Most pharmaceutical statisticians make a contribution to drug development at the study level by designing and analysing clinical studies. There is less involvement at the drug project level, especially when it comes to making a strategic decision for a drug development program. Motivated by the assessment of drug safety, this paper presents a new decision rule on whether or not to continue developing a new drug based on a cost function. The cost function takes into account both type I and type II errors in a decision making problem. It also incorporates both the cost and benefit from a drug development program. This cost-benefit approach enables company senior management to evaluate a drug program objectively and make decisions in a quantitative and rational manner. The probabilities of making correct decisions are illustrated by using a real data example.