Introduction

Bioethics has to date largely consisted of analyses of the moral underpinnings of appropriate behaviour in science and medicine. Although such theoretical writings often make important contributions to public health, they often contain unfounded assumptions about human behaviour. I describe recent examples in which assumptions made in theoretical writings were later repudiated, or in which empirical work served further to clarify positions taken in theoretical pieces. Paralleling the rise of evidence based approaches to clinical medicine, these examples show that further emphasis on empirical analysis is needed to promote more balanced, and evidence based, bioethics.

In 1988 Benjamin Freedman noted that perception, rather than reality, controls the generation and resolution of ethical issues.1 Writing during the first decade of the HIV/AIDS epidemic, he was referring to how the debate on doctors' duties to render care to AIDS patients was influenced by doctors' perceptions of the risk that patients might transmit the virus to them, rather than the actual risk conferred, which was not yet known. In the absence of evidence, intuition and perception have useful roles.

But what ought to guide ethical deliberations once evidence becomes available? Although evidence emerged that the risk of HIV transmission from patient to doctor is extremely low,2 3 resolution of many doctors' fears lagged behind and perception led several doctors to continue refusing to treat patients infected with HIV. I argue that, although Freedman correctly noted that perception often controls the generation as well as the resolution of ethical issues, this statement is only half true when placed in the normative realm: perception ought to govern the generation of ethical issues, but reality must often govern their resolution.

The rise of evidence based medicine

Traditional …