This presentation is part of: Poster Session - CEA: Methods and Applications; Health Services Research
MORBIDITY COST IN COST-EFFECTIVENESS ANALYSIS
John A. Myers, PhD, Yale University, Biostatistics, New Haven, CT, Steven J. McCabe, MD, MSc, Christine M. Kleinert Institute for Hand and Microsurgery, Hand Surgery, Louisville, KY, and Stephan F. Gohmann, PhD, University of Louisville, Economics, Louisville, KY.
Purpose: The handling of morbidity cost in cost-effectiveness analysis presents a challenge to the current recommendations. The current recommendations argue that morbidity costs are intrinsic when assessing the quality of life for a health state. While acceptance of the recommendation is not universal, there has been no empirical evidence demonstrating if morbidity cost should be included in the numerator or denominator of the cost-effectiveness ratio. The current study aimed to provide the first empirical evidence demonstrating whether or not people automatically consider morbidity costs when assessing the quality of life for a health state. Methods: 181 undergraduate students were randomly assigned to one of two groups: (1) those participants who were not informed of morbidity costs and (2) those participants who were informed of morbidity costs. The participants were additionally randomly assigned a level of illness severity (mild, moderate, or severe). Students were asked to read a description of a health state and to assign an assessment of quality of life for the health state described by the use of the paper standard gamble. Two-factor analysis of variance was performed on the instruments completed. Results: The overall mean Quality Of Life (QOL) for the informed group was significantly lower than that of the uninformed group (p<0.0001, F=24.2, df = 1, 179). Similarly, there is a significant difference between illness severity levels in mean QOL (p<0.0001, F=29.5, df = 2, 178). No statistically significant interaction between level of illness severity and prior knowledge was observed (p=0.5904, F=0.53, df = 2, 178). Therefore, we fit a model removing the interaction term. Conclusion: The current study demonstrated that those subjects informed of morbidity costs score quality of life lower than subjects uninformed of morbidity costs. Morbidity costs are part of the cascade of events that result from an intervention. Therefore, morbidity costs are part of the effectiveness of an intervention and should be included in the denominator of the cost-effectiveness ratio. To accurately represent the effectiveness of an intervention, we argue morbidity costs should be included in the description of health states. We recommend that descriptions of health states include morbidity costs when conducting a cost-effectiveness analysis.
