Hester van Eeren

Framework for modelling the cost-effectiveness of systemic interventions | 2 19 | As an illustration an initial assessment of the cost-effectiveness of Functional Family Therapy (FFT) compared to treatment as usual (TAU) is presented. As the aim of the study is the application of the probabilistic decision analytic modelling to interventions aimed at reducing delinquency, the interventions compared could be substituted by other systemic interventions mentioned. The article is structured as follows. The methods section provides information on the health economic model type and general characteristics of the model. The results section elaborates on the applicability of the decision analytic model and outcome measure to the field of systemic interventions specifying necessary adaptations to the health economic approach based on an initial assessment of cost-effectiveness of FFT. The conclusion relates our findings to the general objective of applying health economic methods to systemic interventions not primarily aimed at improving health. Methods Model structure We constructedaprobabilisticMarkov cohortmodel (Briggs et al., 2006). Diseaseprogression in common Markov models is described using transitions between ‘states’, where a subject can move between states or remain in the current state. The transition rates between states are typically estimated based on short run data. Long-term predictions are made based on repetition of transition cycles and assumptions based on for example literature. In order to keep the initial model as transparent as possible, a Markov model was constructed consisting of three states, i.e. A - criminal behavior, B - non criminal behavior and C - dead. The model structure is shown in Figure 1. All subjects in our study started in state A, moved to either state B or C or remained in state A and could then move between criminal and non-criminal states. Death acted as the absorbing state. Note that subjects could also remain in their present state (depicted by the u-turns). A - criminal B - not criminal C - dead tpA2B(1-nmr) tpB2A(1-nmr) nmr nmr tpB2B(1-nmr) tpA2A(1-nmr) Figure 1. Markov model nmr = natural mortality rate tpA2A = transition probability of staying in state A tpA2B = transition probability of moving from state A to state B tpB2A = transition probability of moving from state B to state A tpB2B = transition probability of staying in state B