Hester van Eeren

Framework for modelling the cost-effectiveness of systemic interventions | 2 29 | further research to characterize the uncertainty of the model inputs, including long-term effects (Briggs et al., 2006). Fourth, the cost parameters in the model are depicted from a combination of costs used in health economic evaluations and literature on cost of crime. However, victim costs and intangible costs, which include direct economic losses of the victims and indirect losses suffered by these victims, respectively, are not taken into account (McCollister, French, & Fang, 2010). Addition of these costs could be of value. Finally, model parameters were solely based on the limited evidence base of available literature and where retrieved out of different literature sources. Ideally, these parameters would be retrieved from more comprehensive empirical data. For example, the transition probabilities could be linked to the presence or absence of police contacts, contacts with judicial institutions or committed crimes. Availability of additional data can refine the input data of the model and increase the validity of the model structure and the accuracy of the results. Concluding, we used the methods commonly employed in health economic evaluations to create a framework for determining the value for money of interventions targeted at reducing youth delinquency. The results are encouraging, but important further steps still need to be taken. A first next step may be the collection of empirical data to test the presented methodology. We further suggest the construction of a multidimensional outcome measure that enables researchers to capture the multiple dimensions of the treatment goals, in a preference-weighted manner. A final matter that deserves attention is the value we assign to outcomes such as reduced delinquency. Calculating cost-effectiveness is especially useful when the results can be judged against some ‘threshold’ value. What this should be in this context remains unclear as yet.