Cognitive-Behavioral Intervention for Youth At-risk for Conduct Problems: Future Directions

Abstract

This paper briefly overviews the history of cognitive-behavioral intervention (CBI) for children displaying early signs, or actual diagnoses, of conduct disorders. A series of randomized-control trials have identified evidence-based CBI programs for children with these behavior problems at various developmental stages from preschool through adolescence. While it is critically important for the field to disseminate these existing programs as developed, we argue that it is important to also move beyond the existing evidence-based programs. Research should continue to test new comprehensive, multi-component interventions, fueled by our evolving understanding of active mechanisms that contribute to children’s externalizing behavior problems, the future of research in this area can also benefit from four central areas of intervention research. First, research can address how single interventions can have meaningful impact on a range of transdiagnostic outcomes because the intervention mechanisms may affect those various outcomes. Second, rooted in implementation science, we are beginning to understand better how evidence-based programs can be disseminated in the real-world, examining key issues such as the adequacy of training approaches and the role of therapist and organizational characteristics. Third, a major focus of research can be on how to optimize intervention outcomes, including a focus on microtrials, on tailoring of interventions, on examining rigorously how interventions are delivered, and on the integration of technology and of other approaches such as mindfulness training into cognitive-behavioral intervention. Fourth, research can explore how the therapeutic relationship and the therapists’ characteristics can play substantial roles in effective CBI with conduct problem children.

This paper briefly describes the history of cognitive behavioral intervention with conduct problem children. The paper then discusses four central themes for ongoing and future research in this area, largely drawing on our own current research on cognitive-behavioral intervention. Key directions for future work can focus (a) on the long-term and transdiagnostic effects of interventions, (b) on the dissemination and implementation process for interventions, (c) on the need for optimizing new and existing interventions, and (d) on the critical role of the therapist and therapeutic engagement in cognitive behavioral intervention with conduct problem children.

Roots in Observational Learning, Social Learning and Developmental Theories

Cognitive behavioral interventions focus on children’s cognitions and their behavior, and on their emotional arousal and regulation. Current cognitive behavioral therapies with children have important roots in social learning theories and developmental theories that were being proposed in the 1950s, 1960s, and 1970s (Lochman & Pardini, 2008; Matthys & Lochman, 2017). Research about how individuals’ cognitions could guide their behavior came from the social learning theories (SLT) of Julian Rotter (Rotter, 1954) and his student, Walter Mischel (Mischel, 1973). Rotter’s was the first personality theory to integrate cognitive representations of social information and contingencies (individuals’ expectations for attaining consequences for their behavior) with learning and motivation, in the form of reinforcement. A series of elegant laboratory studies supported this theory, and found that expectancies could be generalized across situations (e.g., internal or external locus of control) or could be specific to the situation. Rotter believed personality was an interaction between a person and that person’s environment. He was an early proponent of the need in treatment to manipulate family and school environments to change child behavior.

Separately, Bandura’s theory (Bandura, Ross, & Ross, 1963) stressed the important role of cognitions by highlighting that a child’s observations of real-life models and symbolic models could influence the child’s subjective perceptions of how rewards and punishments were contingent on their behaviors. To acquire new behaviors through modeling, a child must pay attention to a model (e.g., a peer entering a soccer game), cognitively retain and remember relevant aspects of the model’s behavior, enact the behavior that had been observed, and then be reinforced for the reproduced behavior. Based on early research, it was apparent that children tended to imitate aggressive behavior when they observed a model being reinforced for aggression (Bandura et al., 1963). Mischel (1973) integrated the work of Bandura and Rotter with experimental research on cognitive processing and symbolic mental representations (Neisser, 1967) in an attempt to further explain the complex and dynamic interactions between children and their environment. Mischel’s theory predicted that individuals’ behavioral reactions to environmental stimuli was affected by their encoding strategies, outcome expectancies and values, behavioral regulation systems, and planning abilities. Based on Bandura’s notion of reciprocal determinism, Mischel (1973) also theorized that children’s overt behavior could produce changes in the environment and modify their social stimuli (e.g., a teacher’s behavior to them). Children were viewed as active participants in their social environment, with the ability to influence social exchanges, as well as to be influenced by them.

Developmental psychologists focused on related cognitions and behaviors during this time. Piaget described children as developing an increasingly sophisticated set of mental representations and logical structures in order to master their own behavior and the environment (Thomas, 1996). In the early 1960s Vygotsky (1962) hypothesized from a different perspective that children’s words can direct their mental operations, and that these mental operations in turn control children’s problem-solving behavior. Similarly, Luria (1961) proposed that children go through three developmental stages as they voluntarily initiate or inhibit their own behavior: (a) relying on other- (usually a parent) directed control, (b) controlling their behavior through their overt speech, and (c) controlling their behavior through inner, or covert, speech. Thus, these developmental theorists emphasized the connection between cognitions and behavior, and specifically how cognitive operations are involved in the development of self-control. Their theories had important effects on early cognitive behavioral therapists, such as Donald Meichenbaum, who began working on enhancing children’s self-control.

Early forms of cognitive behavioral therapies and theories with children.

Expanding traditional conditioning paradigms with a multidimensional approach that included covert processes (i.e., cognition), Meichenbaum (1977) and Mahoney (1974) described how internalization of self-statements guided problem solving or behavior. In 1971, Meichenbaum and Goodman found that a combination of modeling and self-instructional training was successful in decreasing children’s impulsive behavior. The program taught children to control their own behavior by modeling self-control verbalizations and instructing children on how to engage in private self-speech while performing sustained attention tasks. Kendall and Braswell (1982) developed a more comprehensive cognitive therapy that taught impulsive children the general steps to problem solving and how to use internalized coping statements to deal with frustration and failures when engaged in goal directed behaviors. Spivack and Shure’s (1974) early research demonstrated that social problem-solving training could be used effectively with children and their parents, and social problem-solving training became a key element of behavioral interventions for children (Allen, Chinsky, Larcen, Lochman, & Selinger, 1976).

Development of Anger Coping and Coping Power Programs addressing social cognition and emotion regulation.

Based on clinical work with low-income aggressive children that began in the late 1970s, Lochman, Nelson and Sims (1981) developed an anger arousal model, and an accompanying Anger Control Program, that incorporated both the self-instruction training methods from Meichenbaum and the social problem-solving training methods from Shure and Spivack. This anger arousal model followed from Novaco’s (1978) conceptualization of adult anger and their efforts to cope with stress, which in turn had been influenced by Rotter’s SLT and Lazarus’ (1966) work on cognitive appraisal processes. Novaco’s (1978) theory indicated that aversive, provocative events in the individuals’ environment had no direct effect on their anger and subsequent behavior except as mediated by the individuals’ appraisal of the events, their expectations and their private speech.

Aggressive children were seen as having two primary areas of cognitive difficulty in this anger arousal model (Lochman et al., 1981). Children first had to accurately perceive and interpret the problematic social situations they encountered (aggressive children had cognitive distortions at this step and were expected to have misperceptions and have accompanying anger arousal) and to develop methods for coping with their high level of arousal (self-statements; relaxation; attention-focusing), and then children had to go through a problem-solving sequence that involved thinking of possible solutions to the problem, considering the anticipated consequences of the solutions, and picking an optimal solution (aggressive children had deficiencies in the number and types of solutions that they could generate, and focal problems in not anticipating the array of consequences they would likely experience following their behavioral choices). A prominent feature of this anger arousal model was that cognitions and physiological arousal related in reciprocal ways, requiring attention to emotional, cognitive and behavioral dysregulation. Social information processing models evolved and expanded on the sequential series of cognitive processes that children display during their social problems (Dodge, 1993), and these social cognitive processing difficulties were found to be evident in both severely and moderately aggressive youth (Lochman & Dodge, 1994).

The Coping Power program (Lochman & Wells, 2002a), which followed from the Anger Coping program, focused on these more discrete social-cognitive processes, and included focus on children’s schematic expectations and social goals (Lochman & Dodge, 1998; McDonald & Lochman, 2012), and on the automatic nature of the primary appraisal phase of these processes (Yaros, Lochman, Rosenbaum, & Jimenez-Camargo, 2014). Based on an expanded contextual social-cognitive model, a series of randomized control trials have found this cognitive behavioral intervention can reduce children’s externalizing behaviors, substance use and delinquency through one- and three-year follow-ups (Lochman & Wells, 2004; Lochman et al., 2013, 2014). In the course of this programmatic series of efficacy studies we have found it has been important, however, to move beyond efficacy research (LaGreca, Silverman, & Lochman, 2009), and the following sections indicate our views of four key next steps in intervention research.

The Long-term Course and Transdiagnostic Nature of Intervention Outcomes

Current evidence-based interventions typically have been developed to focus on immediate intervention effects, and to influence one key outcome area, such as aggression. It will be crucial to consider both the future course of outcomes, as well as the breadth of outcomes that might be influenced by a specific focused intervention. In terms of future course, attention should be paid to the long-term impact of interventions. Although many treatment programs for youth at-risk for conduct problems have demonstrated immediate and short-term outcomes (e.g., 6 months-2 years), more limited research has been conducted on successful longer-term outcomes (Hood & Eyberg, 2003; Schaeffer & Borduin, 2005).

Interestingly, the nature of intervention outcomes may change over time, so it is prudent to explore patterns of long-term outcomes and keep a long-view in mind when evaluating effectiveness. For example, the Conduct Problems Prevention Research Group (CPPRG, 1992) found that the Fast Track Program, a longitudinal, multi-site universal prevention program that targets child, family, school, and community levels has demonstrated varying outcomes over time. The Fast Track intervention initially found significant effects on peer ratings of aggression, disruptive behavior, and ratings of classroom atmosphere. Additionally, moderate positive social effects were reported on children’s social, emotional, and academic skills (CPPRG, 1999). Through the end of elementary school, many of these effects were maintained (CPPRG, 2004). However, by middle school, most of these outcomes had faded. During 7^th grade, other than children’s hyperactive and self-reported delinquent behavior, Fast Track demonstrated no intervention effects on externalizing behavior or social skills, and a negative effect on involvement with deviant peers (CPPRG, 2010b). It was hypothesized that the impact of social environmental influences during this adolescent period temporarily masked intervention effects. By the age 19 follow-up, positive outcome effects, on juvenile arrests, had begun to re-emerge (CPPRG, 2010a), and by the age 25 follow-up, Fast Track participants had lower rates of psychiatric disorders, violent crimes, drug convictions, and risky sexual behaviors, and higher well-being scores compared to control participants (CPPRG, 2015). Taken together, these results suggest that the pattern of intervention effects of an early preventative intervention for youth at-risk for conduct problems can change over the course of time, affected by mechanisms activated early in intervention, supporting the need for a “long-view” when considering outcomes.

In addition to the future course of outcomes, the breadth of intervention outcomes is imperative to consider, as evidenced by the diversity of long-term outcomes described above. In particular, transdiagnostic approaches may facilitate promotion and consideration of broader, wide-ranging intervention effects. Historically, treatment research has taken a disorder- and content-focused approach, which may be too narrowly focused on specific outcomes and tends to exclude those with co-morbidities (Harvey, 2014). By contrast, transdiagnostic approaches are more process-focused, and may group common disorders to identify universal cognitive, emotional, physiological, or interpersonal processes or mechanisms to target. Recent attention has been garnered surrounding such approaches for several reasons, including: high rates of comorbidity in the population, the ability to efficiently address multiple problems concurrently, the potential for more generalizable outcomes, and rapid transfer of treatment breakthroughs to multiple disorders (Ehrenreich-May & Chu, 2014; Harvey, 2014). Furthermore, transdiagnostic treatments may be more ideal for dissemination because they can reduce the need for clinicians to learn multiple treatments for different disorders, and can increase adherence to protocol if flexible enough in application (Ehrenreich-May & Chu, 2014). Given these potential numerous benefits, transdiagnostic research is emerging as a key future direction in the field.

Within a child- and adolescent-specific context, Harvey (2014) argues that a transdiagnostic approach is “uniquely suited to account for developmental differences in typical and atypical development, the presence of multiple reports, and the high rate of comorbidity,” and substantial co-morbidity is expected among children with conduct problems (Reinecke & Clarke, 2004). Furthermore, Nolen-Hoeksema and Watkins (2011) note that transdiagnostic research dovetails with traditional developmental concepts, including continuous and discontinuous development, multi- and equifinality, and convergent and divergent trajectories.

Early support for transdiagnostic approaches to both externalizing and internalizing disorders has been found in nonspecific intervention effects following PCIT (Chase & Eyberg, 2008) and Parent Management Training (Kazdin, 2010). Another example is the Coping Power Program, which was originally designed to reduce anger and aggression, but has found effects on both externalizing and internalizing problems, as well as other behaviors such as substance abuse (Lochman, Powell, Boxmeyer, Ford, & Minney, 2014). These broad outcomes are especially salient when considering developmental cascades of symptomatology, as aggression can lead to increasing youth depression, and co-occurring conduct problems and depression are a risk factor for early substance use (Miller-Johnson, Lochman, Coie, Terry, & Hyman, 1998). By targeting underlying specific risk factors and active mechanisms that may be more malleable, interventions can have broader long-term outcomes and may even be more effective.

Consideration of the Coping Power program provides a unique perspective on how an intervention for youth at-risk for conduct problems has evolved into a transdiagnostic approach (for a more complete description, see Lochman et al., 2014). Coping Power does not focus on specific diagnoses and has no exclusionary criteria for comorbid conditions, which can be treated concurrently. It addresses a number of common risk factors that predict both externalizing and internalizing problems, including difficult temperament (irritability, restlessness, irregular patterns of behavior, lack of persistence, low adaptability), psychosocial stressors in varied environmental contexts (school, neighborhood, peer group), social-cognitive deficits and distortions, and family factors such as harsh discipline practices, low levels of parental warmth and acceptance, multiple changes in family composition, and parental psychopathology, and lower levels of both externalizing and internalizing problems have been found after Coping Power intervention (Lochman, Dishion et al., 2015). Furthermore, Coping Power contains a number of “transdiagnostic features,” including contingency management, personal goal-setting, emotional awareness and regulation, perspective-taking, social problem-solving, and rehearsal and master opportunities, all of which can be flexibly applied to individual treatment at the therapist’s discretion. Research such as this can inform ways of conceptualizing existing interventions as well as guidelines for developing new, potentially broader ones. Given the progress achieved so far and the promising benefits of longitudinal research and transdiagnostic approaches, it will be important to continue exploring the nature of outcomes for youth at-risk for conduct problems, both in terms of future course and breadth.

Dissemination of Cognitive-Behavioral Interventions (CBI) for Conduct Problems

Developing an intervention and documenting its beneficial effects through rigorous research does not ensure that the intervention will become available in routine clinical practice. Disseminating an evidence-based CBI requires focused effort, especially to facilitate high quality and sustained implementation by a wide range of real world providers. Implementation science has developed to study the factors that facilitate the spread of CBI’s beyond the academic setting (Eccles & Mittman, 2006). There is a growing body of implementation science research related to the dissemination of CBI’s for child conduct problems (Boxmeyer et al., 2014).

Training in CBI’s for conduct problems.

Studies have found that approaches to training providers to deliver cognitive behavioral interventions that include more or more frequent consultation result in better intervention outcomes (Sholomskas et al., 2005). This has been examined specifically in dissemination research on cognitive behavioral interventions with children with conduct problems, and this can be an area for future research. Effective elements of consultation following training in cognitive-behavioral therapy have included: the connectedness of the consultant and other therapists; authentic interactions around actual cases; and the responsiveness of the consultant to the needs of individual therapists (Beidas et al., 2013).

With regard to training during dissemination, Lochman and colleagues (2009) compared the effects of three levels of training for school mental health professionals who were learning to implement the Coping Power intervention. School counselors in 57 schools were randomly assigned to one of three conditions: 1) basic training in Coping Power, which included 3 days of initial workshop training and monthly supervisory meetings while delivering the intervention; 2) an enhanced training condition in which counselors also received monthly feedback on implementation fidelity and quality based on review of recorded sessions, and had access to ongoing technical assistance via telephone or email; 3) a service-as-usual comparison condition. Importantly, the basic training condition, which was more intensive than the typical professional development approach (consisting of a 1- or 2-day workshop without any follow-up consultation) did not yield significant improvement in children’s social or behavioral outcomes. Children whose counselors received more intensive training, which included feedback on recorded sessions and access to ongoing consultation, exhibited lower levels of teacher- and parent-rated externalizing behavior problems, lower rates of child-reported aggressive behaviors, and were less likely to expect that aggressive behaviors would lead to positive outcomes for them. The quality of implementation was higher for counselors who received intensive training.

Organizational climate, leadership and personnel characteristics.

Leadership and organizational climate are important aspects of the ‘support system’ for developing and sustaining high fidelity implementation of CBI’s in systems and organizations (Aarons, Ehrhart, Farahnak, & Sklar, 2014). A range of characteristics can influence the implementation and outcomes of CBI’s, such as leadership style and commitment to program implementation, counselors’ personalities, the social environment of the organization, and the relationships among individuals in the work setting (e.g., Aarons et al., 2014).

In the large-scale dissemination trial of Coping Power described above, a second study found that the degree and quality of implementation of Coping Power was influenced by characteristics of the school staff members who were trained to provide the intervention and by the climate of the schools in which the program was delivered (Lochman, Powell et al., 2009, 2015). Counselor agreeableness was positively associated with several measures of program implementation, including completion of session objectives, the number of sessions scheduled, and engagement with parents. Counselor conscientiousness was positively associated with engagement with children. School-level characteristics also predicted implementation outcomes. Counselors who were cynical about organizational change had poorer quality of engagement with children and parents, in particular when they worked in schools with environments that allowed staff members limited autonomy and that had greater managerial control. Dissemination of CBI’s into real-world settings must take into account the social contexts (Aarons et al., 2014), especially their interaction with individual provider characteristics such as agreeableness, conscientiousness and level of cynicism about organizational change. Dissemination of evidence-based interventions may require careful screening and training of clinicians.

Developing a workforce skilled in common elements of CBI’s.

Rather than expecting school and healthcare organizations to provide training for mental health professionals on a variety of CBI’s, a different approach has emerged. This approach is to identify common elements of CBI’s for the most prevalent child mental health problems and to train providers on these common treatment elements (Barth et al., 2012). Garland and colleagues (2008) documented the common elements of CBI’s for children with conduct problems. These included the specific skills taught in parent-directed interventions (e.g., parent-child relationship building, positive reinforcement, effective limit-setting and punishment) and youth skills training interventions (e.g., anger management, problem solving skills, affective education), as well as the specific therapeutic approaches utilized to teach these core skills. An important future direction is to create a highly skilled workforce by emphasizing graduate training on common elements of CBI’s for the most prevalent mental health problems (including conduct problems), rather than expecting school and healthcare organizations to seek professional development opportunities for their providers on a wide range of different CBI’s.

Access to CBI’s and future directions.

One way access to care issues are being addressed is by integrating mental or behavioral health treatment with primary care (Asarnow et al., 2015). Collaborative care models move beyond co-location of mental and physical health care providers and provide a framework for a collaborative treatment team to draw upon shared knowledge, principles, and care plans as the team works toward shared patient health and behavioral health goals (e.g., Richardson et al., 2014). Project LAUNCH (Linking Actions for Unmet Needs in Children’s Health) is an initiative of the Substance Abuse and Mental Health Services Administration to support behavioral health integration into primary care settings for children, among other activities (SAMHSA, 2017). Behavioral health integration is described as the care that results from a practice team of primary care and behavioral health clinicians, working together with patients and families, using a systematic and cost-effective approach to provide patient-centered care for a defined population. Some states have launched initiatives to embed child mental health specialists into primary care practices to enhance screening and coordinate services and supports for families. Parenting interventions such as the Positive Parenting Program (Triple P; Sanders et al., 2008) and Brief Behavioral Intervention for Preschool Children with Disruptive Behaviors (Axelrad & Chapman, 2016) that include the core elements of CBI’s for conduct problems are being offered more broadly via such initiatives.

Optimizing Interventions

In recent years, there has been an emphasis on developing more personalized approaches to healthcare – approaches that are tailored to the individual and that take into consideration biological, environmental, and lifestyle factors (Glenn et al., in press). Emerging research suggests that psychological and biological characteristics of the youth participating in the intervention likely influence “what works for whom” (Albert et al., 2015). A precision intervention approach based on research on moderators and subgroups involves adjusting behavioral interventions to the personal characteristics of high-risk individuals and targeting ‘precise’ groups. This approach involves optimizing interventions, and has the potential to lead to more cost-effective intervention programs that target the individuals most in need and that address the specific deficits of the individual. Methods for optimizing interventions can involve the strategic use of microtrials to increase understanding of mechanisms, a focus on tailoring of interventions, attention to clinical and technological formats for intervention delivery, careful attention to adaptation procedures, and integration of traditional CBI’s with other approaches, such as mindfulness training, which target related mechanisms.

Mini-intervention research.

In contrast to many of our prior randomized-control trials of interventions (which involved large-scale “cocktail treatments” or programs with multiple identified and unidentified mechanisms operating;Leijten et al., 2015), intervention research can also be guided by approaches that focus on microtrials or testing of smaller, more molecular mechanisms. Microtrials have been described as randomized experiments testing the effects of relatively brief and focused environmental manipulations designed to suppress specific risk mechanisms (Howe et al., 2010). Several examples exist. Phillips and Lochman (2003) compared single-session manipulations to address either reactive aggression or instrumental aggression with emotional regulation or consequence-focused mini-interventions (drawn from Coping Power program units), and found these simple interventions to change the two different types of aggression during a game task in expected ways, supporting assumptions about how these mechanisms worked in the larger program. Hiemstra et al. (2018) have used a cognitive bias modification task to reduce hostile interpretations of facial expressions in a game context, leading to reduced hostile interpretations. Although there was no generalization to aggressive behavior from this brief manipulation, the findings suggested new directions for future interventions. Sometimes microtrials can raise questions about our assumptions about common program elements often embedded in large programs. For example, Leijten and colleagues (2015) found that labelled praise in a brief manipulation did not produce greater reduction in mild disruptive behavior problems than did unlabelled praise. Understanding how mechanisms actually operate can contribute to crisper large-scale interventions as well as to specific evidence-based practices.

Tailoring of existing CBI’s.

Interventions using tailored designs hold promise for enhancing engagement, improving provider satisfaction, creating briefer treatments, and promoting positive intervention outcomes, although adaptation and fidelity must be balanced (Schoenwald, Sheidow, & LeTourneau, 2004). Tailored interventions provide a menu of the treatment procedures that would be included in a standard manualized intervention, but the selection and arrangement of those procedures are tailored to support real-time adaptation to a youth’s strengths and needs, response to outcomes, and emergent issues (e.g., poor engagement, life stressors; e.g., Weisz et al., 2012). Weisz et al. (2012) reported that youth treated in the MATCH (Modular Approach to Therapy for Children) condition improved significantly faster than youth in Standard or untreated conditions on internalizing and externalizing outcomes.

The Family Check-up (FCU) is a brief assessment-driven intervention that can be used to assist tailoring of existing interventions, and has led to reductions in child externalizing behaviors (Connell, Dishion, et al., 2007). In a recent pilot example of the use of FCU to tailor a structured intervention, the Coping Power Program was incorporated with the FCU (Herman et al., 2012). Based on feedback from assessments, parents chose relevant intervention components to work on (rather than complete all 16 sessions). The clinician worked with families in Baltimore after the assessment session to develop a plan, identify applicable components to be covered, and establish parental commitment to the chosen plan. Preliminary findings from this study suggested that the integrated Family Check-Up-Coping Power model was acceptable and feasible for families of children with disruptive behavior (Herman et al., 2012).

Group vs individual format for intervention delivery.

It is important to consider possible iatrogenic effects when providing an intervention targeting disruptive behavior in a group delivery format because of concerns about deviancy training (peer reinforcement of children’s deviant talk and behavior; Dishion, Poulin, & Burraston, 2001). Due to Coping Power typically being delivered in small groups, a large-scale study was conducted to investigate how children fare if they received Coping Power in a group versus individual format (Lochman et al., 2015). Results indicated both intervention delivery methods led to similar significant reductions in parent-rated externalizing problems through a one-year follow-up period. However, although teacher-rated externalizing problems also declined significantly for both intervention conditions, the reductions were significantly greater for children receiving Coping Power in an individual format. The main effect was moderated by children’s baseline levels of inhibitory control. Children with fewer problems with inhibitory control responded in similar positive ways to either the group or individual format. Subsequent studies found that children receiving the group intervention who were less prone to social reward (A/A genotype of an oxytocin receptor gene; Glenn et al., 2018), and had better emotional regulation (respiratory sinus arrhythmia; Glenn et al., in press) had better teacher-rated outcomes than their peers.

Internet delivery and intervention length.

One central structural barrier for utilization of mental health services is that intervention can be perceived by participants and practitioners to be too demanding and too lengthy (Kazdin, Holland, & Crowley, 1997). There are encouraging indications that briefer interventions can be effective. A briefer version of Coping Power (24 child sessions, 10 parent sessions) has produced significant reductions in teacher ratings of children’s externalizing behaviors at longer-term follow-ups (Lochman et al., 2014), similar to the effects for the full program. An innovative way of offering briefer interventions is to include internet-based content, making the intervention more accessible and efficient (Taylor et al., 2008). Hybrid versions of existing manualized evidence-based interventions such as Coping Power, including much briefer versions of the programs carefully integrated with internet-based website activities can be created (Lochman et al., 2017). A hybrid version of Coping Power’s child component included 12 small group sessions (instead of the regular 34) and a website which included a brief animated and humorous cartoon series, The Adventures of Captain Judgment, developed to specifically illustrate Coping Power concepts and skills. Relative to a randomized control group, children receiving the hybrid program had lower rates of conduct problem behaviors than untreated control children. Use of technology can also enhance the cost-effectiveness of interventions. The hybrid version of Coping Power had a 60% reduction in the frequency of counselors’ face-to-face meetings (a savings in cost of intervention of 44%) while still producing significant intervention effects on children’s conduct problems.

Video games as intervention tools.

Serious games are video games that are developed with the primary purpose of educating or teaching skills, and a recent meta-analysis found serious games in mental health interventions enhanced engagement and learning (Lau et al, 2017). Serious games can be effective at teaching children new coping skills (Kousmanen et al., 2017). The SPARX-R video game utilizes principles from CBI to enhance emotional regulation in children (Kousmanen et al., 2017), and game play was found to be associated with increased use of positive emotional regulation strategies. However, future research is needed to determine if serious games can influence children’s real-life aggression.

Integration of mindfulness training and CBI.

A growing literature shows that mindfulness interventions with high-risk youth may be feasible and effective. Mindfulness interventions have reduced children’s ADHD symptoms and externalizing problems (van der Oord et al., 2012). An overarching premise of mindfulness involves bringing awareness to one’s present experience. Mindfulness is theorized to increase attentional control and self-regulation (Metz et al., 2013) by creating “space” between perceptions and responses to stress to improve affect tolerance and decrease cognitive and behavioral avoidance. Based on this conceptualization, a feasibility study found that Coping Power integrated with mindfulness training had particular effects in improving at-risk children’s emotional, cognitive, and behavioral regulation, and their self-reports of reactive aggression (Miller et al., 2018)

Adaptation for different cultures.

It is important for interventions to address children coming from diverse ethnic, cultural, and community backgrounds. Cognitive-behavioral interventions may be limited by cultural constraints, and thus need to be carefully adapted. Goldstein, Kemp, Leff and Lochman (2013) outline a stepwise approach for adapting manualized interventions for new target populations. Following a stepwise approach, Coping Power has been adapted for use in other countries and territories. A Dutch version of the Coping Power Program (Van De Wiel et al., 2007; Schuiringa, van Nieuwenhuijzen, Orobio de Castro, Lochman, & Matthys, 2017) produced greater reductions in overt aggression following treatment, and lower marijuana and tobacco use at a four-year follow-up, compared to children in the care-as-usual control condition (Zonnevylle-Bender, Matthys, van de Wiel, & Lochman, 2007). Adaptations of Coping Power have been tested in Canada (Ludmer et al., 2018), Italy (Muratori et al., 2017), Pakistan (Mushtaq, Lochman, Tariq & Sabih, 2017), Sweden (Helander et al., 2018), and Puerto Rico (Cabiya et al., 2008), with greater reductions in conduct problems for children in the intervention conditions than in control conditions. These adaptations have ranged from concrete level translations of the program into a new language, to deeper level adaptations that have required changes to how material is presented. For example, in the Pakistani adaptation of the program, the manual was translated (into Urdu), and the cognitive and emotional skills were presented within the context of Islamic stories and practices (Mushtaq et al., 2017).

Therapeutic Process and Therapist Characteristics in CBT Implementation and Outcomes

Therapeutic alliance is a multifaceted construct that includes the emotional bond between child and therapist, as well as their ability to work together in a collaborative relationship (e.g., Shirk, Karver, & Brown, 2011). In this conceptualization of therapeutic alliance, the establishment of a positive emotional connection is thought to encourage children’s willingness to work cooperatively with the therapist on treatment activities. For older children and adolescents, agreement on treatment goals and therapeutic procedures may also be important aspects of therapeutic alliance. Because treatment for older children and adolescents is likely to be initiated by others (parents, teachers, etc.), resistance is common in this group of youth and resolving treatment-related issues may be particularly relevant to therapeutic alliance. In a meta-analysis of therapeutic alliance in the youth treatment literature, Karver, De Nadai, Monahan, and Shirk (2018) reviewed 28 studies that prospectively evaluated the relation of therapeutic alliance to treatment outcomes. Results revealed a small to medium effect size (r = 0.19). For children specifically receiving treatment for disruptive behaviors and conduct problems, outcomes improve under conditions of better therapeutic alliance (Kazdin & Durbin, 2012).

Children with disruptive behaviors who have better intellectual and social skills are able to form stronger alliances with their therapists (Kazdin & Durbin, 2012). Interestingly, among adolescents with externalizing problems, those with higher levels of callous-unemotional traits report a stronger degree of therapeutic alliance, and these gains in alliance are associated with improvements in behavioral symptoms after treatment (Mattos, Schmidt, Henderson, & Hogue, 2017). Youth with callous-unemotional traits are less likely to demonstrate the intellectual, social, and emotional deficits that often correspond with disruptive behavior, which may allow them to effectively form a relationship within the therapeutic environment (Mattos et al., 2017).

Children with a higher rate of externalizing behaviors can have special difficulties in developing a therapeutic alliance (e.g., Ayotte, Lanctôt, & Tourigny, 2015). Thus, establishing a therapist-child bond early in therapy may be particularly important for the development of the therapeutic alliance in aggressive children. Mitchell and colleagues (2017) examined coder ratings of therapeutic alliance across the implementation of Coping Power. Children who established a bond with their therapist early in the treatment process displayed greater reductions in teacher-reported disruptive behavior. Similarly, research on a sample of adolescents treated for depression revealed that therapeutic alliance in the first therapy session was a strong predictor of reduction in symptoms at session 4 (Labouliere, Reyes, Shirk, & Karver, 2017).

Children’s mental health treatment is affected not only by their own relationship with their therapists, but also by their parents’ relationship with the therapist (Kazdin, Whitely, & Marciano, 2006). Parents are often directly involved in their children’s treatment, and typically provide transportation to treatment, remind and/or assist children with therapeutic assignments, and serve as coaches to encourage children’s use of new skills. In treatment programs for disruptive behavior that include individual child and parenting components, a stronger parent-therapist alliance is associated with more improvement in parenting skills as well as greater symptom reduction child behavior problems (de Greef et al., 2017; Kazdin et al., 2006). Imbalances in therapeutic alliance can predict dropout (Robbins, Turner, Alexander, & Perez, 2003), as a closer alignment between one family member and the therapist may negatively impact other family members’ relationship with the therapist and their investment in the therapy process. In providing services to children and adolescents, much remains to be learned about the various roles, processes, and influencing factors related to family alliance issues.

Therapist factors, such as warmth, interest, openness, flexibility, care, compassion, empathy, respect, and honesty, are positively associated with the therapeutic relationship (Ackerman & Hilsenroth, 2003). Therapist confidence, professionalism, and perceived skill are also positively associated with therapeutic alliance (Ackerman & Hilsenroth, 2003). As noted earlier, therapist factors have also been associated with the quality of their service delivery, with their continued use of evidence-based programs after training, and with child behavioral outcomes (e.g., Lochman, Powell, et al., 2009, 2015). In addition, within treatment sessions, therapists’ warm, nonreactive, professional interactions with children have been associated with less increase in teacher-rated externalizing behaviors at a one-year follow-up (Lochman et al., 2017). Similarly, in Parent-Child Interaction Therapy sessions, parents whose therapists used a constructive criticism communication style used target skills at a higher level than parents whose therapists communicated in a positive or neutral manner (Herschell, et al., 2008). Importantly, this line of research elucidates malleable factors that may prove to be important targets for mental health professionals’ training programs. Strengthening targeted interpersonal and professional qualities in prospective therapists’ training may lead to stronger alliances with clients in practice, and in turn to improved treatment outcomes.

Summary

Cognitive-behavioral intervention research with children with conduct problems is deeply rooted in the history of CBI, and CBI has provided the basis for a number of evidence-based interventions for these children (e.g., Lochman & Matthys, 2018). While randomized trials of new large-scale interventions should continue to be pursued, the future of research in this area can also fruitfully focus on four central areas. First, research can address the newly-emerging interest in how single interventions can have meaningful impact on a range of transdiagnostic outcomes because the intervention mechanisms may affect those various outcomes. Second, we are in a rapidly emerging wave of research on how evidence-based programs can be disseminated in the real-world, using the tools of implementation science to examine key issues such as the adequacy of training approaches and the role of therapist and organizational characteristics. Third, a major focus of research in the coming decade should be on how to optimize intervention outcomes, and that can include a focus on microtrials, on tailoring of interventions, on examining rigorously how interventions are delivered, and on the integration of technology and of other approaches such as mindfulness training into cognitive-behavioral intervention. Fourth, research can explore how the therapeutic relationship and the therapists’ characteristics can play substantial roles in effective CBI with conduct problem children.