Page 118 - Age of onset of disruptive behavior of residentially treated adolescents -Sjoukje de Boer

In chapter 5, the aim was to examine differences between dropouts (n=77) and completers (n=147) on known risk factors. In addition, EO disruptive behavior, and specific types of disruptive behavior and their relation to dropout were considered. The dropout group was further sub classified into withdrawal (n=40) and pushout (expulsion, n=37). It was found that dropouts compared to completers more commonly were male, showed more frequent cannabis use prior to treatment, more often displayed disruptive behavior at school and belonged more often to the EO group. Also, individuals with EO disruptive behavior dropped out almost three times as often as AO counterparts. Further, adolescents who had used cannabis prior to treatment dropped out twice as often as those without previous cannabis use. Within the dropout group, no differences in risk factors or types of disruptive behavior were found between withdrawals and pushouts.
The aim of the study in chapter 6 was to examine the 18 month outcome of the dropouts (n=67) and completers (n=129) after leaving the orthopsychiatric facility. General daily functioning was the outcome assessed. It was found that compared to dropout, treatment completion was associated with more stable living conditions, having a daytime activity, and absence of delinquent behavior. To control for the possible effect of the covariates (of chapter 5) associated with dropout in our sample, EO disruptive behavior, cannabis usage prior to admission and male sex were also examined. The results indicated that treatment completion, compared to dropout, was the best predictor for good general daily functioning at follow-up explaining 10.1% of the variance. EO disruptive behavior, cannabis usage prior to admission and male sex, added to the explained variance over and above treatment completion.
General discussion
Overall, this thesis shows that differentiating between individuals with EO and AO disruptive behavior, within a specific orthopsychiatric sample carries some clinical relevance. Although our group is highly selective, which hampers generalizability, the results are informative at group level. The distinction between EO and AO, together with the other predictors (i.e., cannabis use prior to treatment, male sex), was mainly relevant in the relation to dropout. As the explained variances were low, it must be recognized that other unknown factors carry relevance as well. Nevertheless, at the
116