Disclosure of conflicts of interest by authors of clinical trials and editorials in oncology

Purpose There is concern that financial relationships between sponsors and investigators may bias research results. Our objective was to evaluate the epidemiology of conflicts of interest (COIs) among authors of clinical trials and editorials in oncology and the relationship between COI disclosure and source of funding. Methods We did a cross-sectional survey of clinical trials and editorials of anticancer agents and supportive care medications published in the Journal of Clinical Oncology (JCO) during a 1-year period. Results Of 1,533 articles published in JCO between January 1, 2005, and January 31, 2006, 332 met our inclusion criteria; 289 (87%) were clinical trials, and 43 (13%) were editorials. The pharmaceutical industry entirely or partially funded 44% of the clinical trials. At least one COI was disclosed in 69% of clinical trials and 51% of editorials. The most common types of COI reported by authors were consultancy fees, honoraria, and research funds. The highest monetary levels of interest reported by authors were for research grants, but the majority of authors with COIs received less than US$10,000. In multivariable analysis, authors of clinical trials conducted in North America (North America v Europe: odds ratio [OR] = 2.9, P = .002) and authors of trials funded entirely (industry only v nonprofit: OR = 13.8, P < .001) or partially ( both industry and nonprofit v nonprofit only: OR = 5.8, P < .001) by industry were more likely to report personal COIs. Conclusion COIs are common in clinical cancer research and usually take the form of consultancy fees, honoraria, and research funds. Source of study funding was significantly associated with COI disclosure.