Comparing the Effectiveness of Functional Family Therapy and Multisystemic Therapy Using the Propensity Score Method

Objectives: Governments and youth care organisations want to make informed decisions about their investments in youth interventions. Therefore, evidence for the effectiveness of interventions is becoming more and more... [ view full abstract ]

Objectives: Governments and youth care organisations want to make informed decisions about their investments in youth interventions. Therefore, evidence for the effectiveness of interventions is becoming more and more important. Nowadays various databases provide information about the evidence-base of interventions. For instance, Functional Family Therapy (FFT) and Multisystemic Therapy (MST) are both known as Blueprints models for healthy youth development and have been proven to be effective in reducing antisocial and delinquent behaviour. However, little is known about their relative effectiveness, that is, whether one intervention would outperform the other in a direct comparison. Comparing evidence-based interventions with one another is a next step in evaluating effectiveness, because it may provide insight into what intervention works best for whom instead of only showing that it works. Therefore, the objective of the present study was to compare FFT and MST on their effectiveness and their target populations.

Method: Because we followed regular clinical practice and did not randomly allocate adolescents to one of the two interventions, we used the propensity score method to control for observed pre-treatment differences between the two target populations and balance the treatment arms. In total, 697 adolescents were assigned to either FFT or MST between January 2009 and June 2014 and completed the primary outcome measure (i.e., externalizing problem behaviour reported by the parents and measured with the Child Behaviour Checklist: N = 275 FFT; N = 422 MST). Secondary outcomes were the proportion of adolescents that were living at home, were engaged in school or work, and did not have police contacts during treatment. Data was gathered within Routine Outcome Monitoring of De Viersprong, institute for personality disorders and behavioural problems in the Netherlands. In addition to the full sample, treatment effects were estimated in two subsamples of youth with and without a court order.

Results: Looking at the full sample, the treatment effect regarding externalizing problem behaviour was not significantly different between the two interventions. MST was, however, significantly more effective in increasing the probability of being engaged in school or work after treatment. Within the subsample of youth without a court order, MST was significantly more effective in reducing externalizing problems than FFT. Because balance was not achieved in the sample of youth with a court order, the relative effectiveness of MST and FFT could not be compared in this subgroup: FFT and MST cases in this subsample were probably not fully comparable.

Conclusions: Some significant differences in the effectiveness of FFT and MST were found in the Netherlands. This study also revealed that the target populations of FFT and MST were largely comparable, but that a particular subgroup could not be matched. This result, once more, underlines the importance of controlling for baseline differences when comparing interventions on their effectiveness in regular clinical practice. Using the propensity score method seems to be a feasible and useful approach for this type of research.