Background: Adolescence is considered a critical period for the development of higher-order cognition (Larsen & Luna, 2018). It is also a sensitive period for psycho-social development during which the individual’s life may rapidly pivot for better or worse (Blakemore & Mills, 2014; Dahl et al., 2018). The global prevalence of adolescent mental health problems has increased with dire speed for years (Inchley et al., 2020; WHO, 2018; WHO, 2022). Adolescents residing under care of the child welfare or juvenile justice system are especially vulnerable because they are already characterized by severe, persistent, and co-occurring mental disorders, far exceeding that of their peers (Beaudry et al., 2021; Bronsard et al., 2011; Jozefiak et al., 2015; Schmid et al., 2008; Shufelt & Cocozza, 2006; Underwood & Washington, 2016; Vinnerljung & Sallnäs, 2008; Zajac et al., 2015). Within residential facilities, adolescents with conduct or substance use problems that co-occur with other mental disorders may encounter mental health care, milieu therapy, blunt detainment, and everything in between. The pronounced overlap between emotional and behavioral difficulties in adolescents across residential settings call for transdiagnostic approaches (i.e., targeting mechanisms that contribute to or sustain multiple disorders; Harvey et al., 2004). Another similarity applies to the systems that serve these adolescents: Mental health care, child welfare, and juvenile justice facilities are usually part of complex systems with disparate mandates, responsibilities, and resources. Thus, where adolescents are placed depends on their behavioral expressions but, importantly, it also depends on the layout of the systems they encounter. Practitioners across these systems have high workloads, high levels of job-related stress, and high turnover. Taken together, this makes implementation of novel and evidence-informed practices in these contexts challenging but crucial processes. The similarities between adolescents across settings, and the ensuing need for implementing novel, evidence- informed transdiagnostic interventions can be seen as an excellent opportunity for cross-pollination of knowledge, practices, and research.
Aims: This thesis is an attempt at exploring if and how integrating knowledge from multiple disciplines can promote effective, equitable, and multisectoral care for adolescents. It comprises three articles, each with distinct objectives and methodologies, investigating adolescents with co-occurring problems, the systems they encounter, the challenges associated with implementing interventions for this population, and the research that is for and about them. Ranging from a global overview, via identification of common treatment elements, to a case study of implementation of a transdiagnostic intervention in the Norwegian child welfare system, the three papers address macro- meso- and micro-aspects of research and practice regarding interventions for service-involved adolescents with co-occurring problems. Paper 1 is a review of systematic reviews (i.e., an overview) on transitional interventions for adolescents leaving residential facilities. Paper 2 is a review and meta-analysis, comparing two distinct methodologies to identifying common treatment elements in therapeutic interventions for adolescents across residential settings. Paper 3 is a qualitative exploration of a state-level child welfare initiative to implement a transdiagnostic therapeutic model to municipal child welfare agencies in Norway.
Results: In the first paper, we reviewed eight systematic reviews and 58 primary studies, resulting in the identification of three important gaps of knowledge on transitional interventions for adolescents. First, 95 % of study populations were from the USA and only 13 % of interventions reported on in primary studies were delivered after 2010. Second, seven of the eight systematic reviews were of low or critically low methodological quality and the confidence in effect estimates were low or very low for all outcomes. Third, recidivism was reported as the only outcome in five systematic reviews even though 30 primary studies reported on interventions targeting mental health, substance abuse, social skills problems, education, and employment. The operationalizations of recidivism were vastly disparate, ranging from new convictions, via new court contacts, charges, and arrests, to alleged offenses. In the second paper, we identified 11 common elements across 24 studies of treatments targeting externalizing and/or internalizing symptoms in adolescents living in residential facilities. The paper demonstrates the utility of two different approaches to identifying common elements across studies, one of which can establish statistically significant associations between the presence and absence of an element and treatment effect. The overall treatment effects of all trials were small-to moderate. The two common elements Emotional recognition and differentiation and Psychoeducation were significantly associated with beneficial treatment effects for externalizing symptoms. These elements may be eligible for further experimental testing in future transdiagnostic interventions. Importantly, by contrasting two approaches to identifying common elements, the paper shows the importance of including all effect estimates including those of null or iatrogenic effects, in generating hypothesis about which common elements are most likely to yield beneficial outcomes across disorders and settings. While the first two papers grapple with the methodological quality of research on, and the content of interventions for adolescents who are or have been placed in residential settings, the final paper addresses the issue of implementation in the complex service settings that adolescents often encounter. Through interviews with community child welfare workers, we identified five themes that describe salient issues regarding the implementation of a transdiagnostic intervention for adolescents. The themes reveal that a flexible, methodical, and evidence-informed model for frontline child welfare work is appropriate for and well received by Norwegian practitioners in small communities. The themes also highlight the importance of a pre-meditated and rigorous implementation scheme that includes the active participation of users (welfare workers and families) for the intervention to be sustainable and effective.
Implications: Taken together, the three papers contribute with distinct but interrelated findings of relevance to research on interventions for service involved adolescents with co-occurring problems. There is a dire need for updated research on interventions provided to adolescents, especially those living in and transitioning from residential settings. The first paper points to the potential value of comparing the effect of interventions across residential settings, provided that novel research efforts report outcomes beyond recidivism. The second paper adds to the common elements literature by spearheading the multilevel meta-analytic approach, demonstrating its appropriateness for establishing associations between common elements and treatment effects. This approach enables comparison of interventions across settings, treatments, and outcomes, allowing for these variable to be investigated as moderators. Therefore, multilevel meta-analyses are one promising avenue toward overcoming the siloes that separate research on populations who have similar and co-occurring problems. Finally, the third paper reifies how important implementation is for an intervention to be adopted and sustained. The results provide clear-cut descriptions of how inter-agency collaboration in Norwegian child welfare agencies can and should be modified to enhance implementation success. Overall, I argue that it is time researchers, policymakers, and practitioners stop viewing adolescents with co-occurring problems in terms of the system they encounter. Most of them have much in common and we must use available methods that enable comparison across settings and disorders.