Comparing therapeutic benefit and risk.

Theoretically, the merit of a medical drug can be quantitatively determined by calculating benefit and risks as the drug-attributed gain or loss, respectively, of quality-adjusted life years. This calculation is based on comprehensive data concerning large numbers of patients. For most drugs, however, the data available are more or less incomplete. Therefore, large-scale intensive and continuous data collection needs to become routine in medical and pharmaceutical practice. In addition, important scientific, ethical, logistic and financial obstacles need to be overcome. Preliminary evidence suggests that the 'Principle of Threes' is a useful tool for categorizing the merit of drugs and for comparing drugs of a given therapeutic class. The safety of a drug is a dynamic rather than a constant feature. The merit of a drug is not an absolute quality, but is meaningful only in a given context. It is still uncertain whether merit assessment can present the dilemma that 'a drug causes benefit in many at the cost of serious injury in some' in a way the individual user can understand. Quantitative merit assessment will be welcome but should not be expected to replace individual therapeutic decision making. It remains to be seen whether merit assessment is to result in a gold standard for the acceptability or unacceptability of drugs.