Abstract

Quantitative estimates of risk, and their comparison with quantitative estimates of benefit, contribute usefully to decision-making in many fields. In medicine, our assessments of the probability of harm, and of the likelihood of benefit, resulting from many procedures are at present very limited. Moreover, the comparison of risk and of benefit is difficult to make in any quantitative way, whether for a procedure in general or, even more so, for its application in any particular patient. Yet it must be ethically insecure to propose or to use a procedure without some assessment, however approximate, of the hazards involved, and without some indication of whether those hazards will be clearly offset by the likelihood of benefit that should result from use of the procedure