Theory- and evidence based interventions for children with conduct problems : exploring applicability and underlying assumptions in real world settings

Abstract

Severe child conduct problems (aggressive, disruptive, destructive, oppositional, non-compliant and antisocial behavior) are predictive of serious and violent offences, substance use problems, mental health problems and domestic violence later in life (Moffitt, 2006). Left untreated, as many as 50% to 75% of the children with severe and early conduct problems will exhibit antisocial behavior in adolescence (Nixon, 2002). However, early trajectories of conduct problems can be altered (Brestan & Eyberg, 1998; Lundahl, Risser, & Lovejoy, 2006; Nixon, 2002) and the development of interventions targeting child conduct problems has been an important area of research in the last decades (e.g. Dishion & Patterson, 2006; Webster-Stratton, Reid, & Hammond, 2004). The social interaction learning (SIL) model postulates that contextual factors (e.g. divorce, stress, SES) initiate coercive and aggressive interactions between family members and disturb parenting skills, thereby leading to an early onset path of antisocial behavior (Patterson, 1982). In other words, contextual factors are assumed to impact child adjustment indirectly through parenting practices. <br><br>

The present thesis has examined the applicability and underlying assumptions of theory- and evidence based prevention and treatment interventions in real world setting, namely Parent management training, the Oregon model and the community-wide model called “the Early Intervention for Children at Risk for Developing Behavioral Problems” (EICR). The principles of these interventions are derived from the SIL model. In order to study processes leading to conduct problems, Paper I examined whether the effect of interparental collaboration was mediated through maternal and paternal parenting. Two studies were carried out to investigate the applicability of interventions in real world settings. The study presented in Paper II explored whether the implementation of the EICR model in a community was related to reduced problems behaviors and better child relations in the school context. The study presented in Paper III examined gender differences in behavioral change following PMTO. <br><br>

The cross-sectional findings in Paper I were derived from a sample of families (N = 136) from all health regions in Norway who had been in contact with child and adolescent psychiatric services to get help with their child’s conduct problems prior to receiving PMTO. Paper II, which was based on a quasi-experimental pre-post design, consisted of a sample of school employees (N = 266) recruited from a pool of 271 employees from seven elementary schools (1st–7th grade) in a municipality near Oslo. The sample in Paper III consisted of 323 families who received PMTO, of which 239 (74%) participated in assessment sessions before (pre) and after (post) treatment. <br><br>

Findings from Paper I showed that maternal and paternal parenting practices fully mediated the relation between interparental collaboration and externalizing behavior. However, when the indirect effects of parenting practices were tested separately, paternal parenting functioned as a mediator, whereas maternal parenting did not, indicating that the relation between interparental collaboration, parenting and externalizing behavior was moderated by parent gender. The finding that interparental collaboration influenced child externalizing behavior indirectly through parenting practices suggests the importance of addressing interparental communication skills, teamwork and consensus during parent training interventions when parents are in a problematic relationship: Helping parents communicate and cooperate more efficiently seems to make it easier for parents to interact more positively with their children. <br><br>

The results from Paper II showed that the EICR model had positive short term effects (one year after initiation) in the sense that the prevalence of student problem behavior was significantly lower, and student relations were significantly better in schools located in the intervention area than in schools located in the comparison area. The positive immediate effects indicate that the implementation of the EICR model, which in fact only directly affects a small portion of the child population, can have positive impacts on the overall prevalence of child problem behavior in municipalities. <br><br>

Pre treatment findings from Paper III showed that girls exhibited less externalizing problem behavior according to parents and teachers, more internalizing problems according to parents, and more social competence according to teachers. Parents reported more girls to be within clinical range on externalizing behavior, while teachers reported more boys to be within clinical range. Teachers reported more comorbidity among boys than girls. At post treatment, girls had more likely changed in a positive direction according to teacher-ratings of externalizing behavior and social competence, but not according to any of the parent-reported variables. <br><br>

The findings from the present thesis provide support for the SIL model, thereby suggesting that it may be a beneficial strategy to offer interventions derived from the model to children with conduct problems in Norway. Despite promising findings, further research is needed to develop and refine interventions for children with severe conduct problems.