Conflict of Interest Disclosure

A conflict of interest exists whenever a secondary interest could plausibly influence judgement about a primary one, whether or not it does. The disclosable set is broad: research funding, employment, consultancy fees, speaker honoraria, equity, patents, paid expert testimony, and non-financial commitments such as a publicly staked position. Journals collect these on a standard form covering a defined lookback period, and in the United States industry payments to prescribers are published in a federal database. Disclosure is a transparency mechanism only; it neither removes the interest nor licenses it.

The case for taking it seriously is methodological rather than anecdotal. Systematic examinations of industry sponsorship find that sponsored trials more often report results favouring the sponsor than independently funded trials of the same questions, and that the gap is not explained by the usual risk-of-bias domains. It appears instead through comparator choice, dose selection and endpoint definition.

A declared conflict is an instruction about where to look, not a verdict. It directs attention to whether the protocol was registered before enrolment, whether the reported primary endpoint matches the registered one, and who controlled the decision to publish. A sponsored trial with a preregistered protocol remains stronger evidence than an unsponsored one with none.

The asymmetry is where judgement fails. Disclosure regimes exist in journals and, through payment databases, partially in medicine; they do not exist for clinic websites, supplement marketing or podcasts, so sources showing no conflicts are usually those under no obligation to reveal any. Discounting a study because its funding is printed on the first page, while accepting an anonymous claim naming no interest at all, rewards the least transparent source in the room.