Chapter 6 Michael A. Southam-Gerow, Cassidy C. Arnold, Carrie B. Tully, and Julia Revillion Cox The field of psychological science has made enormous progress developing and testing psychological and combined medication/psychological treatments to address the increasing prevalence of psychopathology in children (e.g., Hoagwood & Olin, 2002). Indeed, a recent review found more than 350 randomized controlled trials testing treatments for childhood mental health problems (Chorpita, Bernstein, & Daleiden, 2011). However, there remains a significant number of children not receiving adequate treatments (e.g., Fulda, Lykens, Bae, & Singh, 2009; Tang et al., 2008), a reminder of the oft-lamented science–practice gap (e.g., U.S. Department of Health and Human Services, 2002). Researchers and other children’s mental health stakeholders have identified dissemination and implementation (D&I) of psychological treatments as a key goal for the field (e.g., Aarons, Hurlburt, & Horwitz, 2011; Proctor et al., 2009; Schoenwald & Hoagwood, 2001; Southam-Gerow, Rodriguez, Chorpita, & Daleiden, 2012). This chapter provides an overview of the D&I field related to child and adolescent mental health treatment and services research. Specifically, we provide a brief conceptual primer related to D&I science; provide a succinct rationale for why D&I has emerged as a focus; outline several prominent D&I models; and describe some recent empirical D&I studies. Given the relative novelty of D&I science, there is not yet absolute convergence on the definition of key terms. Our definitions are based on a careful reading of the literature, with particular emphasis on the work of Chambers, Ringeisen, and Hickman (2005); Fixsen, Naoom, Blasé, Friedman, and Wallace (2005); Proctor et al. (2009); and Rabin and Brownson (2012). First, we provide some clarification of three terms that are similar but distinct: diffusion, dissemination, and implementation. The term “diffusion” means the planned or unplanned/spontaneous spread of an innovation. In other words, “diffusion” refers to both the natural distribution of a new idea and the more intentional spread of that idea (Chambers et al., 2005; Rogers, 2003). The term “dissemination” refers to the directed and planned spread of an innovation or, as Chambers et al. (2005) wrote, the “targeted distribution of a well designed set of information” (p. 323). Because dissemination involves the consistent spread or distribution of information, such as a treatment approach, the process of dissemination involves identifying ways to communicate the innovation to maximize its consistent and accurate receipt by the desired end users. Finally, the term “implementation” refers to the processes and strategies needed to adapt the innovation (e.g., a treatment program) to fit within a specific context. This definition of implementation is similar to that provided by Fixsen et al. (2005) in their treatise on implementation research: “a specified set of activities designed to put into practice an activity or program of known dimensions” (p. 5). Fixsen et al. also distinguished among different degrees of implementation, ranging from paper (i.e., enacting policies consistent with the innovation), to process (i.e., creating procedures that enable training and supervision in the use of a new innovation), to performance implementation (i.e., creating procedures that identify ways to ensure that the innovation is being used properly and that its use is having the expected consequence—that is, benefit for the consumer). It bears mention that others have used the term in a more proximal sense to mean the enactment of a specific treatment approach or treatment strategy, thereby creating some confusion about this term. In general, in the context of D&I science, the term “implementation” has a broader sense and refers to the processes used to bring a program (or other innovation) into a new context. Thus, dissemination involves spreading the word about an innovation, whereas implementation involves specific processes and procedures designed to improve the adoption, correct use, and sustainment of the innovation in a specific place. In short, implementation is the how of dissemination. A few other terms often used in relation to D&I research warrant brief mention. The term “technology transfer” has long been applied to the process of taking scientific findings and adapting them to have broader applications for public use and/or for sale in the commercial sector. Along these lines, then, D&I research can represent a specific example of technology transfer. Additionally, the term “translational research” is used to refer to work that translates “bench science” for the bedside (and vice versa), a process that fits with the aims of D&I science. It is noteworthy that a dictionary of terms is needed for dissemination and implementation. Obviously, the field of treatment of child/adolescent mental health problems is not a build-it-and-they-will-come context. Simple dissemination efforts like publishing papers and generating lists of evidence-based treatments (EBTs) have not been sufficient methods for widespread adoption of EBTs. D&I science has emerged because of the need to identify ways to implement treatments in a variety of settings. But why have simple dissemination strategies failed? That is, why has strong evidence about potent treatments for child/adolescent mental health problems not generated a mass effort on the part of therapists and agencies to adopt these treatments? In this section, we grapple with this question. Most treatment programs are developed for a single, specific child disorder (e.g., obsessive-compulsive disorder) or problem type (e.g., disruptive behavior). The focus on a single child problem is consistent with decades of medical research and has led to incredible and important developments in the science of mental health treatment for children and adolescents. However, as the field has moved toward widespread dissemination of treatments developed in this manner, some limitations of the approach have become apparent. Specifically, multiple factors beyond the child’s disorder appear to influence how potent a treatment will be, as has been described by several scientists (e.g., Damschroder & Hagedorn, 2011; Fixsen et al., 2005; Proctor et al., 2009; Schoenwald & Hoagwood, 2001; Southam-Gerow et al., 2012). For example, Southam-Gerow et al. (2012) describe an array of factors that may have an impact on how well a treatment works in a particular setting; these include child and family factors, therapist factors, organization factors, and service system factors. We review each of these briefly and discuss how they might function as barriers to simple dissemination strategies. A number of child- and family-specific variables can influence how well psychological treatments for youth work. Many of these variables are largely neglected as targets of treatment and/or remain underrepresented within efficacy study samples. Past research efforts across different primary diagnoses have demonstrated that children served by community clinics often present with comorbidities, impaired academic and social functioning, and other stressors (Ehrenreich-May et al., 2011; Southam-Gerow, Chorpita, Miller, & Gleacher, 2008; Weersing & Weisz, 2002), factors that are often unaddressed by manualized treatments (Hammen, Rudolph, Weisz, Rao, & Burge, 1999). When compared to those referred to university-based research clinics, children receiving services in the community are more likely to have parents with less education (Southam-Gerow et al., 2008) and lower incomes (Ehrenreich-May et al., 2011). Single-parent families are also more frequent among this population (Southam-Gerow, Weisz, & Kendall, 2003), and, even when controlling for geographic differences, ethnic minority families are overrepresented (Ehrenreich-May et al., 2011). Echoing these findings, one study of youth receiving school-based services found higher rates of trauma and past suicide attempts when compared to efficacy studies (Shirk, Kaplinski, & Gudmundsen, 2009), highlighting the many ways that clinical complications can impede successful child outcomes. These client and family variable differences have contributed to the rise of D&I science. Efficacy studies attempt to ensure the integrity and maximize the dose of a given treatment by employing therapists who are specially trained and receive ongoing supervision and consultation (usually doctoral-level students/professionals). In contrast, master’s-level therapists with varied training backgrounds comprise the majority of the workforce in community mental health settings (Garland, Kruse, & Aarons, 2003; Weisz, Chu, & Polo, 2004). This difference may influence how an EBT will fare when transported to community settings for a variety of reasons. Community providers’ motivation to learn and use EBTs is variable: Several studies have documented clinicians’ concerns that manualized treatments are inflexible and inhibit individualized case conceptualization and treatment planning (Addis & Krasnow, 2000; Becker, Zayfert, & Anderson, 2004; Walrath, Sheehan, Holden, Hernandez, & Blau, 2006). While openness to and knowledge about EBTs may facilitate the implementation at the therapist level (Aarons, McDonald, Sheehan, & Walrath-Greene, 2007), recent surveys of the attitudes of community-based clinicians toward evidence-based practice have not revealed consistent correlates by theoretical orientation, training, or years of experience (Nakamura, Higa-McMillan, Okamura, & Shimabukuro, 2011). Yet another element to consider is the limited effectiveness of specialized training in EBTs. Self-directed learning and workshops, both in person and online, are by far the most popular training vehicles: Several studies have documented significant gains in clinician knowledge and self-reported efficacy (see Beidas & Kendall, 2010, for a review). There is, however, very little evidence that these methods result in substantive behavioral change (Beidas & Kendall, 2010; Herschell, Kolko, Baumann, & Davis, 2010). However, the use of behavioral strategies, including observation, feedback, ongoing consultation, and coaching, has been shown to increase EBT adoption and impact client outcomes (Becker & Stirman, 2011; Herschell et al., 2010). Therapists within the community are often part of an agency or organization, each with its own unique characteristics and culture that can influence the successful implementation of EBTs. As Glisson et al. (2012) discuss, organizational climate (e.g., norms, expectations) and policies drive clinician behavior (e.g., willingness and ability to pursue training opportunities, or resistance) and attitudes (e.g., job satisfaction, morale). Beyond the diffuse impact on therapist attitudes and behavior, organizations are integral to successful implementation of EBTs in these ways: This organizational infrastructure supports ongoing training, fidelity monitoring, and supervision/coaching in an effort to sustain meaningful change (Fixsen et al., 2005). Just as therapists work within the parameters of an organization, all community mental health organizations are subject to influences at the system level, what Aarons et al. (2011) refer to as the outer context. These influences include relevant local, state, and federal policies; the availability and priorities of funding sources (e.g., public and private insurance, community resources); referral mechanisms; legal obligations or mandates; relationships with other agencies and organizations; and the needs of local mental health consumers (Fixsen et al., 2005; Schoenwald & Hoagwood, 2001). These distal variables can have profound influence on implementation efforts (e.g., Aarons & Sommerfeld, 2012; Metz & Bartley, 2012). Thus, the short answer to why the field has strongly emphasized D&I science in recent years is that myriad complex factors can influence whether and how EBTs (or any new treatment or other innovation) are used in community settings. Developing methods to implement EBTs in community settings, thereby maximizing the public health benefits of our considerable science on treatments, represents the focus of D&I science. Given the complex nature of this work, there has been effort to identify models and frameworks to guide work. In the next section, we consider the most influential current models. As reflected in the previous section, the challenges facing the dissemination and implementation of psychological interventions in community practices are numerous. As a result, the availability of EBTs, directives to use EBTs, and other “push” forces (i.e., forces on the research side of the science–practice gap) seem to be insufficient on their own (e.g., Proctor et al., 2009; Southam-Gerow et al., 2012). At the same time, “pull” forces (i.e., forces on the community side of the gap), such as patients requesting a particular EBT or practitioners requesting training in, organizational support for, and reimbursement of EBT, do not seem to occur organically at a sufficient strength to promote a high rate of dissemination and implementation of EBTs. The sophistication of D&I research and the models that describe them is at an interesting stage. On one hand, the development of D&I models and their use in D&I projects, along with measures of the key variables within models, increases the interpretability of findings from these efforts (Tabak, Khoong, Chambers, & Brownson, 2012). However, the abundance of models and their relative lack of empirical evaluation mean that it is currently impossible to definitively support the use of one model over another (Proctor et al., 2009). To move beyond this stage, for the empirical support to grow and support the use of the best models, more D&I projects that use and compare models with strong research designs are needed. The remainder of this section describes six D&I models, selected because they demonstrate the breadth of characteristics within the D&I models: Rogers’ (2003) diffusion of innovations model is the broadest one presented here and was not originally developed to explain or understand the dissemination of mental health treatments. The model describes the spreading of an idea, practice, or object that is new or perceived to be new to the unit of adoption (e.g., individuals, therapists, hospitals, etc.) through a social system. Within this model, the term “diffusion” refers to the ways an idea moves through a social system via communication between parts or individuals within the system (Rogers, 2003). Unlike the other models discussed in this section, the four factors in Rogers’ model do not explicitly include (or prohibit) effortful processes or require individuals who actively drive the innovation into new areas. That is, the model emphasizes both intentional implementation efforts as well as less (or even un-) intentional dissemination of innovations. The four primary factors emphasized in Rogers’ model are: The communication channels through which information can spread also influence the dissemination of an innovation. Different communication channels (e.g., mass media and interpersonal channels) spread information at different rates, reach different potential adopters, and influence the adopter differently. Time, in Rogers’ model, refers to the lapse between when adopters first learn of an innovation and when they decide to adopt or reject the innovation, the earliness or lateness of adoption versus other members of the system, and the rate at which adoption takes place across the system. The fourth and final major factor is the social system: the complete collection of potential adopters of a given innovation (Rogers, 2003). Within the mental health field, this may be therapists within a community or practice, community mental health centers, localities and states establishing policy relevant to mental health treatment, and/or insurance companies that pay for mental health services. According to Rogers, key aspects of social systems, such as local norms and key opinion leaders, significantly influence the rate of diffusion within them. Although not originally developed with mental health services in mind, Rogers’ model has important implications for D&I efforts. For example, consider the fact that the design of an EBT (i.e., more versus less flexible) appears to influence potential adopters’ perception of that EBT as well as the outcomes achieved by those adopters (Borntrager, Chorpita, Higa-McMillan, & Weisz, 2009; Weisz et al., 2012). The therapist training literature described earlier (e.g., Beidas & Kendall, 2010; Herschell et al., 2010) supports the importance of Rogers’ notion of communication channels. Finally, key opinion leaders (i.e., influential individuals within their social network and whose behavior serves as a model for others) have been found to be instrumental in the implementation of teacher-delivered mental health interventions in schools serving low-income minority children with behavioral problems (Atkins et al., 2008). In their 2005 monograph, Fixsen and colleagues synthesized the dissemination and implementation literature from diverse fields in an effort to provide both an overarching theoretical framework and a concrete step-by-step description of the implementation process, with an eye toward applying their findings to mental health services. While the authors pull broadly from the human services and technology fields, their model for implementation occurs within the context of a specific community (e.g., agency, city) that has unique needs, assets, and challenges. The broad “conceptual framework” for implementation of programs or practices includes five essential components: Building on the work of Rogers and others, Fixsen et al. (2005) also delineate more concrete stages of the implementation model that describes six categories: Along with a thorough review of implementation research efforts, Fixsen et al. (2005) also identify core intervention and implementation components. Core intervention components include the “active ingredients,” the essential techniques and principles of the treatment program being implemented. Similarly, core implementation components are the elements that contribute to the successful implementation of a program with fidelity. These components include training and consultation, selection of staff to directly implement the treatment model, monitoring treatment fidelity, and evaluating program outcomes. The mental health systems ecological (MHSE) model was developed explicitly with children’s mental health services in mind. As discussed earlier, multiple factors influence how potent a treatment will be in treating a child’s mental health problems. The MHSE model (e.g., Schoenwald & Hoagwood, 2001; Southam-Gerow et al., 2012; Southam-Gerow, Ringeisen, & Sherrill, 2006) specifically outlines the importance of four different levels of the ecology to consider when planning D&I science: (1) child and family factors, (2) therapist factors, (3) organization factors, and (4) service system factors. As will be clearer shortly, this broad ecological approach is consistent with other models (e.g., Aarons et al., 2011; Damschroder & Hagedorn, 2011; Fixsen et al., 2005; Proctor et al., 2009).
Dissemination and Implementation of Evidence-Based Treatments for Children and Adolescents
CONCEPTUAL PRIMER ON DISSEMINATION AND IMPLEMENTATION
WHY IS DISSEMINATION AND IMPLEMENTATION SCIENCE NEEDED?
Child and Family Factors
Therapist Factors
Organizational Factors
System Factors
MODELS OF DISSEMINATION
Rogers’ Diffusion of Innovations Model
Fixsen et al.’s Implementation Framework
Mental Health Systems Ecological Model
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
