Application of Quality of Life and Weli-Being Measurement as Input into Cost-Effectiveness Analyses

For the most part, cost-effectiveness analysis has attempted to incorporate the effects of medical treatments on patients' quality of life by adjusting years of survival with an index of the quality of life in the surviving years. The construction of such a quality-adjusted year index is typically based on implicitly value-laden assumptions regarding tradeoffs of hypothetical situations, ie, survival periods with or without a disabling condition. The construction of the index relies on judgments and other procedures that have been shown to generate figures that are widely discrepant from those suggested by empirical studies. Indeed, only empirical studies based on individuals' direct assessments of their quality of life, and those which focus on the persons subjected to the treatments at issue, can provide the necessary valid input to cost-effectiveness analysis.

It is therefore suggested that cost-effectiveness analysis should rely on quality-of-life data generated by empirical studies of persons subjected to the treatments; and it should report separately the cost of treatment per year of life and per quality of life expressed in effect size units. The latter effect size units in quality of life should be specified separately for global measures of physical, mental, and social functioning and well being, as well as for specific measures of the presenting symptoms and disabilities that are produced in these areas of life by the disease. The degree of relationship among these various aspects of quality of life can provide yet another important indication of the overall effect of the treatment. The advantages of focusing on quality-of-life data in terms of effect size units stem from: (a) the greater convergence in the conceptualization of quality of life than in the use of specific measures, and (b) the availability of meta-analytic techniques for synthesizing effect size units from different studies and comparing them across different treatments.