Conflict of Interest Disclosure in Medical Manuscripts: What a 2026 Study Reveals and What Authors Must Do

A study published in Research Integrity and Peer Review in July 2026 searched the Retraction Watch database from its inception through April 23, 2026, for every editorial notice where conflict of interest was explicitly cited as a reason for the action. The database returned close to 900 matches. In 95 of those entries, roughly 11 percent, competing interest was the sole stated reason for the retraction, correction, or expression of concern. The remaining 800-plus mixed it with other causes: data fabrication, plagiarism, ethical violations, peer review manipulation. But in those 95 cases, a financial or relational conflict that should have been declared on the submission form was, by itself, enough to trigger an official editorial response.

That number is probably an undercount. The Retraction Watch database is comprehensive but not exhaustive, and editors do not always list every contributing cause in a notice. Still, the study gives concrete shape to something that has long been a background assumption in medical publishing: the competing-interests statement is not a formality you complete after the science is done. It is part of the science.

The ICMJE issued its updated recommendations in January 2026, and one of the clearer signals in the revision is that the committee is no longer treating vague disclosure as acceptable. If you are preparing a clinical paper, a meta-analysis, or a review this year, here is what the current expectations mean for how you think about and write your competing-interests section.

Why Conflict of Interest Leads to Retractions

Authors sometimes assume that an undisclosed conflict of interest is a bureaucratic failure rather than a scientific one. If the data were collected properly and the analysis was honest, what does a consulting arrangement from two years ago matter? The journals that retract papers for this reason have a clear answer: the reader needed that information to interpret the paper, and the authors withheld it. That is a breach of trust between authors and the scientific community that cannot be undone by showing the underlying data are clean.

There is also a documented link between financial relationships and research outcomes that makes the disclosure problem harder to dismiss. Studies examining industry-funded clinical trials consistently find higher rates of positive results compared with trials with no industry funding. The effect appears across cardiovascular research, oncology, and device studies. The causal pathway is contested and probably multiple, ranging from selective publication to unconscious bias in outcome selection. But editors read this literature, and that reading shapes how they respond when a financial relationship surfaces after publication.

The 32 full retractions in the 2026 study represent papers that could no longer stand in the literature because the editorial team, after learning about the undisclosed relationship, could not confidently certify the integrity of the work. In some cases, post-publication investigation found additional problems. In others, the retraction was a direct response to the disclosure failure itself. Both outcomes damage the author's record in ways that are permanent and searchable.

What the ICMJE Counts as a Conflict of Interest

The ICMJE's definition, which has been stable for several years and was reaffirmed in the January 2026 recommendations, is deliberately broad: a conflict of interest exists when professional judgment concerning a primary interest, such as patients' welfare or the validity of research, may be influenced by a secondary interest. The word "may" is doing significant work. You do not need to believe your judgment was actually influenced. The question is whether a reasonable reader, knowing the relationship, would think it could have been.

Financial conflicts are the most straightforward category: employment by a company with a stake in your results, equity holdings, grants, consulting fees, speaker's bureau payments, honoraria, travel reimbursement, patents, royalties, and paid expert testimony. Most authors who run into trouble in this category have not tried to conceal anything. They have simply forgotten that the consulting arrangement from 30 months ago is still within the disclosure window, or they assumed that a speaker's fee for a disease education talk does not count because it was not specific to the drug under study.

Non-financial conflicts occupy a grayer space. Personal relationships between authors and the subjects of their work, academic rivalries that could affect how competing theories are treated in a review, intellectual or advocacy commitments that predate the research and might bias how results are framed, membership on advisory boards for nonprofit organizations with positions on the clinical question, and leadership in professional societies that have endorsed a particular treatment approach. These are all potentially disclosable. The ICMJE 2026 recommendations are more explicit than previous versions that these relationships belong in the disclosure, not in the author's private judgment about whether they are material.

The ICMJE Disclosure Form: What It Asks and How to Complete It

Most ICMJE member journals, including the New England Journal of Medicine, JAMA, The Lancet, BMJ, Annals of Internal Medicine, and the Canadian Medical Association Journal, require each author to complete a separate ICMJE disclosure form. This is not a box to check. Each co-author fills out their own copy, and the corresponding author collects and submits them as a set. If your submission system asks for a single combined statement, that statement should reflect all authors' disclosures, and the individual forms should be available if the editorial office requests them.

The current form, which the ICMJE revised in 2023 and has maintained in the 2026 update, asks about activities in the past 36 months. That is three calendar years, not the duration of the study. A consulting arrangement that ended before the study began can still fall within the window. The form also distinguishes between payments to you as an individual and payments to your institution for work you performed. Both categories are disclosable. Authors sometimes assume that money paid to their hospital or university rather than directly to them is not theirs to declare. It is.

One practical difficulty is that forms need to match. If your individual form says "none," but a co-author's form lists a relationship with the same company that manufactured the device being studied, that asymmetry will be visible to editors and reviewers. Talk to your co-authors before submission. Differences in what people recall and how they interpret the instructions are common, and it is far better to catch and resolve them internally than to have an editor notice them during review.

Where in the Manuscript the Disclosure Goes

Placement matters. Most journals want the competing-interests statement in the body of the manuscript at a specific location, typically after the acknowledgements and before the references, or in a dedicated section near the end. A disclosure buried in the cover letter is insufficient because the cover letter is not part of the published record. Readers who encounter your paper six years from now will not see your cover letter. The disclosure must travel with the paper.

If you used a methodology or tool that has proprietary components, and the company making that tool provided any support to the study, that relationship may need to be declared both in the competing-interests section and in the methods, where readers evaluate your choices. This is particularly relevant for clinical trials of devices, diagnostic tests, and software-based clinical decision tools, where the tool itself is the intervention. A blanket competing-interests statement at the end of the paper is not a substitute for methodological transparency within the text.

Some journals have begun asking for funding source statements and competing-interests statements as two separate items rather than combining them. Read the current instructions carefully. A statement that lists your grants is not the same as a competing-interests declaration, even if both mention the same organization. Grants may or may not constitute a conflict depending on whether the funder had any role in the study design, data collection, analysis, or decision to publish. When a funder had any of those roles, that too needs to be described.

Financial Conflicts: The Most Common Omissions

Speaker's bureau arrangements are among the most frequently undisclosed relationships, partly because authors frame them as educational rather than promotional. Being paid to present a company's drug or device to physicians at continuing medical education events is a financial relationship with that company, and it belongs in the disclosure regardless of whether you believe the content you delivered was scientifically neutral.

Equity holdings create a different problem. Authors in academic medicine often hold small positions in startup companies they have advised or co-founded. If those companies have any connection to the clinical area being studied, the equity relationship is disclosable. The difficulty is calibrating what "any connection" means. A cardiovascular surgeon who holds shares in a medical device company and publishes a study on a different class of cardiac devices faces a question a journal cannot answer for you. The conservative answer is to disclose and let the editor determine materiality. The risky answer is to decide unilaterally that the connection is too remote to mention.

Consulting fees and advisory board payments are often underreported because they feel transactional rather than relational. Serving on a scientific advisory board for a pharmaceutical company, even briefly, and receiving any payment for that service, creates a disclosable relationship with that company. The relationship persists in the disclosure window even after it ends. Authors who served on an advisory board two years ago and then wrote a paper touching that company's product area without disclosing the prior arrangement have, by the ICMJE definition, filed an incomplete form.

Non-Financial Conflicts: The Gray Areas Most Authors Skip

Non-financial competing interests receive less attention and generate fewer retractions than financial ones, but they appear more often than authors acknowledge. The clearest cases involve clinical guideline authorship. If you co-authored or contributed substantially to a clinical guideline recommending a particular treatment approach, and you subsequently publish a trial whose design or primary outcome was shaped by that guideline, readers should know about your role in the underlying recommendation. That relationship is not financial, but it is a prior intellectual commitment that contextualizes your new paper.

Leadership in professional societies presents a similar question when the society has an institutional position on a clinical issue your paper addresses. An author who serves as president of a cardiology society that advocates for early invasive management of a specific condition, and who then publishes a trial on that management strategy, has a non-financial relationship that readers would want to know about. Journals differ on how strictly they apply this, but the general principle in the 2026 ICMJE recommendations is that authors should err toward disclosure when the relationship is visible and relevant.

Academic rivalry, while rarely disclosed, occasionally comes up in post-publication challenges. Review articles and meta-analyses are particularly vulnerable to this criticism because authors make choices about which studies to include, how to weight methodological quality, and how to frame consensus and controversy. If you have a documented public disagreement with another research group whose work you are evaluating, that context is potentially relevant to readers. It does not require disclosure under current ICMJE language, which is more cautious on this point than on financial relationships, but being aware of the perception is useful.

Institutional Conflicts of Interest: The Category Authors Forget

Individual-level COI disclosure is the focus of most author guidance, but institutional conflicts are increasingly on editorial radar. Your university may hold patents on a molecule or device that is the subject of your research. Your department may have an industry-sponsored research agreement with the company whose product you are studying. Your hospital may have a preferred-vendor relationship or a clinical partnership that creates financial alignment. These are not your personal relationships, but they are relationships that could plausibly influence the research environment you work in.

Most journals do not yet have standardized fields for institutional COI in the submission system, which makes this category easy to overlook. But several high-profile journals, including the NEJM and JAMA, have begun asking explicitly about institutional-level financial arrangements in their instructions, and the ICMJE 2026 language now explicitly acknowledges institutional interests as disclosable. The safest approach is to ask your institution's research integrity or conflict of interest office whether any relevant agreements exist before you complete the submission.

Multisite clinical trials create additional complexity because the relationships that exist at one participating institution may not be visible to investigators at others. The coordinating center, which typically leads the submission, may not know that a co-investigator at a distant site sits on the advisory board of a company with a stake in the results. Establishing a COI disclosure process among all co-investigators before the analysis is complete, rather than asking people to complete forms under deadline pressure, reduces the chance that something important is missed.

When a Reader or Journalist Finds What You Missed

The most stressful version of this situation starts not from an editor's own review but from an external report. Investigative health journalists and science integrity researchers have become adept at cross-referencing author names against Open Payments, the US Centers for Medicare and Medicaid Services database that lists pharmaceutical and device company payments to physicians. ProPublica's Dollars for Docs tool and similar databases outside the United States have enabled readers to verify COI statements against public records in a way that was not practically possible a decade ago.

When a discrepancy surfaces post-publication, journals generally begin by contacting the authors for an explanation. In straightforward cases where the omission appears inadvertent and the underlying research is unaffected, a published correction noting the revised competing-interests statement may be sufficient. Where the omission appears systematic, where the hidden relationship is substantial, or where it connects to other concerns about the paper, the outcome is more likely to be an expression of concern or retraction. Authors have no control over which path the editorial office chooses once the issue has been raised.

A correction for an incomplete COI statement is embarrassing but recoverable. A retraction is not. The 32 full retractions in the 2026 Retraction Watch analysis represent permanent removals from the literature, searchable alongside the authors' names in perpetuity. In competitive clinical specialties where grant committees and promotion panels conduct publication record reviews, a retraction for a disclosure failure is a reputational event that follows a career.

A Pre-Submission Self-Check for Every Co-Author

The most effective way to avoid COI disclosure problems is to treat the disclosure process as a structured team activity rather than an afterthought at the submission stage. That means circulating the questions below to every co-author before the manuscript is finalized, not after it has been revised and accepted. Timing matters because authors' recall of past relationships and their interpretation of what counts as a conflict tend to be more careful when the stakes feel lower.

If any co-author answers yes to any of these questions, the relationship should be recorded and matched against the target journal's specific requirements for what to disclose, how to phrase it, and where to place it in the manuscript. If a co-author is unsure, the default should be disclosure. A competing-interests statement that mentions a relationship later judged to be immaterial costs nothing. One that omits a relationship later found to be relevant can cost everything.

Worth noting: several journals now use the Open Payments database to cross-check US physician author names before peer review is assigned. It takes about three minutes and surfaces any payment above a de minimis threshold. If you publish in a field where your name is in that database, assume that a thorough editor or a persistent reader will check. Write your disclosure accordingly.

Practical Wording for the Competing-Interests Statement

Clear competing-interests statements follow a simple structure: name the relationship, name the party, say when it occurred, and close with responsibility. They do not hedge or explain away. An editor reading your statement should be able to locate it in thirty seconds and extract the relevant relationship without any interpretive work.

Notice what these examples do not include: apologies, explanations of why the relationship did not affect the results, or assertions that the conflict is minor. Those additions invite a reader to debate your framing. A neutral statement of fact does not.

Further Reading