This article uses programme theory reconstruction to investigate three conceptualisations of evidence-based practice (EBP). The first two are taken directly from the literature and contrasted with a third taken from a relevant policy programme. The article argues that 'evidence-based practice' is not a homogeneous concept, and that interested parties within research, practice and policy may not have a shared vision of EBP even though they may use the same terminology. It is important for the successful implementation of EBP within practice and policy that there is a clear understanding of what EBP is and how EBP is supposed to work if research, policy and practice are to work together toward common goals.

his randomized clinical trial assessed the effectiveness of multisystemic therapy (MST) for 156 youths who met the diagnostic criteria for conduct disorder. Sweden's 3 largest cities and 1 small town served as the recruiting area for the study. A mixed factorial design was used, with random allocation between MST and treatment as usual groups. Assessments were conducted at intake and 7 months after referral. With an intention-to-treat approach, results from multiagent and multimethod assessment batteries showed a general decrease in psychiatric problems and antisocial behaviors among participants across treatments. There were no significant differences in treatment effects between the 2 groups. The lack of treatment effect did not appear to be caused by site differences or variations in program maturity. MST treatment fidelity was lower than that of other studies, although not clearly related to treatment outcomes in this study. The results are discussed in terms of differences between Sweden and the United States. One difference is the way in which young offenders are processed (a child welfare approach vs. a juvenile justice system approach). Sociodemographic differences (e.g., rates of poverty, crime, and substance abuse) between the 2 countries may also have moderating effects on the rates of rehabilitation among young offenders. 

Within the context of a randomized trial, this study assessed the costs of treating conduct disorder with multisystemic therapy (MST) or treatment as usual (TAU) for 156 youth in Sweden. From the perspective of the municipal Social Welfare Administration, all intervention costs were collected for the 6-month period starting at randomization. MST was found to cost, on average, US$8,847 per youth. A course of intervention including MST was found to cost significantly more (US$5,038 per youth) than TAU. Although MST was found to reduce the use of and costs associated with non-placement interventions, MST was not found to reduce the use of or costs associated with placement interventions. This is contrary to findings reported from studies undertaken in the USA. Intent-to-treat (ITT) and treatment-of-treated (TOT) analyses are presented.