Hester van Eeren

Framework for modelling the cost-effectiveness of systemic interventions | 2 17 | Introduction Child delinquency poses a high economic burden on society (Welsh et al., 2008). Therefore, crime prevention and treatment of youth delinquents is of great importance to governments, in particular for Justice Departments. Systemic interventions, for instance Multisystemic Therapy (MST), Functional Family Therapy (FFT) or Parent Management Training Oregon (PTMO), are relatively costly interventions in youth health care aiming to reduce delinquent behavior (Aos, Lieb, Mayfield, Miller, & Pennucci, 2004). Cost- effectiveness studies are still limited in the field of youth health care. However, these costly systemic family interventions compete with medical treatments and other interventions for health care budgets, increasing the need for knowledge regarding the operationalization of economic evaluations in this context. In the Netherlands, as part of an ongoing nationwide action plan of the Ministry of Justice, recently a selection was made of evidence-based treatments for delinquent youth (Ministry of Justice, 2008), among which MST, FFT and PMTO were implemented given their apparent effectiveness in reducing criminal activity in youths. The aim of these systemic interventions is not primarily to produce health in the sense of physical health and absence of disease, as measured in the Quality Adjusted Life Years (QALY) outcome. These interventions attempt to improve family functioning and may even intervene with the peers and school environment of the youth (i.e. Hendriks, van der Schee, & Blanken, 2011; Sexton & Alexander, 2000). Still, these treatments are reimbursed by the Dutch social health insurance system and, as such, part of the health care sector. Therefore, like other health care interventions, each intervention needs to demonstrate value for money since it competes for limited funds with other interventions. Efficiency considerations are deemed important in guiding decisions on which treatments to reimburse or initiate. However, given the atypical aim of these systemic interventions, i.e. reducing youth delinquency, an important question is how these types of interventions could demonstrate their efficiency or value for money. The conventional health economic approach of measuring improvements in terms of QALYs may fall short in this context. Indeed, considering the literature on reducing youth delinquency, it becomes clear that important differences exist between economic evaluations performed in the health care sector and evaluations of crime prevention and treatment programs. It seems that both fields commonly perform sophisticated effect studies, including randomized controlled trials, meta-analyses and systematic reviews (Glisson et al., 2010; Henderson, Rowe, Dakof, Hawes, &Liddle, 2009; Kinget al., 2006; Teuffel et al., 2011; Ttofi, Farrington, Losel, & Loeber, 2011). Considering economic evaluations of crime prevention and treatment programs the classical cost-benefit analysis is conventionally used (Loeber & Farrington, 2000; Soeteman & Busschbach, 2008). An extensive cost-benefit evaluation of crime prevention and intervention programs has been performed by Aos and colleagues (2004) in the United States. That evaluation was based on a literature review, computation of average effects per treatment program, assignment of a monetary value to the effects and subsequently calculation of a net present value in a cost-benefit model structure. Furthermore, French and colleagues (2002a; 2002b), for example, conducted cost-benefit analyses on addiction treatment for substance abusers. These cost benefit analyses were deterministic models (Aos et al., 2004; French, McCollister, Cacciola, et