The prevalence and influence of self-reported conflicts of interest by editorial authors of phase III cancer trials.

PURPOSE: To assess the frequency with which editorial authors report personal conflict of interest and investigate a possible association between the presence of such conflicts and favorable opinion.

METHODS: Eligible studies were editorials of phase III clinical trials of antitumor drugs or symptom management published from January 2007 to December 2009 in four top oncology journals. Data from the editorials were collected and each editorial opinion was classified as favorable, neutral or unfavorable with respect to the experimental therapy.

RESULTS: Fifty-four editorials were included: 23 (43%) reported at least one conflict of interest, with the most frequent one being consultancy fees. Conclusions of editorial authors were classified as favorable in 18 editorials (33%). Sensitivity analysis found an association between self-reported conflict of interest of the editorial author and editorial favorable opinion (RR:2.7; 95% CI 1.2-6.1; p=0.019).

DISCUSSION: Besides the high proportion of conflict of interest among editorial authors discussing phase III clinical trials, study interpretation may be influenced by their financial relationships with industry.