Why Publication Governance Needs Explicit Rules
A multi-arm platform can produce several kinds of scientific outputs: reports about the platform as a whole, reports focused on one treatment arm, and analyses that compare an arm with a shared control. These outputs may involve different sponsors, investigators, data, and contractual rights. Without agreed procedures, a sponsor may expect to approve every manuscript containing its arm, while the platform team may expect to publish all results on a common schedule. Neither expectation should be left to assumption.
Publication governance sets out who can prepare and review results, how contributors qualify for authorship, how long review can take, and what happens when parties disagree. Its purpose is not to guarantee favorable findings. It is to support accurate, timely reporting while respecting legitimate confidentiality, participant-protection, and contractual obligations. The governing documents should be agreed before results are known, because rules negotiated after an unfavorable result can create avoidable pressure to delay or reshape reporting.
Define the Output and Its Review Scope
The publication plan should classify outputs before drafting begins. A platform-wide output addresses platform design, shared procedures, or results spanning more than one arm. An arm-specific output reports an individual intervention or comparison, including its relevant control data. Some manuscripts have both features: for example, an arm-specific result may rely on a shared control group or explain a platform-wide allocation rule. The plan should state which governance route applies and how overlapping interests will be handled.
Review scope should follow the content being reviewed, not simply the identity of the sponsor. A sponsor may reasonably check that its product information is accurate and that its protected confidential information is not disclosed. It should not be able to remove valid results from a report, require a favorable conclusion, or prevent publication indefinitely. If a concern is raised, the authors should assess whether it is concrete and within the agreed scope, and document the response.
Before access to results or drafting begins, the platform agreement and publication plan should specify which people or committees receive drafts, which data and documents they may see, and how information from other arms is protected. This builds on the earlier material on data ownership and access rights and confidentiality across competing sponsors: publication review does not create data-access rights that were not otherwise granted. Reviewers should receive only the material needed to assess their defined concerns.
Worked Example: Routing a Platform-Wide Manuscript
Suppose a hypothetical platform has three active arms supported by three sponsors. Investigators prepare a manuscript about recruitment, the shared control policy, and aggregate enrollment across all arms. The plan classifies this as platform-wide because it describes common procedures and includes platform-level information, rather than reporting only one intervention.
The coordinating group circulates the draft simultaneously to the named scientific authors and to each sponsor’s designated reviewer. The plan allows 30 calendar days for review. A sponsor identifies that a table includes an unreleased dosing detail from its arm. The authors check the source and confirm that the detail is confidential and unnecessary to the platform-wide point. They remove or appropriately aggregate that detail, while retaining the enrollment totals and methods needed to support the manuscript.
A second sponsor objects that the aggregate results make its arm look less promising. That objection concerns interpretation, not factual accuracy or protected information. The authors check that the analysis and wording match the underlying results, record the objection and their response, and retain the supported interpretation. The output proceeds through the agreed submission process; the sponsor’s disagreement does not give it a veto.
Authorship and Contributor Recognition
Authorship should reflect intellectual and scientific responsibility, not status, funding, employment, or access to a sponsor’s product. A widely used standard, the International Committee of Medical Journal Editors criteria, requires an author to meet all four conditions: make a substantial contribution to the work’s conception or design, or to data acquisition, analysis, or interpretation; draft the work or critically revise it for important intellectual content; approve the final version; and agree to be accountable for the work, including helping resolve questions about its accuracy or integrity. Journal requirements and institutional policies should also be checked.
The writing group should discuss likely authorship early, revisit contributions as work develops, and confirm the author list before submission. A person who contributed but does not meet all authorship criteria should receive an appropriate acknowledgement, subject to permission and any applicable policy. Group authorship can identify a platform collaboration, but the plan should still make individual responsibility and the complete contributor list clear where the journal requires it.
A sponsor’s funding or provision of an intervention does not automatically qualify a representative for authorship. Conversely, a qualified contributor should not be excluded solely because their contribution came through a sponsor. Authorship decisions should be based on documented contributions and the agreed criteria. A contributor who has not reviewed and approved the final manuscript should not be listed as an author under the criteria above.
Worked Example: Resolving an Authorship Request
In a hypothetical arm-specific paper, a sponsor employee supplied product background and attended two planning meetings, but did not contribute substantially to the design, analysis, interpretation, or drafting. The employee requests authorship because the sponsor funded the arm.
The writing group applies the stated criteria rather than treating sponsorship as a qualification. On the facts given, the employee has not met the substantial-contribution criterion, and has not drafted or critically revised the manuscript. The group explains that funding and background support alone do not establish authorship. If the employee provided a relevant service, the group may propose an accurate acknowledgement, subject to permission. If the employee later makes a substantial intellectual contribution, critically revises the manuscript, approves the final version, and accepts accountability, the group can reassess eligibility against all four criteria.
A Bounded Review and Submission Workflow
A useful publication plan states the responsible roles, review windows, and escalation path in advance. The exact periods depend on the agreement, the type of output, and applicable requirements; the example below is a workable default to adapt, not a universal rule. Statutory reporting, registry requirements, journal deadlines, and other binding obligations must be addressed separately and must not be displaced by an internal review period.
Classify the output as platform-wide, arm-specific, or mixed; identify the writing group, relevant reviewers, data-access basis, and authorship criteria before drafting is complete.
The writing lead sends the manuscript and a clear response deadline to all required reviewers at the same time. The draft should identify the results and materials under review.
For example, allow 30 calendar days for review. Reviewers identify the passage or result at issue, the factual or protected-information concern, and a proposed correction where possible.
The writing group considers each in-scope comment, corrects substantiated errors, and records its response to unresolved objections. Silence by the deadline is handled under the pre-agreed procedure, not treated as endorsement of the findings.
After author approval and required internal checks, submit the manuscript. A narrowly defined additional delay may be allowed for a specific protectable invention or similar issue, but it should have a stated maximum and must not become indefinite suppression.
The logic of a bounded review can be stated directly. Review serves legitimate purposes only if reviewers have enough time to identify a specific concern, and scientific reporting remains useful only if the process has an endpoint. Therefore, the plan should provide a meaningful review window, require a concern to be identified with enough detail for evaluation, and specify what happens when the deadline passes or a disagreement remains. A fixed deadline alone is not sufficient: the plan also needs a defined escalation route and a rule against open-ended delay.
Working Through Sponsor Disagreements
A sponsor disagreement should be separated into its type before it is escalated. A reproducible factual error, an undisclosed confidential detail, a substantiated participant-protection concern, and a difference of scientific interpretation are not the same issue. Authors should investigate evidence-based concerns and correct errors. They should also avoid disclosing information that is genuinely protected and unnecessary to the report. But a sponsor’s preference for a different conclusion is not, by itself, grounds to rewrite the results or stop submission.
A clear escalation sequence is: the writing lead and reviewer first attempt resolution against the agreed criteria; if unresolved, a designated platform publication lead or governance body reviews the specific issue; if the disagreement remains, the decision and rationale are documented and the manuscript proceeds unless a defined legal, ethical, or contractual restriction applies. Any restriction should be narrow, supported, and time-limited where possible. The process should not invent a new approval right after results are known.
Worked Example: A Sponsor Challenges an Arm-Specific Conclusion
In a hypothetical platform, an arm-specific manuscript reports an estimated treatment difference of 1.8 points on a symptom scale, with an uncertainty interval that includes no difference. The sponsor argues that the conclusion should say the treatment is effective because the point estimate favors its intervention. The analysis plan and manuscript describe the estimate and uncertainty accurately.
The writing lead asks the sponsor to identify a factual, analytical, confidentiality, or other in-scope issue. The sponsor instead reiterates its preferred interpretation. The authors check the calculation and wording against the analysis and find no error. They retain the cautious conclusion, which reflects both the estimated direction and the uncertainty, and record the sponsor’s comment and the reason it was not adopted. If the sponsor disputes the analysis itself, the authors can examine the specific methods concern through the stated escalation route; disagreement alone does not justify an indefinite delay.
Timelines, Diagnostics, and Limits
For clarity, the plan can use a calendar such as: draft circulation on Day 0; comments due on Day 30; the writing group’s response and revisions by Day 40; and a decision on unresolved in-scope objections by Day 50. If a specific intellectual-property review is justified, the agreement might permit a further delay of up to 30 days to file a patent application. These dates are illustrative. The important feature is that each interval has a defined start, responsible person, and endpoint. Any extension should be agreed and recorded rather than silently resetting the clock.
A publication process may be failing if drafts repeatedly circulate without a named decision-maker, comments arrive after the deadline without a reasoned extension, the same concern is raised after it has been addressed, or sponsors seek access to unrelated arm-level information. Other warning signs include authorship being offered as a reward for funding, reviewers being asked to approve conclusions, or manuscripts being held without a specific unresolved issue. Documenting the date, comment, response, and decision makes such patterns visible and gives the parties a factual basis for correcting the workflow.
A governance plan cannot guarantee publication in a particular journal or dictate the journal’s editorial decision. It cannot replace applicable reporting obligations, journal standards, or binding agreements. Nor does a publication deadline justify publishing information that would violate a valid protection or compromise participants. The appropriate response is to identify the narrow restriction and seek a compliant way to report the scientifically necessary information, not to assume either unrestricted disclosure or an unrestricted right to withhold results.
| Approach | Advantage | Risk or limitation |
|---|---|---|
| Unstructured, case-by-case review | Flexible for unusual manuscripts | Deadlines and decision rights may be renegotiated under pressure after results are known |
| Pre-agreed review with fixed windows | Predictable, auditable, and compatible with timely reporting | Requires agreement in advance and a workable escalation route |
| Sponsor approval before submission | May appear to provide a simple sign-off point | Can turn review into a veto, blur scientific responsibility, and create indefinite delay |
Common Mistakes
- Equating review with approval. A reviewer can identify a defined concern without having authority to dictate the conclusion or block submission for any reason.
- Promising authorship for funding or seniority. Apply the stated contribution and accountability criteria to every proposed author.
- Using one review route for every output. Platform-wide and arm-specific manuscripts may involve different data and confidentiality boundaries; classify the output and reviewers accordingly.
- Allowing an open-ended hold. A concern should be specific, routed to a named decision-maker, and resolved under a procedure with a defined endpoint.
- Treating silence as scientific agreement. Silence can be handled procedurally after a deadline, but it does not establish that a reviewer endorses the findings.
Check Your Understanding
Answer each question, then compare your reasoning with the short solutions.
- A sponsor requests 60 additional days to review an arm-specific draft but identifies no specific concern. What should the writing lead check before granting an extension?
- A contributor funded the study but did not contribute intellectually, revise the manuscript, approve the final version, or accept accountability. Should funding alone establish authorship?
- A sponsor points out that a table contains a factual dosing error. How should the authors respond?
- A platform-wide paper includes aggregate enrollment information and a confidential detail from one arm. What principle should guide revision?
- Why should the publication plan distinguish a scientific disagreement from a confidentiality or factual concern?
Solutions. (1) Check the agreed review window, whether there is a specific in-scope reason for extension, who can authorize it, and the maximum endpoint; do not reset the clock automatically. (2) No. Apply the authorship criteria to contributions and accountability, and consider acknowledgement if appropriate. (3) Verify the source and correct the error if substantiated, then document the change. (4) Protect the genuinely confidential detail while retaining the aggregate information needed to support the report, if that can be done accurately and consistently with the agreement. (5) The categories call for different responses: correct verified errors and address valid protection concerns, but do not treat a preference about interpretation as a reason to suppress supported findings.