Can We Trust Positive Findings of Intervention Research? The Role of Conflict of Interest.

In recent years, there has been increased attention to the issue of conflict of interest within prevention research. The aims of this paper are to discuss these developments and to relate them to discussions of conflict of interest in the broader scientific literature. Although there has been concern expressed about the extent to which conflicts of interest can be defined and measured, empirical research suggests that financial conflicts can be easily identified and assessed in meta-analyses focused on their effects on research quality. Research evidence also shows that conflict of interest is associated with use of flexible data analysis practices and the reporting of chance positive findings, both within prevention research and related disciplines such as public health and psychology. However, the overwhelming majority of published studies report positive results, and there are a number of other influences within academia (such as pressure to publish) that account for this and for the use of flexible data analysis practices. Accordingly, introducing measures to improve research quality in general, rather than just focusing on problems specific to research in which there is a clearly identifiable conflict of interest, may prove more effective and less controversial. Most such efforts focus on introducing greater transparency into research design, practice, and reporting. These both curtail employment of flexible data analysis practices and make their use transparent to investigators seeking to assess their effects on research quality. Also, requiring detailed disclosures of conflicts be reported by all investigators (not just senior authors) would improve current disclosure practices.