Measuring the potency of risk factors for clinical or policy significance

For those in the prevention field who need to use risk factors to identify those at risk, mere statistical significance is not sufficient to establish clinical or policy significance. For that purpose, some measure of "potency" of risk factors is needed, that is, an effect size interpretable and meaningful in a clinical or policy context. The numerous possible candidate measures available often lead to contradictory conclusions. Which of the candidate measures is right or wrong for clinical and policy decision making depends on the fit between the trade-offs between false-positive and false-negative decisions implicit in the definition of the measure (and often unknown to the user) and those related to a specific context of application. (September 1999)