Conflicts of interest can be either tangible, primarily financial, or intangible, involving academic activities and scholarship like the peer review process . They can lead to biases in research. Federal regulations require institutions establish standards and procedures to avoid conflicts of interest in the design, conduct, or reporting of research. Academic societies and associations have developed similar policies.
Having a conflict of interest is not, by definition, scientific misconduct.
If the potential gain is large, then principles that guide responsible conduct in research may be compromised.
Judgments can be biased unintentionally to favor a desirable outcome.
When large sums of money are involved, it may be difficult for the public, legislators, the judicial system, and even colleagues to be convinced that results were not biased by personal gain. Perceived impropriety can result in consequences as damaging as if intentional misconduct had been committed.
In theory, the adverse consequences of conflicts of interest will eventually be mitigated by the structure of science (e.g., objectivity, blinding of experimenters, repetition of studies, peer review, and disclosure). In practice, this strategy does not address the harms to subjects in clinical trials, misinformation entering the literature, and increased cynicism on the part of the public, the funders of research, and the researchers themselves.
Although it is not possible to avoid all sources of conflict, it is in the best interests of the community of science and the individual scientist to recognize conflicts of interest and to take steps to nullify (e.g., sell shares in the company, turn down the research support, abandon the project) or mitigate those conflicts.
If conflicts cannot be avoided, then those conflicts should be disclosed. As a minimum, the institution and any other parties with a significant interest should be made aware of the extent and nature of the conflict. This includes the audience at meeting presentations as well as journal editors (either for submitting or refereeing manuscripts).
Disclosure is often not enough because of the risks of bias, the temptation for irresponsible conduct, public and regulatory concerns about the possibility of misconduct, and the appearance of impropriety. For every step of the research process, attempts should be made to isolate the conflicted individual(s) from all decisionmaking functions. For example, steps should be taken to maximize the objectivity of patient selection, data collection, the selection of data for publication, and interpretation of the findings. In such cases, management of the conflict of interest means that such functions should be the responsibility of, and/or reviewed by, an unconflicted individual or group.
Competing commitments of time and effort may mean that one responsibility is favored to the detriment of another. Although conflicts of commitment can sometimes be subtle, the problem is still of significant concern. For reasons similar to those described for conflicts of interest, every effort should be made to: (a) recognize conflicts of commitment; (b) attempt to eliminate or minimize those conflicts (e.g., discontinue non-academic activities or limit such activities to a few hours on the weekend); and (c) find mechanisms to manage any conflicts that cannot be eliminated (e.g., disclose the conflicts to responsible university officials and arrange a mutually agreeable system for tracking of time and effort).