignment after the system confirms enrollment.
If the assistant cannot access the complete sequence, modify the allocation algorithm, or obtain assignments before enrollment, the arrangement can preserve allocation concealment.
By contrast, a recruiter who maintains an accessible spreadsheet containing all future assignments cannot provide the same protection simply by promising not to inspect it.
The critical distinction is between performing several operational steps and possessing information that could influence those steps.
Should the Sequence Generator Be Independent?
Independence is generally desirable because the person generating the allocation sequence may have access to all future treatment assignments.
An independent statistician or trained trial methodologist can create the sequence and transfer it securely to an allocation system without participating in recruitment.
However, independence is not an absolute requirement imposed on every trial. Smaller studies may use appropriately configured systems or qualified collaborators to achieve functional separation.
The key question is whether future allocations remain inaccessible to personnel who can influence enrollment.
The responsibilities involved in generating the randomization sequence should therefore be specified separately from the responsibilities for releasing assignments.
What Is Centralized Randomization?
Centralized randomization uses a secure system or independent allocation service to administer treatment assignments.
Recruitment personnel submit the information required for allocation, and the system releases the assigned condition only after the relevant enrollment requirements are satisfied.
In a multicenter trial, this may involve a web-based platform or an independent coordinating center.
Centralization can reduce opportunities for recruiters to access future assignments and can provide consistent procedures across sites.
Nevertheless, a computerized system is not automatically secure. Its access permissions, sequence storage, assignment logic, and audit records must be appropriately configured and tested.
Where allocation depends on stratification variables, the system should also validate that the required baseline information is entered correctly.
What Is the Correct Order of Enrollment and Assignment?
The appropriate sequence should be established prospectively and implemented consistently.
Confirm eligibility.
Apply the approved inclusion and exclusion criteria and document the decision.
Complete required consent.
Obtain informed consent where required and confirm that participation is voluntary.
Finalize enrollment information.
Record required identifiers, baseline characteristics, and stratification variables according to the protocol.
Request allocation.
Submit the completed enrollment record to the secure assignment mechanism.
Release and implement treatment.
Obtain the assigned condition, document the allocation event, and communicate the information to authorized intervention personnel.
The sequence should not permit recruiters to request an assignment for a hypothetical participant and then decide whether to complete enrollment after seeing the result.
Why Is Allocation Timing Important?
The time at which an assignment is revealed determines whether that information can influence enrollment.
Suppose a research assistant enters a candidate's information into a randomization system before consent has been completed. If the assignment is revealed immediately, the assistant might subsequently influence whether the candidate formally participates.
Similarly, allowing investigators to reserve assignments for participants who have not yet met eligibility criteria can create opportunities for selective allocation.
A secure system should therefore enforce the protocol's required enrollment conditions before assignment release.
Where a participant withdraws after randomization, the event should be documented according to the protocol. The assignment should not simply be erased or reassigned to another participant.
What Happens if a Participant Is Randomized Before Eligibility Is Confirmed?
Randomizing an ineligible participant is a protocol deviation that requires assessment.
The appropriate response depends on the nature of the eligibility criterion, participant safety, treatment exposure, and intended analysis.
Investigators should document the circumstances and consult the relevant trial governance or ethics personnel where necessary.
Removing the participant automatically from every analysis may be inappropriate, particularly if exclusion decisions are made after treatment assignment is known.
Similarly, investigators should not silently substitute another participant or repeat randomization to conceal the error.
The protocol should establish how eligibility errors and other allocation deviations will be managed and reported.
What If a Participant Receives the Wrong Intervention?
An intervention delivery error is distinct from a randomization error.
A participant may be correctly assigned to treatment A but inadvertently receive treatment B. The original randomized assignment remains part of the trial record.
The event should be documented, and any safety implications should be addressed promptly.
The primary analysis may continue to compare participants according to randomized assignment when that approach corresponds to the prespecified estimand.
Additional analyses may examine treatment receipt or adherence, but such analyses require careful consideration because actual treatment receipt may no longer be randomized.
The distinction between assignment and delivery should be preserved in both the database and the final report.
How Can Small Research Teams Maintain Allocation Concealment?
Small teams may not have enough personnel to assign a different individual to every stage of trial implementation.
One solution is to use a secure allocation service that releases assignments only after enrollment. Another is to involve an independent qualified collaborator who maintains the sequence and communicates assignments on request.
Where envelopes are used, they require independent preparation, sequential numbering, opacity, tamper protection, and strict procedures for opening after enrollment.
The appropriate method depends on the trial's resources and complexity.
What should be avoided is an arrangement in which the recruiter can view, select, skip, or modify upcoming assignments.
Operational simplicity is valuable, but it should not come at the expense of a credible randomization process.
How Should Enrollment and Assignment Responsibilities Be Documented?
The protocol should identify who assesses eligibility, obtains consent, enrolls participants, generates the allocation sequence, administers assignments, and delivers interventions.
It should also explain when allocation is released and which personnel can access the complete sequence.
Audit records should document relevant enrollment and allocation events, including timestamps, authorized users, and any deviations.
For blinded trials, the documentation should additionally explain how treatment codes are managed and how authorized emergency unblinding can occur where necessary.
CONSORT 2025 and SPIRIT 2025 provide guidance for describing the randomization process in trial reports and protocols.
Transparent documentation allows readers to assess whether the planned separation of responsibilities was actually maintained.
How Does This Relate to Selection Bias?
Selection bias can arise when treatment allocation becomes associated with participant selection through foreknowledge or manipulation.
For example, a recruiter may be more likely to enroll a participant with favorable prognostic characteristics when the upcoming assignment is to the preferred intervention.
Such a process can undermine the independence between treatment assignment and baseline characteristics that randomization is intended to establish.
The Cochrane RoB 2 framework evaluates the randomization process, including allocation concealment and evidence suggesting that the process was compromised.
Baseline imbalance alone does not prove selection bias. Chance differences can arise after valid randomization.
The strongest evidence about allocation integrity comes from the actual procedures and records rather than from whether the baseline table appears balanced.
Who Is Ultimately Accountable for the Process?
Although responsibilities may be delegated, the principal investigator and relevant trial governance structure retain responsibility for ensuring that the approved protocol is implemented appropriately.
Delegation should therefore be accompanied by training, supervision, and clear documentation.
Personnel should understand which activities they are authorized to perform and how to report deviations or technical problems.
In regulated clinical trials, additional sponsor and investigator responsibilities may apply under relevant good clinical practice requirements.
The organizational arrangement may vary, but accountability should never depend on an informal assumption that someone else is managing the allocation process.
