A variety of economic and ideological forces threaten the integrity of public health research.1^–4 In response, the editors of the top biomedical journals now require conflict of interest disclosures for submissions. Their increasingly extensive disclosure policies have resulted in some progress towards protecting the integrity of public health research by disclosing for readers potential sources of author bias.5 6

Yet the journal editors appear to be reaching the limits of what they can realistically accomplish. Letters to the editor reporting undisclosed conflicts,7 editorial retractions of non-compliant articles8 and occasional news scandals9 10 are still far too common. Several empirical studies have also confirmed problematic levels of author non-compliance with journal disclosure policies.11 12 And journal editors themselves seem frustrated by the scope of the conflicts problem and their limitations in addressing it. As Catherine DeAngelis, editor-in-chief of JAMA, explained: “I’m not the F.B.I.” and cannot look “in the hearts, minds and souls of authors”, much less oversee the completeness of the disclosures accompanying the over 6000 articles submitted annually to her journal.13 Moreover, as Dr DeAngelis notes, even if JAMA could miraculously identify all …