Project 3: Economic Evaluation

König et al.
Economic evaluation of cognitive behavioral therapy for positive symptoms in psychotic disorders

Background and Aims
Demographic change and technological progress will lead to a growing demand for health services in Germany which is confronted with limited resources. Hence, the cost-effectiveness of single health services increasingly gains importance. In order to estimate cost-effectiveness, the ratio of the additional costs and health effects of a health service is computed. Regarding costs, one can differentiate between direct costs (costs for medical treatment, e.g. hospital stays, drugs) and indirect costs (disease related lost productivity, e.g. sickness absence or reduced productivity).
In this project we conduct a full economic evaluation of cognitive behavioural therapy (CBT) from a societal perspective. Therefore we measure the direct and indirect costs as well as health effects of CBT and the comparator "supportive treatment" (ST) and calculate the incremental cost-effectiveness ratio (ICER). The ICER describes the ratio of the differences in costs and health effects of CBT compared to ST. Thus, an ICER corresponds to the costs for one health effect gained by conducting CBT instead of ST.

Methods and Instruments
The direct and indirect costs of both study arms are calculated prior, during and after therapy using a modified „Client Sociodemographic and Service Receipt Inventory" (CSSRI). The CSSRI assesses the overall resource utilization of patients as well as productivity losses. To estimate costs these quantities are then valued with market prices. When market prices are not available administrative prices or mean costs are used to estimate so called "shadow prices".
To assess the health effects of CBT and ST, two different measures are used. On the one hand the EQ-5D, a generic measure of subjective health related quality of life comprising a health profile and a visual analogue scale is used prior, during and after therapy to estimate quality adjusted life years (QALYs). On the other hand an objective measure of positive symptoms (PANSS-Score) is used to quantify treatment response.
A Markov model is built to estimate the ICER using long term costs and effects beyond the time frame of the study. Markov models simulate the course of a disease over time and thus allow to calculate long term costs and effects.