03 · What You Need to Know
A late conflict requires reconstruction, not assumptions
First determine what should actually have been disclosed
Before treating an omitted relationship as a formal violation, establish which policy applied at the relevant time.
Was the interest reportable under the institution's policy? Did the funder require disclosure? Did the journal's disclosure form ask for that relationship? Was the researcher serving in another role, such as peer reviewer or advisory committee member, with separate conflict requirements?
Disclosure standards vary. A relationship omitted from one process may not necessarily have been reportable under another. The relevant policy, time period, thresholds, definitions, and researcher's role therefore matter.
Do not wait for certainty before notifying the responsible authority
If the interest appears potentially reportable, the researcher should not privately investigate for months before informing the body responsible for applying the policy.
Report the facts promptly through the appropriate process and allow the institution, journal, funder, or other authority to determine what additional review is required.
This follows the broader principle that researchers should not be the final judges of their own conflict-of-interest disclosure obligations.
An undeclared conflict and biased research are separate findings
Suppose a researcher should have disclosed company stock but failed to do so. The disclosure failure can be established from the relevant policy and financial relationship.
Whether that financial interest biased the research requires a different assessment.
Disclosure failure
A relevant interest or relationship was not reported when the applicable policy required disclosure.
Unmanaged conflict
A conflict existed without the safeguards that should have been applied to it.
Bias or research impact
The conflict actually affected, or evidence indicates that it affected, research design, conduct, reporting, interpretation, or another consequential process.
These findings may occur together, but one should not be assumed merely because another has been established. The distinction between conflict of interest and bias remains important even when disclosure was late.
Reconstruct when the interest arose and what role the researcher held
A retrospective assessment needs a timeline.
When did the financial or other interest begin? Did it exist when the protocol was designed? During participant recruitment? During data collection? When outcomes were adjudicated? During analysis? When the manuscript was written?
Then identify the researcher's authority at each stage. A conflict arising after data collection but before interpretation presents a different set of possible influences from an interest that existed while the researcher selected outcomes, recruited participants, controlled the dataset, and performed the analysis.
Review the decisions the conflict could realistically have influenced
The retrospective review should focus on pathways of influence rather than merely noting that the conflict existed.
Relevant materials may include the protocol, preregistration, statistical analysis plan, raw or processed data where appropriate, analytic code, outcome definitions, meeting records, manuscript drafts, correspondence, sponsor agreements, disclosure forms, and documentation of who made consequential decisions.
The objective is to determine whether the research record shows evidence that the competing interest affected the study and whether any earlier safeguards existed despite the missing disclosure.
NIH has a specific retrospective-review process for covered FCOI noncompliance
The U.S. Public Health Service financial conflict-of-interest regulations provide a concrete example of how late discovery can be handled. For covered NIH-funded research, when an FCOI was not identified or managed in a timely manner because a significant financial interest was not disclosed or reviewed, or because a management plan was not followed, the institution may be required to conduct a retrospective review.
Current NIH guidance states that the retrospective review must be completed within 120 days of the institution's determination of noncompliance in the circumstances specified by the regulation.
The review documentation includes information such as the project, investigator, entity involved, reason for the review, methodology used, findings, and conclusions.
Policy-Specific Requirement
The NIH 120-day retrospective-review requirement applies within the covered PHS/NIH FCOI regulatory framework. It is not a universal deadline for every undeclared conflict in every country, discipline, institution, or journal.
If bias is found, the response moves beyond late disclosure
Under the NIH framework, if a retrospective review determines that an FCOI biased the design, conduct, or reporting of covered NIH-funded research, the institution must notify NIH promptly and submit a mitigation report.
NIH's policy materials indicate that the mitigation report includes the key elements of the retrospective review together with information about the impact of the bias and actions taken or planned to eliminate or mitigate its effects.
This illustrates an important principle even outside that specific regulation: once evidence of actual impact is identified, the problem is no longer merely that a disclosure form was incomplete. The research itself may require corrective action.
The public record may need correction even when the findings remain valid
If a published paper omitted a relevant conflict, readers may have evaluated the article without information the journal expected them to have.
The appropriate response depends on the journal's policy and the seriousness of the omission. A journal may need to update the article's competing-interest statement, publish a correction or other notice, or take additional editorial action after assessing the circumstances.
A corrected conflict statement does not imply that the results themselves were corrected. It can simply restore information that should have accompanied the paper from the beginning.
Retraction is not the automatic response to an undeclared conflict
An undisclosed conflict can be serious, but retraction should not be treated as the default consequence merely because an interest was omitted.
The central questions concern the reliability and integrity of the research record, the applicable journal policy, the nature of the nondisclosure, and what the subsequent assessment discovers.
If the research remains reliable and the primary problem is an incomplete disclosure statement, correction may be more proportionate. If investigation uncovers serious distortion, unreliable findings, fabricated information, or other major integrity problems, stronger editorial or institutional responses may become appropriate.
Intent matters for accountability, but the record still needs correction
A researcher might omit an interest deliberately. Another might misunderstand the disclosure form, apply the wrong time period, fail to recognize an indirect interest, or simply make an administrative error.
Those circumstances can matter when an institution or journal assesses compliance and accountability.
They do not make an inaccurate public disclosure accurate. Even an inadvertent omission may need correction so readers receive the information that should have been available originally.
Do not rewrite the historical record silently
Quietly editing a laboratory webpage, updating a CV, or adding the interest to future papers does not necessarily address the research produced while the conflict was undisclosed.
The responsible authority needs to know when the interest existed and which studies, decisions, publications, presentations, or reports may have been affected.
Where corrections are required, they should be made through the appropriate institutional, funding, journal, or other formal process rather than through undocumented private edits.
Previously presented findings may require additional disclosure
Under the PHS/NIH framework, there is a specific additional requirement for certain clinical research evaluating the safety or effectiveness of a drug, medical device, or treatment when HHS determines that the project was designed, conducted, or reported by an investigator with an FCOI that was not managed or reported as required. In that situation, the institution must require disclosure of the FCOI in each public presentation of the research and request an addendum to previously published presentations.
This is another policy-specific requirement rather than a universal rule. It demonstrates how retrospective remedies can extend beyond a single journal article.
Future management does not resolve past influence
Once a conflict is discovered, the institution may implement a management plan for ongoing work. That is important if the project, follow-up studies, analyses, or publications continue.
But prospective management answers the question “How do we protect the work from now on?” It does not answer “What happened while the conflict was unmanaged?”
That is why retrospective assessment and prospective conflict management may both be necessary.
Preserve records before trying to resolve the problem
When a potentially significant undeclared conflict emerges, preserve relevant research and disclosure records. Do not delete correspondence, overwrite earlier disclosure forms, alter analysis histories, or reconstruct documents in ways that obscure what existed at the time.
A credible retrospective assessment depends on an accurate record of what the researchers knew, what interests existed, what decisions were made, and when they were made.
The response should be proportional to what the review finds
Late discovery can lead to very different outcomes.
One review may conclude that an interest should have been disclosed but arose only after all consequential research decisions had been completed. Another may find that the conflicted researcher controlled subjective outcome assessment and changed the analysis after seeing unfavorable results.
Treating those situations identically would make little sense. The remedy should respond to the disclosure failure, the nature of the unmanaged conflict, and any demonstrated effect on the research.
