The cost-effectiveness of health-care interventions is often evaluated using quality-adjusted life years (QALYs) as a measure of outcome. However, QALYs are valid only under several restrictive assumptions. Furthermore, QALYs are ethically controversial, as they receive their strongest support from utilitarian theory, which is often considered an unacceptable ethical theory. The purpose of this work is to present a non-utilitarian approach to cost-effectiveness analysis, which avoids calculating QALYs, but still is able to aggregate and compare different clinical outcomes. By capturing benefits in terms of adverse events (AEs) avoided, the approach is based on one of the fundamental metrics of clinical epidemiology and thus moves the assessment of cost-effectiveness closer to that of clinical outcomes in clinical trials. Furthermore, it directly incorporates the two most important ethical values with regard to setting priorities in health care, ie, a concern for health gain as well as for health without treatment. The approach aggregates the different types of AEs avoided, by introducing weights that reflect their value. In order to project weights on an interval scale, ranking data, the time trade-off, or the standard gamble method can be used.