Small Gifts, Conflicts of Interest, and the Zero-Tolerance Threshold in Medicine

In their paper "All Gifts Large and Small," Dana Katz, Arthur L. Caplan, and Jon F. Merz (2003) tackle an issue that has been percolating for decades within circles of bioethicists and medical opinion makers; namely, can small gifts to physicians fall below a threshold of ethical, professional, and public concern. It is noteworthy that this paper was funded, in part, by Pfizer, Inc., a leading pharmaceutical company, through an unrestricted gift to the University of Pennsylvania Center for Bioethics. Of course, it should be obligatory for bioethicists, no less than biomedical and clinical researchers, to disclose their sources of funding and potential conflicts of interest. As a preface to my comments on the substance of their paper, I should underscore an inescapable psychological insight about perceived conflict of interest. If the results of their paper were consistent with the sponsor's position on small gifts, some indeterminate but consequential number of readers would be inclined to infer, without seeking further evidence, that the funding, however small, was related to the outcome of the study. Because the study is more like a legal brief (a compilation of evidence in support of an argument) than a test of an empirical hypothesis, which involves original data, the opportunities for a sponsor's funding to influence the outcome, through, let us say the selection of studies, is heightened.

As it happens, the conclusion of the Katz, Caplan, and Merz study—namely that "from a moral and regulatory perspective, policies that determine the acceptability of a gift according to its size are unsound"—appears to be at odds with the values and practices of large pharmaceutical companies such as the one that provided unrestricted support for the study. Thus, our suspicions of a funding effect are immediately put to rest. Moreover, when researchers publish results that are antithetical to the values or practices of their sponsor, we are inclined to assign higher credence to their findings. This is exemplified in the case of Dr. Dong, the pharmacologist at the University of California, San Francisco, who reported that her sponsor's drug exhibited no greater efficacy than a generic agent (Rennie 1997). When the public learned that Dr. Dong refused her sponsor's request to suppress her findings, her results achieved a higher a priori standing than they might have were they consistent with the sponsor's interests.

Readers' suspicions about the influence of funding on research outcome (of which Katz, Caplan, and Merz's paper is a countervailing instance) does not depend on the amount of funding but on the mere existence of it. That is the psychological reality in regard to publications. There is growing evidence of a "funding effect" on the outcome of studies in certain areas of biomedical research (Stelfox et al. 1998; Krimsky 1999; 2003). By this I mean that the for-profit sponsor's interests are a potential biasing factor in the execution and interpretation of research. The general psychological factors at play in the funding effect also apply to gift-giving.

"All Gifts Large and Small" calls attention to two issues about gift-giving to physicians. The first is the voluntary code issued by the Pharmaceutical Research and Manufacturers of America that permits gifts when they amount to less than $100. The second relates to the potential consequences of adopting a low-threshold norm for gift-giving. Little is known about the effectiveness of the voluntary code. If small gifts matter, then large gifts matter more. It is, therefore, of some significance that we understand so little about whether the voluntary code will be effective in stopping the flow of large gifts to physicians. While the issue of large gifts falls outside the scope of Katz, Caplan, and Merz's paper, it should not be lost upon us, even as we evaluate the effects of small gifts on prescribing behavior, that large gifts to...