Stability of the effects of a social competence training program for children with oppositional defiant disorder/conduct disorder: a 10-month follow-up

Oppositional defiant disorder (ODD) and conduct disorder (CD) are among the most common disorders for which children and adolescents are referred for mental health treatment [1]. Longitudinal studies have shown that ODD/CD are highly stable over time and predict later emotional and behavioral problems in adolescence and adulthood [2‐4]. With regard to psychotherapeutic interventions, ODD/CD are notably responsive to parent management training (PMT) [5‐8], child-centered treatments (cognitive behavioral interventions (CBT) including social skills training) [9, 10], and a combination of PMT and CBT programs [5, 11‐13].

While PMT is considered as the gold standard, this approach still has some limitations, and some factors remain untreated, in particular those related to the child (e.g., child’s inability to handle anger due to affect regulation deficiencies) [14]. Moreover, regarding therapeutic effects on behavioral problems outside the family (e.g., behavior at school; peer relationship problems), CBT interventions that focus on social skills and cognitive interventions (e.g., problem-solving and anger management strategies) seem to increase the effects of PMT [13, 15, 16].

However, many studies only consider the post-treatment effects and neglect to examine the long-term effects, which are of great importance given the negative long-term prognoses of ODD/CD. In this context, a meta-analysis by van Aar and colleagues [17] encompassing 40 studies examined the long-term effects [follow-up (FU) assessment up to 3 years] of PMT on disruptive child behavior and found that the post-treatment effects remained stable. However, the long-term effects of CBT interventions were not considered. Furthermore, a meta-analytic review by Fossum and colleagues [18] identified 56 studies (published between 1984 and 2010) that examined the longer term impact (with a mean FU period of 8.9 months; SD = 6.5) of PMT (k = 34), CBT (k = 7), a combination of PMT and CBT (k = 12) or family therapy (FT; k = 3) on conduct problems and oppositional behaviors (within the clinical range before treatment). Overall, the findings again indicated a stability of treatment effects. None of the included studies found a significant deterioration in ODD/CD symptoms in the FU period compared to post-treatment, and overall, there was even a small (non-significant) further reduction in conduct problems after treatment was over (effect size = 0.08). Moreover, the meta-analysis revealed that changes in conduct problems were significantly larger in the studies that examined PMT in combination with CBT or CBT alone compared to studies that examined PMT alone or FT. The study by Lochman and Wells [13] comes to comparable findings. Their analyses showed that the Coping Power Program (a child CBT program) have sustained effects 1 year after the program has ended, especially when parent components are added. Children of the intervention group showed lower rates of self-reported covert delinquent behavior and of parent-rated substance use at FU that at the baseline than did the control group.

However, most of the studies considered here predominantly examined the stability of treatment effects (i.e., within-group change: comparison of post-treatment versus FU), while only a small number of studies, e.g., [19, 20], also investigated whether there were differences between the treatment and control group at FU (i.e., between-group comparison). These studies mostly did not find any significant differences between the groups at FU, possibly due to the fact that control group participants had received adequate therapy in the interim.

Overall, it appears that the number of studies on long-term effects of cognitive interventions in ODD/CD is considerably low. A closer look also shows that there are hardly any studies that analyze the difference between the treatment and the control group at FU. Further studies are strongly needed in this area.

The aim of the present study was therefore to evaluate the long-term effects of the Treatment Program for Children with Aggressive Behavior (THAV) [21]. THAV mainly consists of a patient-focused intervention program (CBT), which is combined with parent interventions (PMT) and teacher- and peer-focused interventions according to the individual needs of the patient.

The effectiveness of the program has been demonstrated in two previous analyses [11, 22]. In a study employing a within-subject design which included 6–12-year-old boys with peer-related aggressive behavior, growth rates for all parent-rated outcome measures were found to be significant during treatment, and a comparison with the waiting phase indicated a stronger decrease in aggressive behavior and a stronger increase in prosocial behavior during treatment [22]. In a randomized controlled trial (RCT) evaluation, THAV (n = 50) was compared with an active control involving group play (PLAY; n = 41). The between-group evaluation showed mostly moderate treatment effects for THAV compared to PLAY in parent ratings on aggressive behavior, comorbid symptoms, psychosocial impairment, quality of life, parental stress, and negative expressed emotions at post-assessment [11].

In the current analysis, we subsequently examined whether these effects (based on parent ratings) remain at a 10-month FU assessment. Besides testing the stability of treatment effects, the present study also seeks to shed light on the differences between THAV and PLAY at 10-month FU. In sum, our hypotheses were as follows: (a) improvements in child behavior (based on multiple parent-rated outcome measures) through THAV treatment would be sustained 10 months later (comparison of post-treatment versus 10-month FU controlling for pre-treatment); (b) the significant effects of THAV compared to PLAY shown at post-assessment would also be apparent at 10-month FU.

Sample characteristics including ICD-10 diagnosis at baseline, age at FU assessment, treatment received during the FU period, and average waiting time are reported in Table 2. Chi-square tests indicated no significant differences between THAV and PLAY regarding treatment received during the FU period [psychological outpatient treatment (Χ²(1) = 3.25, p = 0.07); pharmacotherapy (Χ²(1) = 0.03, p = 0.95); training therapy (e.g., on reading and spelling training) (Χ²(1) = 0.74, p = 0.39)].

Results of the MMRM including means and SDs (post and FU) as well as effect sizes are detailed in Table 3. On a descriptive level, both THAV and PLAY yielded a reduction in problem behavior from post-assessment to FU assessment, with effect sizes in the low-to-moderate range (d_post/FU = 0.06–0.43). These differences were not significant, with the exception of the scales SCL-DBD ODD for both THAV and PLAY, the SCL-ADHD total for THAV and the CBCL-Internalizing problems for PLAY. Here, the values at FU assessment were significantly lower than those at post-assessment.

A comparison of THAV and PLAY at the FU assessment revealed, on a descriptive level, lower problem behavior and higher prosocial behavior in the THAV group compared to PLAY, with effect sizes in the low-to-moderate range (d_THAV/PLAY = 0.22–0.71). These contrasts were significant for the scales SCL-DBD ODD and SCL-DBD Prosocial behavior.

Figure 2 illustrates the course of mean values over pre-2 (after the waiting period), post (after the treatment period), and FU (after the FU period) for THAV and PLAY for the primary outcomes (FAVK-F total Peer and SCL-DBD ODD).

THAV is an individualized CBT treatment program for boys aged 6–12 years with a diagnosis of ODD/CD, which is combined with parent- (PMT), teacher- and peer-focused interventions. In the previous studies, THAV has been shown to be effective in reducing aggressive behavior problems [11, 22]. In particular, a study employing a within-subject design, which compared the course of symptoms during a 6-week waiting period with the course during the subsequent THAV treatment period, demonstrated a significant decrease in peer-related and adult-related aggressive behavior and a significant increase in prosocial behavior. Furthermore, the RCT evaluation indicated that when compared to an active control involving group play (PLAY), THAV is specifically effective in reducing peer-related aggression, with a low-to-moderate-effect size d_THAV/PLAY = -0.46) and in improving aggressive and rule-breaking behavior and prosocial behavior, with moderate-effect sizes.

The aim of the present study was to examine the stability of the effectiveness of THAV at a 10-month FU. We hypothesized that the improvements in child behavior, as rated by the parents, through THAV treatment would be sustained 10 months later from post-treatment to FU, and that the significant effects of THAV compared to PLAY shown at post-assessment would also be apparent at the 10-month FU. Overall, the results of our study confirm our hypotheses. Comparable to previous studies on long-term effects of CBT and PMT interventions on ODD/CD [17, 18], the comparison of post-treatment versus 10-month FU indicated the long-term maintenance of treatment outcomes. In particular, in THAV, there was no significant deterioration in symptoms of CD and prosocial behavior, or in problem-maintaining and problem-moderating factors. Moreover, there was even a further significant improvement in oppositional behavior and comorbid ADHD symptoms. Similar to Fossum and colleagues [18], we mostly found small-to-moderate-effect sizes. When clinical significance analysis is considered, it appears that 90% of patients in the THAV condition improved at FU assessment compared to pre-treatment. Similar findings are also seen in the study by Lochman and Wells [13]. Here, it was shown that the Coping Power intervention moved aggression boys’ self-, parent-, and teacher-rated behavior problems from a nonnormative range at pre-treatment to a normative range by the 1-year FU.

The described stability and even partial improvement are evident not only in the THAV but also on almost all scales in the PLAY condition. On the scale of CBCL-Internalizing behavior, the PLAY condition even shows a stronger improvement than the THAV condition over the FU period. This can possibly be explained by the fact that the group condition offers good opportunities for improving social skills and an easier transfer to the natural environment. The only exception is the SCL-DBD-Prosocial Behavior, which shows a slight, non-significant deterioration in the PLAY condition over the FU time.

When comparing THAV and PLAY at FU, it emerged that the significant differences between the two conditions found at post-treatment assessment were mostly also apparent 10 months later (SCL-DBD-ODD, SCL-DBD-Prosocial Behavior). In contrast to post-treatment [11], no significant difference between the two conditions was found on the FAVK-F-Peer and CBCL-Externalizing behavior at FU. However, these values narrowly missed significance (p = 0.07). With regard to effect sizes, comparable values between post-treatment and FU were evident, although some values were slightly higher at the FU (SCL-DBD-CD: d_post = 0.27 vs. d_FU = 0.41; SCL-DBD-Prosocial behavior: d_post = 0.42 vs. d_FU = 0.71; CBCL-Internalizing behavior (d_post = 0.28 vs. d_FU = 0.49).

To summarize, as the RCT study on post-treatment effects already demonstrated [11], the present findings now indicate that the multimodal approach of THAV is superior to the active group play of PLAY—not only immediately after therapy but also 10 months later. At this point, it is difficult to compare our findings with previous research, because we found only two studies [19, 20] that, besides investigating the stability of effects of CBT (and PMT) on ODD/CD, also looked at the differences between the groups at the FU assessment. The first study, by Cavell and Hughes [19], examined the post-treatment and long-term effects of a CBT program (training in problem-solving skills, and consultation with parents and teachers) compared to a ‘Standard Mentoring’ program (that was carried out with minimally trained, unsupervised mentors) in highly aggressive children. The authors did not find any group differences at the FU assessment. However, the study also did not reveal any significant group differences at the post-treatment assessment. The second study, by Pepler and colleagues [20], investigated the efficacy of a combined CBT and PMT program compared to an untreated comparison group (waiting list) in aggressive girls. The study revealed a difference between the groups at post-treatment, but did not find any significant differences between the groups at FU, with the latter being due to the fact that the control group participants had received adequate therapy in the interim. These two studies are therefore also difficult to compare with our findings, as in our study, there was no significant difference between THAV and PLAY with respect to treatment received during the FU period.

The present study has several strengths and limitations. One strength is that we looked not only at the stability of the THAV treatment but also at the stability of the differences between THAV and PLAY at FU assessment. To the best of our knowledge, our study is the first to demonstrate on one hand that the therapeutic effects of a multimodal therapeutic approach for ODD/CD with a focus on child-based interventions remain stable over a certain period of time, and on the other hand that the effects are superior to those of an active control group in the long term. A further strength of the study is that we examined an active control group instead of a waitlist control group. It can be assumed that the differences found between THAV and the control group would have been even greater if children of the control group had received no therapy at all. Besides other strengths of the study, such as the high percentage of participants (84%) retained in the FU, several limitations have to be mentioned. First, ten patients dropped out from the PLAY condition after randomization because of ethical objections by the therapists. For these patients, the more effective treatment according to the hypotheses of the trial was strictly indicated according to the regulations of the ethics committee, which required that the patients must not be put at a disadvantage due to their participation in the trial. Even though it represents a methodological shortcoming of our study, we decided to exclude the patients from further analyses, since the alternative—an imputation of missing values in this control group (e.g., with last observation carried forward)—would have favored the treatment group and thus the research hypotheses. Second, the results are restricted to boys, and further research should therefore also focus on girls. Third, the influence of callous-unemotional traits was not assessed in the FU. The impact of limited prosocial capacities and of psychopathic or callous-unemotional personalities on the effect of THAV and PLAY should be considered in further studies. Fourth, all outcome measures were based on parent ratings. As such, the results are influenced by the parents' perceptions and expectancies and may be biased due to effort justification. However, this bias may be similar in both conditions. Moreover, parents cannot make an accurate statement about their children’s behavior at school. Here, teacher ratings would have been helpful. However, although we also collect the teachers’ questionnaires at FU assessment, we did not include these in the analyses due to a high number of missing data. Likewise, we also collected self-ratings from the age of 10 years, but here too, there were too few available data. Future studies should consider not only parent ratings but also self- and teacher ratings to draw a more complete picture. Fifth, parents could not be fully blinded, because they actively took part in the intervention. Nevertheless, parents were blinded to the specific hypotheses of the study and many of the parents in the PLAY condition actually believed that this was the intervention that was expected to be more effective. Finally, adherence to the interventions should be considered in further studies. Information on adherence on THAV can be found in the RCT article [11], but an analysis of the differences in adherence between THAV and PLAY also seems interesting, as PLAY may be an effective treatment method and not simply an unspecific control group.

Overall, the study found that the effects of the individually tailored treatment program THAV, which combines patient-focused (CBT) with parent- (PMT), teacher-, and peer-focused interventions, were retained at a 10-month FU assessment. The effects that were found at the end of the treatment remained stable and even partially improved over the FU period. Additionally, the differences between THAV and the active control group PLAY, which were apparent at the end of the treatment (post), mainly also persisted 10 months later.