03 · What You Need to Know
How Protocol Amendments Work After Stage 1 Acceptance
Why Does Stage 1 Acceptance Create Special Responsibilities?
During Stage 1 review, the journal evaluates the proposed research question, design, measurement procedures, sampling plan, and intended analyses before the principal results are known.
In-principle acceptance establishes a conditional publication commitment based on that evaluated proposal. The researchers are subsequently expected to conduct the study in substantial accordance with the approved protocol and report the findings transparently.
This creates a different situation from ordinary preregistration. A preregistered study may be amended while preserving the original time-stamped record, but an accepted Registered Report also involves an editorial commitment to a particular reviewed design.
Consequently, a major change may require more than an updated registration. It may need the journal's agreement that the revised study remains capable of addressing the approved research question.
The conditions governing in-principle acceptance and final publication therefore provide the starting point for deciding how to manage amendments.
What Counts as a Change to the Approved Protocol?
A protocol change occurs when the study's actual or intended procedures differ from those approved at Stage 1.
Changes may concern recruitment, eligibility, intervention delivery, measurement, sample size, analysis, or other aspects of implementation.
Not every difference has the same scientific importance. Rescheduling an interview because a participant is unavailable is different from replacing the study's primary outcome after observing the treatment effects.
| Type of change |
Illustration |
Potential significance |
| Administrative adjustment |
Changing internal staff assignments without altering study procedures. |
Usually limited inferential implications. |
| Operational modification |
Changing the delivery platform while retaining the intended intervention. |
May affect implementation fidelity or participant experience. |
| Recruitment amendment |
Extending recruitment or changing eligibility criteria. |
May affect sample composition, feasibility, or statistical inference. |
| Measurement change |
Replacing an approved assessment instrument. |
May change the construct, measurement properties, or comparability. |
| Design modification |
Changing the allocation procedure or comparison condition. |
May alter internal validity and the intended estimand. |
| Analysis amendment |
Changing the primary statistical model. |
May affect confirmatory inference and interpretation. |
| Research objective change |
Replacing the principal hypothesis or research question. |
May fundamentally alter the study accepted at Stage 1. |
These categories are illustrative rather than formal classifications applied uniformly by journals. The appropriate response depends on the actual consequences of the proposed revision.
Which Changes Usually Deserve Editorial Consultation?
Researchers should consult the journal's current Registered Report policy and IPA correspondence. Where those documents do not clearly address a consequential change, contacting the handling editor before implementation is generally prudent.
Particular attention is warranted when the revision affects the primary outcome, the principal hypothesis, sample-size target or stopping rule, intervention or comparison conditions, participant eligibility, randomization, or confirmatory analysis.
These decisions may have influenced the journal's original assessment of whether the study could produce informative evidence.
By contrast, minor administrative changes that do not affect participant protection, measurement, implementation, or inference may not require the same level of review. Nevertheless, the journal's written requirements take precedence over assumptions about what is minor.
Researchers should not assume that silence from the journal constitutes approval. Where formal authorization is required, obtain a clear editorial decision.
Does the Timing of a Change Matter?
Yes. Timing influences whether a revised decision could have been shaped by the evidence it is intended to evaluate.
| When the change occurs |
Main concern |
Recommended response |
| Before relevant data collection |
Whether the revised design remains scientifically adequate. |
Document the rationale and obtain required approvals before implementation. |
| During collection, before examining relevant outcomes |
Potential influence of recruitment or operational information. |
Record the information available and explain why the revision is necessary. |
| After blinded data-quality review |
Whether blinding adequately prevented outcome-informed selection. |
Describe the blinding arrangements, diagnostic evidence, and amendment. |
| After examining relevant results |
Risk of selecting procedures because they favor particular findings. |
Disclose outcome access, consult the editor, and distinguish revised analyses from original confirmatory commitments. |
Access to a dataset does not always mean access to the relevant treatment effects or hypothesis tests. Nevertheless, researchers should specify what information was available when the decision was made rather than simply claiming that the amendment occurred before analysis.
Can You Change the Sample Size or Recruitment Stopping Rule?
Potentially. Recruitment problems are among the foreseeable challenges of implementing an approved study.
A study may encounter lower-than-expected participation, institutional closures, unexpected attrition, or changes in the availability of eligible participants.
Extending recruitment because of a documented feasibility problem can have different inferential implications from extending recruitment after examining a nearly significant treatment effect.
Researchers should identify the original target, achieved recruitment, proposed revision, and the consequences for the intended analysis. If interim results have been examined, the statistical framework may require modification.
Changing sample-size and stopping commitments may affect assumptions underlying the Stage 1 evaluation, so substantial revisions should be discussed with the journal.
A smaller achieved sample does not automatically make the study worthless, but it may reduce precision or compromise the ability to answer the approved question.
Can You Replace the Primary Outcome or Measurement Instrument?
A measurement change may be justified when the approved instrument proves unavailable, defective, unsafe, or otherwise unsuitable.
However, the replacement should not be selected merely because it produces a more favorable result.
Researchers should assess whether the alternative measures the same construct, whether its measurement properties are suitable, and whether the change affects comparability across participants or study phases.
Replacing an instrument is not necessarily equivalent to replacing the primary outcome. Two instruments may operationalize the same construct differently, while a change in outcome may alter the study's central research objective.
Because the primary outcome specification is often central to Stage 1 acceptance, consequential changes require particular transparency and potentially editorial approval.
What If the Approved Statistical Model Becomes Inappropriate?
A statistical model may prove unsuitable because of an unforeseen data structure, an implementation problem, or assumptions that cannot reasonably be maintained.
For example, an approved analysis may assume independent observations, but implementation may reveal an additional clustering structure requiring appropriate treatment.
Researchers should investigate the methodological issue and determine whether the original analysis remains interpretable. When a revision is necessary, explain why the alternative is more appropriate.
Where meaningful, reporting both the approved and revised analyses may help readers understand the consequences of the change. However, an invalid analysis should not be presented as authoritative merely because it was approved prospectively.
Changes should be distinguished from additional exploratory or sensitivity analyses. Replacing a required confirmatory model raises different issues from adding an analysis to investigate robustness.
The principles for prespecifying design and analysis decisions remain relevant when evaluating the consequences of an amendment.
What Happens When an Ethical or Safety Concern Requires Immediate Action?
Participant protection takes precedence over preserving an unchanged research protocol.
If a procedure presents an unexpected safety risk, confidentiality concern, or serious ethical problem, researchers should take appropriate protective action and follow applicable institutional or regulatory requirements.
Depending on the circumstances, this may involve suspending recruitment, modifying procedures, notifying an ethics committee, or implementing an urgent safety measure.
Researchers should document what happened and inform the journal as required. Editorial approval does not replace ethics or regulatory authorization.
Conversely, researchers should not delay necessary participant protection merely because a journal has not yet responded to an amendment request.
Watch Out
Journal approval, ethics approval, and regulatory authorization are distinct. A change accepted by the handling editor may still require approval from an institutional review body or another responsible authority before implementation.
How Should You Request Approval for a Consequential Amendment?
A useful amendment request allows the editor to evaluate the change without reconstructing the entire study history from scattered correspondence.
Explain the approved procedure, the problem encountered, the proposed replacement, the scientific rationale, and the consequences for the study's original objectives.
Also identify the stage of data collection, whether relevant outcomes have been examined, and whether the change requires other approvals.
Where appropriate, provide revised protocol text, an updated analysis plan, or supporting feasibility evidence.
Do not implement a nonurgent consequential change before receiving required approval. Retain the editorial decision and incorporate the approved amendment into the study's documentation.
Does Every Amendment Require Another Full Stage 1 Review?
No universal rule applies. The journal may handle a minor amendment through editorial correspondence, request targeted reviewer input, or require more extensive evaluation when the proposed revision substantially changes the accepted study.
The appropriate procedure depends on the nature of the amendment and the journal's policy.
A small operational correction may leave the original scientific rationale intact. Replacing the primary research question or fundamental design could require reconsideration of whether the existing IPA remains appropriate.
Researchers should avoid assuming that a revised protocol automatically retains the same acceptance conditions.
Can an Amendment Cause the Loss of In-Principle Acceptance?
Yes, in some circumstances. IPA is a conditional commitment to a study evaluated at Stage 1.
If a proposed change fundamentally alters the research question, removes essential methodological safeguards, or makes the study incapable of answering the approved question, the journal may determine that the original acceptance conditions are no longer satisfied.
However, a justified amendment does not automatically invalidate IPA. The consequences depend on its significance, the adequacy of the revised design, and the journal's procedures.
Researchers should distinguish an unavoidable methodological problem from a discretionary change made because the results are disappointing.
The latter creates greater concern about results-dependent research decisions.
How Should Changes Be Reported at Stage 2?
The Stage 2 manuscript should accurately distinguish the approved protocol from the procedures actually implemented.
For consequential deviations, report what changed, why the change was necessary, when it occurred, whether relevant results were already known, and whether the journal or other authority approved it.
Where feasible, identify the effects of the change on the sample, measurements, analysis, or interpretation.
Additional analyses may help evaluate the robustness of conclusions, but they should be labeled according to their actual role and timing.
Researchers should not retrospectively rewrite the methods as though the amended procedure had always been the approved plan.
The broader distinction between exploratory analyses and approved confirmatory analyses remains important during Stage 2 reporting.
What Documentation Should Be Preserved?
Maintain the approved Stage 1 manuscript, IPA correspondence, original analysis plan, amendment requests, editorial decisions, and relevant ethics approvals.
A dated deviation log can identify the affected procedure, reason for the change, implementation date, responsible personnel, and implications for analysis.
Where appropriate, version-controlled protocols and analysis scripts can help reconstruct the research process.
These records should be managed securely, particularly when they contain confidential participant information or sensitive institutional details.
The objective is an auditable history of consequential decisions, not an administrative record so elaborate that it becomes impossible to maintain.
