Conflicts of Interest
Core summary
A conflict of interest (COI) exists when a researcher's judgment about their primary obligation (scientific validity and participant safety) could be influenced by a secondary interest (money, career advancement, personal relationships). COIs are not inherently wrong — they must be disclosed and managed so readers and institutions can assess potential bias.
Detailed explanation
Detailed explanation
Financial COIs are the most recognized: consulting fees from a pharmaceutical company whose drug you study, stock ownership in a medical device company, or patent royalties related to the research topic. But conflicts extend beyond money. Intellectual COIs occur when a researcher has strong theoretical commitments — they may be biased toward confirming their own published theory. Personal COIs include studying a collaborator's work, reviewing a paper by a former mentor, or having family connections to the research subject. Disclosure requirements vary by context: for journal submissions, ICMJE requires disclosure of all financial relationships from the past 36 months. For grant applications, NIH requires disclosure of significant financial interests over $5,000. For IRB membership, members must recuse themselves from reviewing protocols where they have a conflict. Management goes beyond disclosure. Institutions may require: independent oversight committees, blinded data analysis by unaffiliated statisticians, restriction from certain study roles, or in extreme cases, prohibition from conducting the research. The goal is not to eliminate all conflicts (that would halt most industry-academic collaboration) but to ensure transparency so that bias can be identified and mitigated.
Clinical example
You receive a $50,000 consulting fee from a company whose drug you are studying in an RCT. You disclose this in your IRB application, grant submission, and manuscript. Your institution requires that an independent Data Safety Monitoring Board oversee the trial and that the statistical analysis be conducted by a blinded statistician at another university.
Research example
A meta-analysis published in JAMA found that industry-funded studies of drugs were significantly more likely to report favorable results than independently funded studies of the same drugs. This does not mean industry studies are all flawed — but it demonstrates why COI disclosure is essential for evidence appraisal.
Knowledge check
Q1. Is having a conflict of interest automatically unethical?
Q2. What is an intellectual conflict of interest?
Q3. How does COI management go beyond disclosure?