P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
Style file version July 26, 1999
C 2002) Journal of Abnormal Child Psychology, Vol. 30, No. 1, February 2002, pp. 1–17 (°
The Implementation of the Fast Track Program: An Example of a Large-Scale Prevention Science Efficacy Trial The Conduct Problems Prevention Research Group1 Received December 13, 2000; revision received July 9, 2001; accepted July 12, 2001
In 1990, the Fast Track Project was initiated to evaluate the feasibility and effectiveness of a comprehensive, multicomponent prevention program targeting children at risk for conduct disorders in four demographically diverse American communities (Conduct Problems Prevention Research Group [CPPRG], 1992). Representing a prevention science approach toward community-based preventive intervention, the Fast Track intervention design was based upon the available data base elucidating the epidemiology of risk for conduct disorder and suggesting key causal developmental influences (R. P. Weissberg & M. T. Greenberg, 1998). Critical questions about this approach to prevention center around the extent to which such a science-based program can be effective at (1) engaging community members and stakeholders, (2) maintaining intervention fidelity while responding appropriately to the local norms and needs of communities that vary widely in their demographic and cultural/ethnic composition, and (3) maintaining community engagement in the long-term to support effective and sustainable intervention dissemination. This paper discusses these issues, providing examples from the Fast Track project to illustrate the process of program implementation and the evidence available regarding the success of this science-based program at engaging communities in sustainable and effective ways as partners in prevention programming. KEY WORDS: conduct-problems; implementation; prevention.
Prevention science begins with the assumption that effective prevention efforts will target risk and protective factors identified by developmental theory and research as causally associated with the emergence and course of a disorder (Coie et al., 1993; Institute of Medicine, 1994). The prevention of conduct disorder is well-suited to a prevention science approach toward community-based preventive intervention, because the critical elements needed to support the application of this model exist (Weissberg & Greenberg, 1998). That is, a rich prospective data base
has elucidated the epidemiology of risk for conduct disorder, and identified causal processes linking individual, family, school, and neighborhood characteristics to risk and protection (Loeber et al., 1993; Patterson, Capaldi, & Bank, 1991). In addition, several promising intervention strategies have been developed, demonstrating evidence of effectiveness in short-term trials targeting single risk and protective factors associated with the development of conduct disorder (Kazdin & Weisz, 1998). Yet, despite evidence that child-conduct problems can be reduced in short-term clinical trials, only one Canadian study to date suggests that conduct disorder can be prevented effectively in longer-term trials (Kazdin, 1987; Tremblay, Masse, Pagani, & Vitaro, 1996). In addition, processes of implementation and the impact of programs preventing conduct problems have not yet been examined in communities that vary widely in terms of their demographic characteristics and cultural heritage. Thus, one critical “next step” in this field was the implementation and evaluation of a large, multicomponent, long-term trial.
1 Members
of the CPPRG in alphabetical order include Karen L. Bierman (Pennsylvania State University), John D. Coie (Duke University), Kenneth A. Dodge (Duke University), Mark T. Greenberg (Pennsylvania State University), John E. Lochman (University of Alabama), Robert J. McMahon (University of Washington), and Ellen Pinderhughes (Vanderbilt University). 2 Address correspondence about this article to Karen Bierman, Department of Psychology, Pennsylvania State University, State College, Pennsylvania 16801. Address requests for copies of this article to Robert J. McMahon, University of Washington Fast Track, 146 N. Canal Street, Suite 111, Seattle, Washington 98103.
1 C 2002 Plenum Publishing Corporation 0091-0627/02/0200-0001/0 °
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
2 In 1990, the Fast Track project was initiated to evaluate the feasibility and effectiveness of a comprehensive, multicomponent prevention program targeting children at risk for conduct disorders (Conduct Problems Prevention Research Group [CPPRG], 1992). The Fast Track program involves a developmentally-based, long-term, comprehensive intervention designed to prevent conduct problems in adolescence, applied in four demographically diverse communities. The evaluation design includes a high-risk control group and a normative comparison group being followed longitudinally at each of the four sites. The implementation of this large-scale prevention science efficacy trial has faced several key challenges. In previous efforts, the prevention science approach has been criticized for generating short-term programs that are not ecologically valid or sustainable (Weissberg & Greenberg, 1998). Fast Track was designed to be a longterm, multifaceted community-based prevention program, which could be embedded within existing community support structures. As such, it represented a test case, illustrating the challenges facing a prevention science approach when taken to the level of a large-scale efficacy trial. Central challenges included: (1) the engagement of community members and stakeholders; (2) the maintenance of intervention fidelity balanced with the flexible adaptation to local norms and needs; and (3) the long-term retention of participants, maintenance of community engagement, and creation of a foundation for sustainability. Community psychologists have suggested that the involvement of local community members as central participants in the design of prevention may be critical to the long-term success of the programs, as such involvement fosters better understanding of the intervention goals, greater commitment to high-quality implementation, and more appropriate adaptation of programs (Connell, Kubisch, Schorr, & Weiss, 1995). Of necessity, the prevention science approach involves some “top-down” planning, as research plays a central role in determining the selection of targeted risk and protective factors and intervention strategies. Hence, a critical challenge facing Fast Track involved attaining the engagement of community members and key stakeholders given that intervention planning occurred with “top-down” direction, with researchers playing a central role in the design of the prevention program. That is, Fast Track was committed to a theoretical model that prespecified the desired change targets, change agents, and change methods, and, hence, faced the challenge of engaging community members around this particular prevention model. A second concern raised regarding the prevention science approach is whether programs based on controlled empirical trials can have the flexibility in form and
Style file version July 26, 1999
The Conduct Problems Prevention Research Group function needed for ecological validity (Dumas, 1989). That is, the concern exists whether university-developed programs based on empirical linkages regarding risk and outcomes can offer sufficient flexibility to adapt to varying contexts and communities—adaptation which may be crucial to the effective implementation and evaluation of prevention services (Elias & Branden, 1988). The issue is that, although experimental research may provide useful intervention materials and strategies, contextual factors may affect the motivation of community members, as well as the effectiveness of different prevention strategies, which may be influenced by different ecologies (Dumas, 1989). The challenge to Fast Track, then, was to deliver a prevention program that maintains fidelity regarding the critical change targets and processes, but offers flexibility for local adaptation in implementation. Finally, prevention science approaches are often time-limited, lasting through the course of a short-term trial, but not beyond (Weissberg, 1995). The short-term nature of many prevention science efforts raises questions regarding the feasibility of making sustainable changes with programs developed “outside” the community, and whether such programming can create systemic changes in schools or community systems to support longer-term prevention efforts. The Fast Track program involves a 10-year span of prevention activities, making the completion of the trial, itself, a long-term engagement with schools and communities. The challenge was to retain the interest and active engagement of families, youth, and schools over this sustained period, as well as to build a foundation of support for the continuation of prevention efforts following the end of the trial. In describing the tenets of the prevention science approach to prevention, Weissberg and Greenberg (1998, p. 483) wrote The RD&D [Research, Design, and Development] and IOM’s prevention research cycle both assume that there is a rational sequence—research, development, packaging and dissemination—for evolving and applying a new intervention. They also assume that there is a rational consumer who accepts and adopts the innovation. It is assumed that programs are effective because they incorporate the most recent findings from theory and research and that they have been through systematic development and field-testing. The fact that most of these programs are never used on a wide-spread basis illustrates the danger in not understanding more about the user end of the RD&D continuum.
Fast Track represents one of the first attempts to take a prevention science approach to the next step, by creating a science-based intervention trial that includes multiple levels (universal and indicated prevention components) and multiple stakeholders within a community
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
Fast Track Implementation (interfacing with schools, families, and community service and youth agencies). The prevention program is multifaceted, and is carried out over a 10-year time frame that is compatible with the science-based developmental model. This paper describes some of the issues faced in moving from short-term controlled trials of isolated intervention components to a long-term, comprehensive prevention approach that requires, for its success, reciprocal commitments between researchers and the host community that extend beyond the demands of short-term interventions and data collection. It provides some information about “lessons learned” in trying to merge the sensibilities of community psychologists with a university-directed prevention program and to better understand “the user end of the RD&D continuum,” and issues for future research and development. After a brief description of the developmental model that provides the basis for the design of the Fast Track program, the paper focuses on three core areas of implementation: (1) entry into school and community systems and engaging teachers and parents in the intervention; (2) intervention implementation and the processes associated with the delivery of the various components of the intervention, including the balance between maintaining fidelity while adapting flexibly to local cultural/ethnic variations; and (3) the processes associated with retaining families and youth in the long-term intervention, and building a foundation for the support of longer-term sustainability of the prevention model. The paper focuses primarily on the first 3 years of intervention, but includes some information on the sustainability of the early-grade school intervention components after the trial moved into later years. THE DEVELOPMENTAL MODEL AND CORRESPONDING PREVENTION DESIGN The Fast Track project design involves a randomized, controlled trial of a multifaceted, long-term prevention program embedded within a longitudinal study of normative and high-risk youth. At each of four sites (Seattle, WA; Durham, NC; Nashville, TN; and rural central Pennsylvania), sets of schools, matched on size, ethnic composition and poverty, were randomly assigned to intervention or control conditions. Three cohorts of families and children are participating currently in a long-term developmental study that provides ongoing comparative information about the development of children at each of these sites and about the progress and outcomes of children who are or are not receiving the prevention services. (Cohorts entered first grade in 1991, 1992, and 1993.) The prevention program involves a 10-year span of prevention activities (from Grades 1 through 10), covering
15:54
Style file version July 26, 1999
3 the important developmental transitions of entry into elementary, middle, and high school. The intervention is guided by a developmental theory positing the interaction of multiple influences on “earlystarting” antisocial behavioral development (CPPRG, 1992, 2000). In this research-based model, early childhood characteristics, such as impulsivity, irritability, and inattention (Bates, Bayles, Bennett, Ridge, & Brown, 1991; Campbell, Breaux, Ewing, & Szumowski, 1986) combine with family stressors such as poverty, single parenting, and parental conflict (Dodge, Bates, & Pettit, 1990; Offord, Alder, & Boyle, 1986; Richman, Stevenson, & Graham, 1982) to produce children who rely on aversive behaviors to get what they want. Coercive parent–child interactions, combined with inadequate parental support, contribute to delays or distortions in the child’s development of cognitive skills and adaptive emotion regulation capabilities, and also contribute to aggressive social-cognitive and behavioral tendencies (Cook, Greenberg, & Kusche, 1994; Dodge, Pettit, McClaskey, & Bown, 1986; Lochman & Dodge, 1994). These behavior problems are often exacerbated when children enter elementary school. Many highrisk children attend schools with a high density of other children who are similarly unprepared, and who are negatively influenced by disruptive classroom contexts and punitive teacher discipline practices (Rutter, Maughan, Mortimore, Ouston, & Smith, 1979). Peer rejection and learning difficulties fuel escalating aggressive exchanges and reduce interest and positive engagement in school (Dodge, Pettit, & Bates, 1994; Ladd, Price, & Hart, 1990; Moffitt, 1990). This process of escalating risk extends from the elementary school years into adolescence where deviant peer group influences and dysfunctional personal identity development contribute to serious and persistent antisocial behavior. Based upon the developmental model, one would anticipate that effective prevention programs would need to be multifaceted, addressing problems in both home and school settings, and include key socializing agents, such as parents, teachers, and peers, along with targeted high-risk children. Clearly the dysfunctional development that is associated with the early-starting pattern of conduct problems is multiply determined and is embedded in transactions among family, peer, school, and neighborhood influences and child characteristics. Hence, prevention efforts should target both the promotion of individual competencies and the promotion of protective contextual supports. In keeping with the extended developmental framework implicated by research findings, the Fast Track program includes two phases of prevention activities. The focus of this paper is on the first phase of the intervention, which targeted the risks associated with the early
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
Style file version July 26, 1999
4
The Conduct Problems Prevention Research Group
onset and display of behavior problems during the initial school years. In elementary school, the Fast Track program included three levels of intervention: (1) universal intervention, conducted at the school level; (2) standard indicated interventions for children identified as at risk for future conduct problems during a kindergarten screening; and (3) individualized indicated interventions to address heterogeneity in risk factors and severity among the highrisk group. At the universal level, the intervention included a classroom curriculum that was taught by teachers at intervention schools (see Bierman, Greenberg, & CPPRG, 1996). An adapted version of the PATHS (Promoting Alternative Thinking Strategies ) Curriculum, originally developed by Kusche and Greenberg (1994) was used, which focused on skills in four domains related to school success: (a) prosocial behavior and friendship skills, (b) emotional understanding and self-control skills, (c) communication and conflict resolution skills, and (d) problem-solving skills. This curriculum included a developmental progression of skills across Grades 1–5. At the end of first grade, our findings indicated significant effects reducing peer ratings of aggression and hyperactive-disruptive behavior and improving observer ratings of classroom atmosphere (CPPRG, 1999b). For the indicated level of intervention, high-risk children were identified in kindergarten on the basis of teacher- and parent-reports of child-conduct problems at home and school. Intervention began in the first grade with high-risk children, their adult caregivers, and their teachers. The elementary school phase of the prevention program addressed six areas of risk and protective factors derived from the developmental model: parenting, child social problem-solving and emotional coping skills, peer relations, classroom atmosphere and curriculum,
academic achievement, and home-school relations. Indicated interventions included parent training, home visiting, parent–child relationship enhancement, social-skills training, peer-pairing social skills support, and academic tutoring (Bierman et al., 1996; CPPRG, 1992; McMahon, Slough, & CPPRG, 1996). All of the identified high-risk children received all of the indicated interventions in Grade 1, and the indicated parent groups, child social-skill training groups, and parent–child relationship enhancement sessions through the elementary years. In Grades 2 and beyond, however, the level of home visiting and the provision of peer-pairing and academic tutoring depended upon individualized assessments. Children with poor adaptation in any of these domains (family functioning, peer relations, reading and academic performance) received additional intervention support. For example, children who were rejected by peers qualified for peer-pairing; children whose reading performance placed them in the lower third of their class qualified for tutoring; and frequency of home visiting (weekly, biweekly, or monthly) depended upon staff ratings of child and family functioning. Parent groups, child social-skill training groups, and parent–child relationship enhancement sessions were provided weekly in Grade 1, biweekly in Grade 2, and monthly in Grades 3–6 (see intervention organization in Table I). Increased child competencies and reduced child behavior problems have been documented at the end of first and third grades (CPPRG, 1999a, 2002). The core staffing for the intervention involved two professional positions and one paraprofessional position. Educational Coordinators (ECs), who had education, special education, or counselor education backgrounds and teaching experience, had three core job responsibilities: (1) providing training and consultation to teachers and other school personnel in the implementation of the
Table I. Fast Track Intervention Components in Grades 1 Through 3 Component Universal components PATHS curriculum Indicated components Child social-skills groups Parent training groups Parent–child sharing Individualized components Academic tutoring Peer-pairing Home visiting
Grade 1
Grade 2
Grade 3
Yearly curriculum
Yearly curriculum
Yearly curriculum
22 sessions 22 sessions 22 sessions
14 sessions 14 sessions 14 sessions
9 sessions 9 sessions 9 sessions
60 sessions 22 sessions 11 visits
0 or 60 sessions 0 or 14 sessions 8, 16, or 32 visits
0 or 60 sessions 8, 16, or 32 visits
Note. In second grade, children received peer-pairing only if they were rejected by peers or had elevated teacher ratings of aggression; in second and third grades, children received academic tutoring only if their reading skills placed them in the bottom-third of the classroom; families received weekly, biweekly, or monthly home visits depending upon staff ratings of family functioning.
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
Style file version July 26, 1999
Fast Track Implementation PATHS Curriculum and classroom behavior management, (2) leading the child social skills training groups and coleading the parent–child interaction sessions, and (3) training and supervising paraprofessional tutors. Paraprofessional tutors, who received 40 hr of training and ongoing supervision, conducted the tutoring and peer-pairing sessions with the target children, under the guidance of the ECs. Family Coordinators (FCs) had social work or counseling psychology degrees, and/or extensive work experiences in human services that involved relating to and working with high-risk families comparable to our participants. These staff members led the parent group and parent–child interaction sessions, and conducted home visits. They also took responsibility for general case management and referral for other services, when needed. ECs and FCs worked in pairs, serving an average of 15–18 families in three groups. The implementation of the multifaceted Fast Track program required the partnership of schools (administrators and teachers) and parents and youth. Of particular importance was the involvement of teachers who would implement the classroom-based PATHS program, and the engagement of the families and youth designated “highrisk” by the population-based screen.
ENTRY PROCESSES: PROMOTING THE ENGAGEMENT OF SCHOOLS AND FAMILIES One of the major concerns raised about the prevention science approach to community-based prevention programs is the degree to which a university-developed and research-directed program can effectively activate community members’ feelings of program “ownership” and concomitant commitment and participation (Elias & Clabby, 1992). Fast Track requires the active involvement of community members. To be effective, school personnel and parents must become committed to the program and willing to invest in the process of making the changes supported by the program. The challenge was substantial for two reasons: (1) in the school setting, the universal program was building-based and could run effectively only with the active commitment of the participating teachers (who were responsible for implementing the program) in each elementary school building; and (2) in the community setting, the program was designed to attract and involve the most high-risk families (who usually do not volunteer for school programs). Although one approach to establishing school or community “buy in” is to involve participants in the selection and design of the prevention program, as noted earlier, this approach would have required a large step away
5 from the guiding premises of the prevention science approach. That is, based upon the extensive research base on critical risk and protective factors for conduct disorder, we felt fairly confident that a large-scale effort to prevent this disorder could be effective only if it addressed all of these factors. Hence, the specific goals of the prevention trial were not open to negotiation, and recruitment processes had to be designed to encourage collaboration in other ways. In the following sections, we describe the specific approaches used to recruit and form effective partnerships, first with participating schools and then with participating families.
Establishing Collaborative Partnerships With Schools The first step in entering schools involved acquiring the approval and initial support of school administrators (superintendents, principals, school-board members). This proved to be relatively easy. Presented with an opportunity to provide a set of academic and social support services to at-risk students and their parents at no cost to the school district, almost all administrators that we contacted were interested in hearing more about the program. There was particular enthusiasm about the acquisition of parent training and home-visiting services for parents of at-risk students, because most administrators felt that they lacked the resources and expertise to offer these services themselves. The focus on promoting competencies as a prevention strategy at school entry also interested administrators, because their available funding was often limited to serving children with identified problems. Administrators were required to agree to the randomization process, recognizing that a particular school would be assigned to either intervention or control status, but were assured that, within each district, approximately half of the participating schools would receive the intervention. In general, the administrators felt that teachers would benefit from additional training and consultation in working with high-risk students with conduct problems, and supported the provision of the universal prevention curriculum (PATHS). Hence, administrative support was easily acquired; only one of the eight school districts contacted about the project was not interested in meeting with the investigators, and all of the district administrators who met with investigators agreed to participate in the project. The second, critical level of support needed was the active engagement of the teachers who would serve as providers of the classroom-based intervention (the PATHS Curriculum), and “gate-keepers” of the school-based
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
6 indicated interventions (the pull-out peer-pairing and academic tutoring programs). Being on the “front lines,” teachers had a number of practical concerns about the prevention program, including concerns about the time needed for the intervention, the overlap with other prevention curricula in the school, and the usefulness of “pull-out” programs that took children out of the classroom. Some wondered about the utility of an approach that emphasized skill training and positive contingency management programs rather than more punishment-oriented discipline strategies for students with aggressive–disruptive behavior problems. In some cases, teacher concerns reflected a history of poor administrator–teacher relations, and a concern that new programs were added routinely without a true understanding of the complexities of running consistent and coherent educational programs in the classroom. The process of gaining positive engagement from teachers involved several steps. Prior to the initiation of the program, informational meetings were held at each identified elementary school in each district to allow teachers to hear about the program and to decide (on a building level) whether they wanted to participate in the program. Schools varied in the particular process they used to decide whether they would participate, some using more formal voting procedures, and some using more informal discussions to reach consensus. Two schools in Seattle declined to participate at this point: one had just begun an extensive curriculum initiative and felt overwhelmed, and another had a temporary principal who felt that this type of decision should be made by a permanent principal. All of the other 54 invited schools agreed to participate. One key attraction for the participating schools was the fact that program costs were covered by the program grant, providing schools with a wide array of “free” services for their high-risk students. (As discussed later, however, schools who wanted to maintain these services after the grant-supported period had to find alternative sources of financial support.) The second step of engaging teachers involved a 2-day introductory workshop, held at each of the Fast Track sites. Teachers were paid according to their inservice pay scale for attendance at this workshop, or they received CE credit (whichever they chose) to elicit their participation. During this workshop, the principles underlying the design of the PATHS Curriculum were reviewed, and teachers were provided with samples of different lessons. Discussion sessions allowed teachers to react to and comment on PATHS, and to raise concerns or issues regarding the implementation of Fast Track in their schools and communities. This workshop played an
Style file version July 26, 1999
The Conduct Problems Prevention Research Group important role in providing teachers with an understanding of the developmental rationale underlying PATHS and Fast Track, and also offered the opportunity to initiate personal relationships with the Fast Track program staff. The discussions alerted staff to issues that needed to be discussed further in individual or small group meetings with teachers. The third step in engaging teachers was a prolonged one, and involved the building of personal relationships between teachers and Fast Track ECs. The fact that ECs had teaching experience themselves provided Fast Track with some “face validity” among the participating teachers and maximized the likelihood that the Fast Track staff would understand the perspectives and concerns raised by teachers. Gaining the support of individual teachers required an investment of staff time, including multiple visits to the school and classroom, designed to help staff gain a full understanding of the context and challenges the teachers faced. Issues such as curriculum overlap and integration, the scheduling of pull-out sessions, and providing consultation on classroom behavior management issues were more easily negotiated when Fast Track staff were able to spend the time and become accustomed to the routines and structure of the various classrooms and schools. Staff made themselves available to meet with teachers according to the teacher’s schedule and preferences. During these meetings, program staff were instructed to listen carefully to the concerns raised by teachers, provide information about the rationale for various program design features, and to remain flexible, taking a joint problemsolving orientation with teachers and balancing the goals of preserving the integrity of the field trial with the preferences or concerns of various teachers. In addition, ECs observed a PATHS lesson in the classroom approximately three times per month during the teacher’s first year of implementation, providing a mechanism for additional communication and monitoring of program fidelity. As discussed in the next section, flexible adaptation of some program components was needed in order to create a better fit between the program and the various school and classroom settings. Overall, these strategies were effective in gaining the “buy-in” of the majority of teachers. Program implementation data show that, in first grade, the mean number of PATHS lessons taught by Grade 1 teachers was 48.2 (SD = 9.7, range = 13–57) out of a possible total of 57 lessons. In summary, keys to successful entry in the schools included the combination of: (a) informational meetings, (b) opportunities to voice opinions, (c) establishment of personal relationships between program staff (who were
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
Fast Track Implementation experienced teachers) and teachers, (d) program staff willingness to spend time in the schools and understand the challenges of program implementation from the perspective of the teachers, and (e) some flexibility around that implementation (an issue discussed further in a later section). Establishing Collaborative Partnerships With Families A second critical entry challenge involved the successful recruitment and active involvement of at-risk families into the program (McMahon, Slough, & CPPRG, 2001). For the prevention program to be successful at the community level, it was important to recruit a large majority of the high-risk families in the school system into the prevention program. Yet, a number of the family factors that contribute to elevated rates of child-conduct problems (e.g., low socioeconomic status, maternal depression, marital discord and instability, insularity, and single-parent status) also reduce the likelihood that at-risk families will volunteer for extracurricular school programs or that they will seek out community support programs (Dumas & Wahler, 1983; Harnish, Dodge, Valente, & CPPRG, 1995; Offord, Boyle, & Racine, 1991; Rutter & Giller, 1983). Parents of children who are at risk for conduct problems tend to face multiple ongoing life stressors that are disorganizing and create immanent life pressures, undermining long-term problem-solving approaches to childrearing issues. In addition, many of these families had previous experiences with schools and other service agencies that were conflict-ridden and left some families feeling that these agencies were judgmental, culturally or ecologically insensitive, and nonsupportive (Orrell-Valente, Pinderhughes, Valente, Laird, & CPPRG, 1999). Thus, Fast Track was also likely to be viewed initially in these same ways. The program directors and school representatives both felt that it was unlikely that announcements and advertisements would provide an adequate means of recruiting the families most in need of the intervention. To increase the representativeness of the high-risk family participants, for the purpose of both prevention effectiveness and maintaining the integrity of the research design, a population-based sampling technique was employed to identify target families. A multistep process was developed for recruitment within this sampling design. The multistage, population-based screening procedure used to identify children at risk for future conduct problems is described in detail elsewhere (Lochman & CPPRG, 1995). Basically, teachers completed ratings on
15:54
Style file version July 26, 1999
7 all children attending kindergartens in the participating schools during January and February (N = 9,594), making it possible to identify children who were struggling to adjust behaviorally to the demands of school. The parents of children who scored in the top 38% (N = 3,600) with regard to school adjustment problems were then identified for contact, along with a representative sample of parents. These parents were invited to participate in a brief (15-min) school-sponsored telephone survey on children’s adjustment to kindergarten. (Parents who failed to return the permission form for this survey, or who did not have a telephone, were visited at their homes.) Parents were compensated financially ($15) for their participation in this survey. With some variation across sites, 91% of the parents contacted agreed to participate in this survey. During this survey, parents were asked to complete ratings describing their child’s behavior problems at home during the kindergarten year. Kindergarten teacher- and parentreports of child behavior problems (at school and at home, respectively) were used as a combined screen to identify children for the high-risk sample (i.e., children scoring in the top 10% in terms of combined cross-setting assessment of behavior problems by teachers and parents). A second contact was made with these identified families during the summer between the child’s kindergarten and first-grade year, in order to recruit them into the longitudinal study associated with the project. Families were asked to participate in an in-depth developmental study of children’s adjustment to school, which would involve annual home interviews and tracking of the child’s school adjustment. The study involved a 2–3-hr summer home interview with parent and child, and follow-up telephone interviews. Parents were compensated financially for their participation ($75). Families who agreed and completed the home interview became participants in the Fast Track longitudinal study. Depending upon the school where their child entered first grade, they became participants in either the intervention or the comparison group for the purpose of the research trial. Fast Track thus used an “intent to intervene” model, in which participants are identified for the intervention prior to their recruitment for intervention. This model guarantees greater comparability between families in the intervention and comparison group than a model in which only help-seeking volunteers are included in the intervention group. However, it increases the challenge to recruitment, as failures to recruit eligible families for prevention services result in weakened effects in the trial. That is, eligible families who are not successfully recruited into the intervention remain in the intervention group for analyses, despite a lack of participation in the
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
8 intervention.2 This design feature motivates staff to continue to encourage participation from families who are initially hesitant or hard to reach. It also provides a more accurate estimate of the impact of the prevention trial on the community, as failure to recruit high-risk families into ongoing community and school prevention programs weakens the likelihood that those programs will affect population-level incidence or prevalence of problem behaviors. Ninety-two percent of eligible families whose child actually enrolled in first grade at a core school agreed to participate in the longitudinal study. Attempts to recruit families into the Fast Track intervention were initiated at the end of the summer or during the 1st month of the child’s first-grade year for those children who were in schools that had been randomly assigned to intervention. At that point in time, parents had experienced two previous personal contacts with project staff (i.e., the brief telephone survey in the spring and the longer home visit interview in the summer). Thus, project staff were no longer strangers. During a home visit, an FC described the prevention program to parents and invited them to attend a program orientation at the school. During a group “dessert party” orientation in the afternoon or evening, families were invited to their local school. Child care and transportation to the meeting were provided. The Fast Track staff described the program components, and invited discussion and comments from the parents. Parents were invited to participate in all aspects of the program, but were also given the option of participating in selected components of intervention. In addition to using a gradual process with repeated personal visits and invitations to establish positive relationships with parents, several principles guided our recruitment efforts. From the beginning, we felt that the effective engagement of family participants would be influenced heavily by the success with which staff could establish positive working relationships with parents. Therapy process research suggests that the quality of the therapist– client working alliance is a key predictor of client par2 One
consequence of the design feature was that, at the point of participant recruitment into the developmental study, participants were not given information about the intervention and control comparisons planned for the future. Rather, they were told that the purpose of the developmental study was to examine child developmental progress and adjustment at home and school, and how family and school experiences contributed to developmental progress. Later, the families of children attending intervention schools were offered the prevention support services at home and school, but participation in these services was not tied specifically to the developmental study. Some families offered intervention did not participate in the intervention, but continued in the developmental study. Conversely, intervention was not withdrawn if families chose not to participate in the developmental study in later years.
Style file version July 26, 1999
The Conduct Problems Prevention Research Group ticipation in therapy (e.g., Horvath, 1994). In addition, Patterson, Chamberlain, and colleagues have found that the therapist–family relationship is a critical determinant of the rate, and possibly also the quality, of parent participation (e.g., Chamberlain, Patterson, Reid, Kavanagh, & Forgatch, 1984; Patterson & Chamberlain, 1994). To strengthen the likelihood that staff could develop positive working relationships with high-risk families, an attempt was made to recruit professional staff with similar cultural and ethnic backgrounds as the families (Orrell-Valente et al., 1999). We were able to recruit clinical staff (including clinical supervisors) who were representative of the ethnic composition of the families, as a function both of sustained efforts and flexibility in recruiting staff with a range of different types of relevant professional training and experience. Efforts were also made to recruit support staff (particularly the paraprofessional tutoring staff) from the neighborhoods and schools being served by the program, in order to increase the local grounding of the program and to increase community ownership. A second guiding principle in the recruitment process involved the program’s promotion of family and child competency rather than focusing on risk. In a number of cases, families being recruited into the program had already experienced conflict with schools or other social service agencies because of their children’s misbehavior. Some were wary of pejorative labeling or a deficit-oriented perception of themselves or their children. In addition, given the young age of the children involved, it was more appropriate from a developmental standpoint to focus on the promotion of skills as a central goal of the program than the reduction of problems. Almost universally, we found that parents in the project hoped their young children would be successful in first grade, and would learn to read. Almost all of them were interested in additional tutoring that might help their children attain this goal. Focusing on the opportunity to participate in an enrichment program that could offer their children a “boost” in terms of school skills, and offer the parents support and ideas about positive parenting experiences was viewed as a much more effective recruitment strategy than trying to engage parents by warning them about the possible long-term consequences of their child’s early problem behaviors and the need for them to take remedial action. In addition to the presentation by our staff, we also found it important to monitor how the program was portrayed by school personnel and in the community in order to maintain a focus on positive skill promotion rather than a deficiencies (i.e., a program for problem parents or children.). The fact that the prevention program included components for all students (i.e., PATHS, peerpairing), in addition to program components for high-risk
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
Style file version July 26, 1999
Fast Track Implementation students, helped focus community members on the positive orientation of the program to building school-wide competencies. Third, the staff were instructed to maintain a flexible orientation toward working with families (McMahon et al., 1996). One key adage describing their orientation was “take the families where they are,” reflecting the flexible willingness of staff to adapt to parental choice in the selection of intervention components and to let families “ease into” the prevention program on their own terms. A second key adage was “never give up,” encouraging staff to see familial participation in the intervention as a longterm goal rather than a short-term accomplishment. Staff were instructed to put the highest priority on establishing positive long-term relationships with families (rather than exerting pressure to attain a short-term goal of parental attendance at sessions), with the understanding that family priorities, stressors, and motivations change over time, and that through an ongoing positive relationship, staff were most likely to be in the position to offer help when it was sought out by the family. Fourth, but certainly not least, a critical factor in increasing the engagement of parents involved the reduction of pragmatic obstacles to their participation, and the provision of incentives for involvement (McMahon et al., 1996). Specifically, sessions were held at local school buildings, to increase accessibility. Transportation was provided, as was on-site child care for siblings (e.g., in the school gymnasium). A welcoming atmosphere and refreshments were offered at each session. In addition, financial compensation was offered to parents who attended the parent training sessions. Parents were treated as colleagues (rather than clients), were given the message “You are your child’s most important teacher,” and parallel to the collaborating teachers, parents were paid for the time they spent in program training sessions ($15 for each 2-hr group program session that they attended). In summary, the strategy for the recruitment of families and the establishment of positive collaborative working relationships began with a population-based sampling strategy to identify at-risk children. A gradual process of sequential contacts with parents provided them the opportunity to become familiar and comfortable with program staff, and provided a foundation for communication and building positive relationships. These sequential contacts began with a low-risk/low-demand telephone contact, followed by an in-person, moderate-demand interview contact, then by an in-person recruitment visit, and finally, an invitation to a group orientation. In addition, several principles guided staff selection and staff communications about the program. Staff were selected with similar ethnic backgrounds as the families served, and for their ability to
9 Table II. Parent and Child Group Participation in Grades 1 Through 3
% Attended
% Sessions attended
% Attending at least 50% of sessions
Grade
Parent
Child
Parent
Child
Parent
Child
1 2 3
96 88 80
98 92 86
71 72 72
78 79 77
79 79 78
90 87 84
connect well with the families served. Staff focused on the competency-building opportunities provided by the prevention program, rather than focusing on child or family deficits. Finally, families were provided with pragmatic support (transportation, childcare) and incentives for involvement (friendly atmosphere, refreshments, financial compensation for training sessions). These recruitment strategies appeared effective in eliciting parent and child involvement in the intervention (see Table II). In Grade 1, nearly all of the children (98%) attended at least one of the social skills group sessions, and all but a few parents (96%) attended at least one parent group. Of those who attended any sessions in the first grade, parents attended an average of 71% and children attended an average of 78% of the sessions. We think it is very unlikely that we could have generated this level of parent participation and representativeness had we not employed our extensive set of recruitment procedures and incentive structures. Similar to the teachers, parents did vary in their interest in, and responses to, various components and aspects of the intervention. The next section describes some of the issues and tensions that emerged when balancing flexible responsivity to teacher and parent preferences with the goal of maintaining the core fidelity of the preventive intervention. INTERVENTION IMPLEMENTATION: FLEXIBILITY WITHOUT LOSS OF FIDELITY The Fast Track prevention program was designed to be sensitive to the wide range of individual preferences (including preferences associated with gender and ethnic differences) that existed both across the multiple sites with diverse populations and within sites, reflecting the specific needs of individual children and families. To maintain the fidelity of the prevention program, it was important to maintain a central focus on the protective and risk factors identified in developmental research, and to employ intervention strategies that had proven effective in previous clinical trials. Yet, at the same time, flexibility was needed to adapt the intervention in order to engage the
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
10 heterogeneous participants who represented a range of demographic characteristics and cultural backgrounds. In general, we focused on maintaining similarity across sites and groups in the principles of intervention, but allowing the process and implementation strategies to vary within these limits. Examples of adaptations made to various aspects of the intervention follow. School-Based Program The PATHS Curriculum provided teachers with comprehensive, step-by-step lesson plans, including example dialogue and practice activities. Teachers were asked to teach 30-min lessons 2–3 times per week. However, the effectiveness of the program rested on the extent to which teachers accepted the basic tenets of the program and integrated the concepts into their teaching style, rather than solely on their strict adherence to the dialogue of the lesson plans. We anticipated that a teacher who read every lesson from the manual verbatim, but failed to incorporate the communication and problem-solving skills into his/her everyday interactions with the children, would be less effective than the teacher who veered from the literal script, but adapted lessons to incorporate the core concepts effectively into his/her teaching activities and style. Indeed, outcome data at the end of first grade demonstrated that the number of PATHS lessons taught was not directly correlated with outcomes, whereas the extent to which a teacher used the core concepts in regulating his/her classroom did predict positive child outcomes (CPPRG, 1999b). Teachers requested a number of alterations to the “standard” presentation of PATHS lessons. Many of these alterations were adjustments in the style, pacing, or wording used, and as such, posed minimal threats to program fidelity. Examples included: adjustments in the number and length of sessions taught, integration of selected lessons with similar lessons from other prevention curricula, changes in the wording of lesson presentation or the examples used, and changes in the order of lesson units. Teachers met regularly, individually or in small groups, with ECs to discuss their implementation of the program, and ECs spent time in the classroom each week, either modeling lessons or observing. Over time it became clear that coteaching lessons was often a more effective way to build a strong working relationship with a teacher, rather than observing or modeling. Teachers usually are not accustomed to being observed and evaluation issues can create resistance in this circumstance. On the other hand, some teachers would use the ECs solo presentation as an opportunity to take a break, and often did not attend to this presentation as closely as when coteaching was employed. In general, the flexibility
Style file version July 26, 1999
The Conduct Problems Prevention Research Group in program implementation allowed teachers to adapt the program to fit their circumstances and to make the program their own; the partnership of the EC provided support as well as a mechanism to assure that program adaptations did not undermine the fidelity of the core concepts of the program. Examples of Adaptations to the Child Social Skills Training Groups Similar to the PATHS Curriculum, the Friendship Group Curriculum given to ECs to lead the child socialskill training groups was detailed and comprehensive. For each lesson, the goals, objectives, and methods were provided, along with sample dialogue, activity plans and materials (including modeling videotapes and stories, and practice games and exercises). However, it soon became clear that groups varied in their responsivity to various presentation materials. For example, some groups responded well to videotape presentations, some groups enjoyed role play exercises, some groups were responsive in open problem-solving discussions, whereas other groups were not well-engaged by these activities. Hence, over time, a “menu” organization was developed for the Friendship Group curricula. The goals and objectives of a lesson were specified, along with a menu of interchangeable skill presentation and practice activities. This allowed group leaders to tailor the activities used in each session to the interests of the children in their group. For example, a skill concept might be presented by video in one group, through a modeling story in another group, and by a coach demonstration in a third group. Across all groups, the same skill concept was presented, but the mode of presentation varied, with the goal being the fullest possible engagement of group members. Clinical supervisors at each site held weekly meetings with ECs to discuss the focus of the skill training curriculum, and made periodic observations of group sessions. This ongoing supervisory process was critical to maintain the cross-group fidelity of the conceptual core of the skill training curriculum, and to assure that adaptations of the activity menu remained “true” to the underlying concepts and goals of the program. A similar level of flexibility was developed in the parent training and parent–child sharing group curricula, as well as in the home-visiting procedures. Examples of Adaptations in the Parent Program Parallel to the child group intervention manuals, the parent group and parent–child sharing sessions were described in detail in staff manuals. Also parallel to the child group sessions, it became evident during the first 2 years
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
Style file version July 26, 1999
Fast Track Implementation of the project that parent groups differed in their interest in and responsivity to different forms of skill concept presentation and practice. Videotapes, role play practices, and open discussions were more or less successful across the different groups. Some effects seemed site-based. For example, some of the parents in the primarily African American sample at the Durham site found it difficult to relate to some of the language and scenarios in the videotapes (even though they included ethnically-diverse actors). The primarily European American sample at the rural Pennsylvania site found open discussions difficult, and felt uncomfortable talking freely about their personal problems. Other group preferences appeared to be a function of the process dynamics that emerged in particular groups; for example, the degree to which a group enjoyed participant role plays. In addition, “special interest” topics emerged in some groups. For example, some groups consisted primarily of single parents who had a particular set of shared issues; other groups included a large proportion of members who had current or past problems with alcohol or other substance use. Group preferences and group needs thus varied somewhat across groups. To adjust for these differences, parent group leaders were encouraged to present identified concepts in several different ways (for example, with a videotape model, a leader demonstration, or a parent participant role play). In Grade 2, menu options were added to include alternate activities for the parent–child sharing session. By Grade 3, two (of the nine) sessions of the parent group curriculum were “open” sessions, which allowed the group leader to revisit or introduce concepts and skills of particular relevance to certain groups. These sessions also gave parents the opportunity to ask for information or focus discussion on parenting topics of particular interest to them. Previous research suggests that group composition and dynamics may affect how involved and interested parents are in group sessions (e.g., Shen, Sanchesz, & Huang, 1984). Hence, after the first year, FCs (and ECs) shuffled some of the group compositions, in order to maintain or increase a high level of engagement. The driving motivation in these adaptations was to increase parental interest and engagement in the parent groups, and their incorporation of the central parenting skills in their everyday interactions with their children. Allowing some adaptations in the curriculum (the intervention given) was viewed as a method of enhancing the fidelity of the intervention received. Like the ECs, the FCs met with a clinical supervisor weekly, and had weekly staff meetings to discuss the parent groups. The supervisory process was critical in maintaining the conceptual fidelity of the program. One critical challenge for staff was to distinguish flexibility in adapting to local culture and group norms from flexibility that
11 served to support parental resistance to change and undermined movement toward improved parenting. Patterson and others have shown that, in samples similar to the Fast Track sample, three out of four parents of children with conduct problems exhibit significant resistance to the intervention (e.g., Chamberlain et al., 1984; Patterson & Chamberlain, 1988). These investigators have noted that parent resistance to intervention tends to increase during critical phases of the intervention, when therapists are focusing on changing dysfunctional parenting practices (Patterson & Forgatch, 1985). Flexibility that takes the intervention staff away from a focus on changing these parenting practices does not serve the program well, but rather allows parents to disengage or divert their efforts away from the difficult challenge of making changes. Hence, flexible adjustments of program content must be made with care, recognizing that effective intervention may not always be as comfortable or fun as intervention sessions that focus on less difficult subjects. Adaptations were also made to engage the “hard to reach” families. For example, some parents were unwilling or unable to attend parent groups. When parents missed a group session, FCs made efforts to visit them in their home to review the materials. When parents were uncomfortable meeting staff in their home, alternative meeting places were selected (e.g., at a coffee shop over donuts or lunch). EC and FC work schedules included evening and weekend hours, to allow flexibility in meeting parents and scheduling groups. Maintaining Cross-Site Fidelity Several strategies were employed to maintain fidelity of intervention implementation across sites. One strategy involved the detailed and comprehensive intervention manuals that were prepared for each of the core intervention components. In the spring of the year prior to the initiation of intervention, each site conducted a short series of pilot sessions, implementing a 4-week program that included all intervention components with a group of school-referred first-grade children at one of the schools at each site. During each of the first 2 years of program implementation, cross-site staff training was conducted. ECs and FCs, clinical supervisors, and principal investigators from all four sites met for 2.5 days each year to review each of the programs in detail. During these same years, Mark Greenberg traveled to each of the sites to train teachers in the PATHS Curriculum. In this way, initial training was comparable at each of the sites. Weekly cross-site conference calls (monthly in Grade 3 and beyond) were conducted to discuss the implementation of each of the core intervention components (PATHS, child social-skill
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
12 training groups, academic tutoring, and parent training groups and home visiting). These calls provided a mechanism for clinical supervisors across sites to discuss intervention issues, and provided an opportunity for the crosssite input and feedback from the principal investigators. Family and school engagement in the intervention was viewed as critical, both to foster short-term growth in parent and child competencies (in Grades 1–3) but also to establish longer-term relationships that would lead to: (1) the retention of participating families in future intervention (Grades 4–10), and (2) the sustainability of the school-entry prevention model (Grades 1–3) as the trial moved forward to Grade 4 and beyond.
RETAINING FAMILIES AND BUILDING A FOUNDATION FOR SUSTAINABILITY As a long-term, comprehensive prevention effort, Fast Track represents a trial that will test the extent to which therapeutic relationships with high-risk families can be sustained over the difficult course through elementary school and into the risky early adolescent years. It is also a test of whether schools and communities will become engaged and interested in sustaining prevention efforts when these efforts are driven by prevention science and initiated by university researchers. The long-term success of Fast Track in achieving these goals awaits future evaluation; however, the following sections report some relevant issues and experiences to date. Retaining Families With respect to the retention of children and parents in the intervention, the overarching principle throughout the elementary-school phase of the intervention has been to “keep them involved at all costs” (McMahon et al., 2001). We will focus on three issues that are central to the retention of families in the Fast Track intervention: (a) the long-term nature of the intervention; (b) the transience and mobility of families; and (c) family reactivity to changes in intervention structure and content. Long-Term Nature of the Intervention As noted above, the intervention began in Grade 1 and is scheduled to go through Grade 10. This longevity presents both advantages and disadvantages. On the positive side, it provides staff with the luxury of time necessary to develop a relationship with a resistant parent, child, or family. In many cases, it took nearly a year for an FC to
Style file version July 26, 1999
The Conduct Problems Prevention Research Group develop a trusting relationship with a Fast Track participant. The fact that the FC conducted both the parent group and the home visits was also beneficial, in that as the parent got comfortable with the FC through the home visits, the parent was also more likely to participate in the parent group run by that same FC. (In other situations, the reverse occurred—the parent might attend the parent group, but be unwilling to let the FC come to her home. However as the parent became more comfortable with the FC through the parent group, the parent became more open to the possibility of allowing home visits.) Another positive aspect of the long-term nature of the intervention was the increased opportunity for social support to develop among the group members. In some cases, relatively high levels of such support occurred within the first year, but in other groups, social support among the group members developed more slowly over the course of several years. There have been multiple instances of parents offering to provide childcare or transportation for another parent, and assisting parents with various personal and family issues. Parents and children remained in the same intervention groups across time whenever possible, to enhance the development of group support. In some cases, family moves or problematic group dynamics led to group reassignments. Whenever possible, in such cases, staff overlapped their work with families for a few contacts to facilitate smooth transitions. Staff turnover has remained fairly low, fostering the program’s capacity to develop lasting relationships with parents and children. There are also several negative outcomes that may occur when conducting an intervention of such long duration. Some families may become overly dependent upon Fast Track, and have unrealistic expectations as to what the program can or should provide. At its most extreme, it can be described as the “What Have You Done for Me Lately?” syndrome. For some staff, there is an increased possibility that they may develop too strong an attachment either to individual children or families, or to their caseload in general. At its most extreme, the staff member may become enmeshed in family dynamics. Finally, both families and staff may suffer from fatigue and burnout as a function of formal contact over such a long period. There are few models for the clinical supervision of staff who retain cases over such an extensive time period. Staff burnout can also lead to increased staff turnover. This presents another set of issues around reestablishing a relationship between the new staff member and the family. To help guard against these potential negative effects, Fast Track maintained a consistent emphasis on building family self-sufficiency and empowering parents to problem solve and take action to cope with their life challenges and be positive advocates for their children (Dunst, Trivette, &
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
Fast Track Implementation Deal, 1989; Wasik, Bryant, & Lyons, 1990). The provision of group supervision allowed staff to share successes and frustrations, providing an avenue to both support staff and brainstorm around challenging cases. Individual supervision provided a context for careful case review, increasing the fidelity of intervention as well as providing an opportunity to monitor the staff–family relationship and staff morale. Transience/Mobility of Families One of the characteristics of high-risk families is a high rate of transience (Offord et al., 1991). This can wreak havoc on an extended intervention program, because it means that the family moves away from the locale in which service is being provided before the intervention has been completed. Interestingly enough, one of the advantages of implementing a long-term intervention such as Fast Track is that, although children do move out of original core schools and families may move out of their original neighborhoods, over time, many of those same children and families move back into the service area. In these latter cases, we are able to restore service to previous levels for many of these families. The question arises as to what extent this dispersion has affected the delivery of services in Fast Track? By the end of Grade 4, 88.5% of the families were living in local service areas, and were thus able to receive all Fast Track services. The remaining families lived outside of the area in which all services could be delivered. However, half of those families (i.e., 5.9% of the sample) lived close enough so that some services were still be made available. For example, a child might switch out of a core school into a neighboring school district. Although that child would not receive PATHS (which was administered only in the core schools), other school- and family-based services would still be available. If a family moved even farther away, then the only services that might be provided might be tutoring and home visiting. The remaining 5.6% of the sample lived too far out of the service area to receive any of the intervention components. Thus, at the end of Grade 4, more than 94% of the sample was available to receive at least some Fast Track services. Reactivity to Changes in Intervention As noted above, the Fast Track intervention was derived from a developmental model of conduct problems. As such, there were a number of planned changes in the intervention with respect to structure, intensity, content,
15:54
Style file version July 26, 1999
13 and staffing over the course of the intervention. The most salient change during the first 3 years of the intervention had to do with the frequency of the parent and child group sessions. In Grade 1, these sessions occurred on a weekly basis during the school year (from October to April). Thus, in Grade 1 there were 22 sessions (with each parent group and child group lasting 60 min, followed by a 30-min parent–child sharing session and a 30-min reading tutoring session). In Grade 2, sessions were biweekly, with a total of 14 across the school year (from September to April). (Group sessions were expanded to 90-min in Grade 2 and beyond, so that actual hours of parent group/child group/parent–child sharing contact were 33 and 28 hr, respectively, for Grades 1 and 2.) In Grade 3 (and beyond), the groups were scheduled monthly, with 9 groups scheduled over the course of the school year (from September to May). We anticipated that many high-risk children and families would respond well to the intensive prevention services provided in Grades 1 and 2, and be ready for a less intensive set of “maintenance and support” sessions by Grade 3. At the same time, we expected variability in intervention response and in the life-course trajectories of the families that contributed to the risk status of their children. The provision of individualized services (e.g., home visiting and tutoring) allowed us to continue higher levels of prevention services to children and families with greater needs. In general, parents and children reacted minimally to the shift from weekly to biweekly sessions that occurred between Grades 1 and 2. Indeed, as the children began to participate in various structured activities (e.g., sports, music), the decreased frequency of the group sessions appeared to facilitate their participation. This was not the case for the shift to monthly groups that occurred after Grade 2. Although some families were experiencing marked improvements and welcomed a reduction in group time, other families reported that they felt that the meetings were not being held frequently enough. Staff also noted that it was more difficult to sustain child and parent interest from one session to the next, or to maintain “momentum” in terms of facilitating change. A number of other significant changes in the intervention occurred subsequent to Grade 3. These included the introduction of a mentoring component in Grade 4, a focus on the transition to middle or junior high school in Grades 5–7, and the shift to the adolescent phase of the intervention in Grade 7. Finally, there are a host of termination issues that continue to arise as we approach the end of the intervention itself. Because they occur after Grade 3, they will not be discussed in this paper. (See McMahon et al., 2001, for a discussion of reactivity to these aspects of the Fast Track intervention.)
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
14 Participation in the Parent and Child Groups As noted by McMahon et al. (2001), there were high levels of participation in both the parent group and the children’s group over the first several years of the intervention. Table II presents data concerning attendance from Grades 1 through 3. As noted earlier, nearly all of the children and parents attended at least one of the group sessions. Participation in both groups did drop somewhat from Grade 1 to Grade 3 (from 98 to 86% for the child groups and from 96 to 80% for the parent groups). A significant portion of this decrease appears to be due to families moving out of the service area (see above). In general, rates of participation in both groups were relatively high. Children attended 77–79% of the sessions, whereas parents attended 71–72% of the sessions (column 2 of Table I). Eighty-four to 90% of the children who attended group in a given year attended at least 50% of the sessions. For the parents, 78–79% of the parents who attended a parent group in a given year attended at least 50% of the sessions (column 3 of Table I). Given the high-risk nature of the sample we started with, our goal was to provide intensive interventions at the particularly risky transition points of entry into elementary school, middle school (preadolescence), and high school (adolescence). We expected some heterogeneity to develop within the sample, as some children and families would improve markedly during the initial school years whereas others would continue to struggle. Yet, we hoped to maintain therapeutic relations with all, recognizing that even the improved children might need support during later risk periods. By Grade 3, the monthly sessions were designed to provide developmental support appropriate for all of the families, with the additional criterion-based services (e.g., tutoring, additional home visiting) serving children and families with greater needs. The high rates of group participation through Grade 3 suggest that this strategy was successful in maintaining the involvement of the majority of children and families. Given the positive outcomes evident during the early school years (CPPRG, 1999a, 2002), the preventive efforts appeared worth sustaining, even as the prevention trial itself moved into later years. The next section describes community efforts at sustaining the early grade school prevention program. Dissemination of Fast Track: Evaluation of Feasibility, Sustainability, and Effectiveness As noted above, the Fast Track prevention trial includes three successive cohorts of children and families. Although the trial is still underway, all three cohorts have
Style file version July 26, 1999
The Conduct Problems Prevention Research Group completed the elementary school phase of the intervention. The Fast Track investigators and prevention staff are still working in the communities and with the schools, but have moved into the middle and high schools, making it possible to explore sustainable dissemination at the elementary level without compromising the trial. This opportunity has provided a chance to determine whether and how a university-initiated prevention science program might transition into a community-run sustainable venture. The issues regarding the feasibility of sustainable dissemination are somewhat different for the universal versus the indicated prevention program components, because the universal intervention is administered within a single, existing service delivery system (i.e., classroom teachers administer the program in schools), whereas the indicated prevention components require the coordination of multiple service agencies and the development of new delivery systems (e.g., mental health personnel along with educators are needed to administer the program, and new delivery systems are needed to provide the parent/child group programs, home visiting, and in-school peer-pairing and tutoring). Although more straightforward in dissemination than the indicated prevention components, the sustainable dissemination of the universal intervention has been affected by a variety of local issues. For example, one of the school districts (Durham) reorganized and merged with a neighboring district midway through the trial, integrating students across intervention, control, and new schools, thus making it impossible to complete the control versus intervention comparison of the universal PATHS intervention at that site. (PATHS continued to be a part of the intervention for high-risk students involved in the trial, with sessions taught outside of the school setting.) In this school district, the pressures of adjusting to the merger took precedence in administrative discussions, relegating discussions about school adoption of the PATHS curriculum to the “back burner.” Schools at the other three sites have varied in the degree to which they have pursued opportunities to sustain the PATHS program at the elementary level. As part of the initial agreement, the Fast Track staff offered to provide basic training in the PATHS intervention to all control schools at the completion of the trial. As of this writing, teachers from 60% of the control schools have participated in PATHS training. Two school districts in Pennsylvania have taken a further step and adopted PATHS district-wide, requiring it of all elementary teachers and developing contracts for ongoing teacher training and consultation independent of Fast Track. It is not clear exactly which issues affected district decisions to “institutionalize” the use of the PATHS curriculum, but it is notable that the highest level of institutionalization
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
Style file version July 26, 1999
Fast Track Implementation occurred at the rural site where school districts were smaller and had more stable teaching and administrative staffs than at the urban sites. Continuation of other components of the Fast Track intervention required schools and collaborating agencies to work together to coordinate their services and find a source of funds for the new delivery systems providing parent groups, child social-skill training groups, home visiting, and peer-pairing/tutoring services. (None of the participating communities provided this set of prevention services prior to the initiation of Fast Track.) Fast Track investigators often played a key role in helping to bring multiple agencies to the table to discuss and plan dissemination efforts, providing incentives for collaboration of schools and community agencies in the form of promised technical support and assistance developing grant proposals. At three sites, school districts successfully pursued funding that allowed them to continue some components of the intervention. For example, State Department of Education funds supported the continuation of parent and child groups in one of the school districts in Pennsylvania, and funds from the Alberta (Canada) provincial government allowed a school district in Calgary to initiate the full Fast Track model with the technical support and guidance of staff from the Seattle site. External funds also supported the development of a “school readiness” adaptation of the Fast Track parent and child groups for Head Start in Durham and a parallel developmental adaptation of PATHS for Head Start in Pennsylvania. Most recently, a federal funding opportunity (Safe Schools, Healthy Students Initiative) allowed schools at two of the four sites (Pennsylvania and Durham) to write prevention program grants that involved the local collaboration of mental health, early childhood, and law enforcement agencies with school districts in each of the communities. This opportunity has allowed the schools and communities to initiate an adapted replication of the Fast Track model, owned and directed by local school districts and community agencies. At both sites, the communities are using core components of the Fast Track model, while also adapting the model, adding new components (e.g., an early childhood extension, an in-school supervised suspension program). In a sense, the communities are using the Fast Track program as their base, and adapting or adding components to “make the program their own.” The key to the success of these efforts will be determined by the extent to which these adaptations maintain the fidelity of the core components of the program, whereas also becoming successfully housed within existing school and service agencies in the community for long-term sustainability. University-based investigators at each site continue to work with these community groups, conducting
15 evaluations to determine the effectiveness of the dissemination efforts. One issue that is sometimes raised in discussions of the Fast Track program is whether a 10-year program with universal and indicated components can properly be called a prevention program, or whether such sustained and intensive efforts better qualify as treatment. Our belief is that effective prevention is most likely when schools and communities collaborate in the provision of a comprehensive network of services, which include multiple levels of prevention (e.g., universal, selective, indicated), which offer sustained support during childhood and adolescence, and which provide a natural interface with more intensive treatment services when needed. In the Fast Track program, all levels of prevention services were provided by our “stand alone” agency, but sustainable infrastructures for this program will most likely involve multiagency collaboratives rather than single agency efforts, as evidenced in the dissemination projects under way. For example, in both of the Safe Schools, Healthy Student initiatives currently using the Fast Track model, prevention services are being provided and cofunded by multiple agencies (e.g., the school, community mental health, early childhood, law enforcement), and blended with treatment services, such that children and families can move into and out of more intensive individualized services as needed. Indeed, the orientation and organization of the Fast Track model is very consistent with some of the more recently developed family-based treatment models, such as Multisystemic Therapy (Henggeler & Borduin, 1990). Such ideological commonalities may provide the foundation for the future development of integrated systems of prevention and treatment services.
CONCLUSION In the prevention field, leaders are calling for comprehensive prevention programs that have the potential to provide coordinated networks of services, and to bridge the efforts of education and mental health agencies (Coie et al., 1993; Weissberg & Greenberg, 1998). The need for a comprehensive and long-term set of coordinated prevention services is especially critical for the prevention of conduct disorder, given the developmental research base which has documented the multidetermined developmental trajectories of the “early-starting” delinquent youth (Loeber et al., 1993). Based upon developmental research, the Fast Track prevention efficacy trial represents a prevention science approach to the design and evaluation of long-term, comprehensive prevention services. Key challenges to implementation included entering communities,
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
15:54
16 engaging schools and families, adapting the program to the variety of needs and cultural backgrounds of the participants and communities, and beginning to build a foundation for long-term sustainability. The Fast Track experience suggests that a research-driven prevention science model can effectively initiate and sustain school, family, and community engagement, but also illustrates the extensive efforts at partnership-building, the flexibility in program presentation and implementation, and the longterm commitment by university investigators needed to produce this outcome.
ACKNOWLEDGMENTS Support for this project came from the National Institute of Mental Health through grants R18MH48083, R18MH50951, R18MH50952, and R18MH50953 to CPPRG. The Center for Substance Abuse Prevention also provided support through a memorandum of support with the NIMH. Support has also come from the Department or Education grant S184430002 and NIMH grants K05MH00797 and K05MH01027. Appreciation is expressed to the parents, teachers, students, and school district personnel who supported this research in the Durham, Nashville, central Pennsylvania, and Seattle areas.
REFERENCES Bates, J. E., Bayles, K., Bennett, D. S., Ridge, B., & Brown, M. M. (1991). Origins of externalizing behavior problems at eight years of age. In D. J. Pepler & K. H. Rubin (Eds.), The development and treatment of childhood aggression (pp. 92–120). Hillsdale, NJ: Erlbaum. Bierman, K. L., Greenberg, M. T., & the Conduct Problems Prevention Research Group. (1996). Social skills training in the Fast Track Program. In R. D. Peters & R. J. McMahon (Eds.), Preventing childhood disorders, substance abuse, and delinquency (pp. 65– 89). Thousand Oaks, CA: Sage. Campbell, S. B., Breaux, A. M., Ewing, L. J., & Szumowski, E. K. (1986). Correlates and prediction of hyperactivity and aggression: A longitudinal study of parent-referred poblem preschoolers. Journal of Abnormal Child Psychology, 14, 217–234. Chamberlain, P., Patterson, G., Reid, J., Kavanagh, K., & Forgatch, M. (1984). Observation of client resistance. Behavior Therapy, 15, 144–155. Coie, J. D., Watt, N. F., West, S. G., Hawkins, J. D., Asarnow, J. R., Markman, H. J., Ramey, S. L., Shure, M. B., & Long, B. (1993). The science of prevention: A conceptual framework and some directions for a national research program. American Psychologist, 48, 1013– 1022. Conduct Problems Prevention Research Group. (1992). A developmental and clinical model for the prevention of conduct disorders: The Fast Track program. Development and Psychopathology, 4, 509–527. Conduct Problems Prevention Research Group. (1999a). Initial impact of the Fast Track prevention trial for conduct problems: I. The highrisk sample. Journal of Consulting and Clinical Psychology, 67, 631–647.
Style file version July 26, 1999
The Conduct Problems Prevention Research Group Conduct Problems Prevention Research Group. (1999b). Initial impact of the Fast Track prevention trial for conduct problems: II. Classroom effects. Journal of Consulting and Clinical Psychology, 67, 648–657. Conduct Problems Prevention Research Group. (2000). Merging universal and indicated prevention programs: The Fast Track model. Addictive Behaviors, 25, 913–927. Conduct Problems Prevention Research Group. (2002). An end of thirdgrade evaluation of the impact of the Fast Track prevention trial with children at high risk for adolescent conduct problems. Journal of Abnormal Child Psychology, 30, 19–35. Connell, J. P., Kubisch, A. C., Schorr, L. B., & Weiss, C. H. (Eds.). (1995). New approaches to evaluating community initiatives: Concepts, methods, and contexts. Washington, DC: Aspen Institute. Cook, E. T., Greenberg, M. T., & Kusche, C. A. (1994). The relations between emotional understanding, intellectual functioning and disruptive behavior problems in elementary school-aged children. Journal of Abnormal Child Psychology, 22, 205–219. Dodge, K. A., Bates, J. E., & Pettit, G. S. (1990). Mechanisms in the cycle of violence. Science, 250, 1678–1683. Dodge, K. A., Pettit, G. S., & Bates, J. E. (1994). Socialization mediators of the relation between socioeconomic status and child conduct problems. Child Development, 65, 649–665. Dodge, K. A., Pettit, G. S., McClaskey, C. L., & Bown, M. (1986). Social competence in children. Monographs of the Society for Research in Child Development, 51 (2, Serial No. 213). Dumas, J. E. (1989). Primary prevention: Toward an experimental paradigm sensitive to contextual variables. Journal of Primary Prevention, 10, 27–40. Dumas, J. E., & Wahler, R. G. (1983). Predictors of treatment outcome in parent training: Mother insularity and socioeconomic disadvantage. Behavioral Assessment, 5, 301–313. Dunst, C. J., Trivette, C. M., & Deal, A. (1989). Engaging and empowering families: Principles and guidelines for practice. Cambridge, MA: Brookline Books. Elias, M. J., & Branden, L. R. (1988). Primary prevention of behavioral and emotional problems in school-aged populations. School Psychology Review, 17, 581–592. Elias, M. J., & Clabby, J. F. (1992). Building social problem-solving skills: Guidelines from a school-based program. San Francisco: Jossey-Bass. Harnish, J. D., Dodge, K. A., Valente, E., & the Conduct Problems Prevention Research Group. (1995). Mother-child interaction quality as a partial mediator of the roles of maternal depressive symptomatology and socioeceonomic status in the development of child behavior problems. Child Development, 66, 739–753. Henggeler, S. W., & Borduin, C. M. (1990). Family therapy and beyond: A multisystemic approach to treating the behavior problems of children and adolescents. Pacific Grove, CA: Brooks/Cole. Horvath, A. O. (1994). Research on the alliance. In A. O. Horvath & L. S. Greenberg (Eds.), The working alliance: Theory, research, and practice (pp. 259–280). New York: Wiley. Institute of Medicine. (1994). Reducing risks for mental disorders: Frontiers for preventive intervention research. Washington, DC: National Academy Press. Kazdin, A. E. (1987). Treatment of antisocial behavior in children: Current status and future directions. Psychological Bulletin, 102, 187–203. Kazdin, A. E., & Weisz, J. R. (1998). Identifying and developing empirically supported child and adolescent treatments. Journal of Consulting and Clinical Psychology, 66, 19–36. Kusche, C. A., & Greenberg, M. T. (1994). The PATHS Curriculum. Seattle, WA: Developmental Research and Programs. Ladd, G. W., Price, J. M., & Hart, C. H. (1990). Preschoolers’ behavioral orientations and patterns of peer contact: Predictive of peer status? In S. R. Asher & J. D. Coie (Eds.), Peer rejection in childhood (pp. 90–115). New York: Cambridge University Press. Lochman, J. E., & the Conduct Problems Prevention Research Group. (1995). Screening of child behavior problems for prevention
P1: GDW/GVG Journal of Abnormal Child Psychology
pp320-jacp-363409
January 31, 2002
Fast Track Implementation programs at school entry. Journal of Consulting and Clinical Psychology, 6, 549–559. Lochman, J. E., & Dodge, K. A. (1994). Social-cognitive processes of severely violent, moderately aggressive and nonaggressive boys. Journal of Consulting and Clinical Psychology, 62, 366– 374. Loeber, R., Wung, P., Keenan, K., Giroux, B., Stouthamer-Loeber, M., Van Kammen, W. B., & Maughan, B. (1993). Develomental pathways in disruptive child behavior. Development and Psychopathology, 5, 103–133. McMahon, R. J., Slough, N. M., & the Conduct Problems Prevention Research Group. (1996). Family-based intervention in the Fast Track Program. In R. D. Peters & R. J. McMahon (Eds.), Preventing childhood disorders, substance abuse, and delinquency (pp. 65– 89). Thousand Oaks, CA: Sage. McMahon, R. J., Slough, N., & Conduct Problems Prevention Research Group. (2001). Fostering parent and child participation in the Fast Track prevention trial. Manuscript in preparation. Moffitt, T. E. (1990). Juvenile delinquency and attention deficit disorder: Boys’ developmental trajectories from age 3 to age 15. Child Development, 61, 893–910. Offord, D. R., Alder, R. J., & Boyle, M. H. (1986). Prevalence and sociodemographic correlates of conduct disorder. American Journal of Social Psychiatry, 6, 272–278. Offord, D. R., Boyle, M. H., & Racine, Y. A. (1991). The epidemiology of antisocial behavior in childhood and adolescence. In D. J. Pepler & K. H. Rubin (Eds.), The development and treatment of childhood aggression (pp. 31–54). Hillsdale, NJ: Erlbaum. Orrell-Valente, J. K., Pinderhughes, E. E., Valente, E., Laird, R. D., & the Conduct Problems Prevention Research Group. (1999). If it’s offered, will they come? Influences on parents’ participation in a community-based conduct problems prevention program. American Journal of Community Psychology, 27, 753– 783.
15:54
Style file version July 26, 1999
17 Patterson, G. R., & Chamberlain, P. (1994). A functional analysis of resistance during parent training therapy. Clinical Psychology: Science and Practice, 1, 53–70. Patterson, G. R., Capaldi, D. M., & Bank, C. L. (1991). An early-starter model for predicting delinquency. In D. J. Pepler & K. H. Rubin (Eds.), The development and treatment of childhood aggression (pp. 139–168). Hillsdale, NJ: Erlbaum. Pattterson, G. R., & Forgatch, M. S. (1985). Therapist behavior as a determinant for client noncompliance: A paradox for the behavior modifier. Journal of Consulting and Clinical Psychology, 53, 846– 851. Richman, N., Stevenson, J., & Graham, P. J. (1982). Preschool to school: A behavioral study. London: Academic Press. Rutter, M., & Giller, H. (1983). Juvenile delinquency: Trends and perspectives. New York: Guilford. Rutter, M., Maughan, B., Mortimore, P., Ouston, J., & Smith, A. (1979). Fifteen thousand hours: Secondary schools and their effects on children. Cambridge: Harvard University Press. Shen, W. W., Sanchesz, A. M., & Huang, T. D. (1984). Verbal participation in group therapy: A comparative study on New Mexico ethnic groups. Hispanic Journal of Behavioral Sciences, 6, 277–284. Tremblay, R. E., Masse, L. C., Pagani, L., & Vitaro, F. (1996). From childhood aggression to adolescent maladjustment: The Montreal prevention experiment. In R. D. Peters & R. J. McMahon (Eds.), Preventing childhood disorders, substance abuse and delinquency (pp. 268–298). Thousand Oaks, CA: Sage. Wasik, B. H., Bryant, D. M., & Lyons, C. M. (1990). Home visiting: Procedures for helping families. Newbury Park, CA: Sage. Weissberg, R. P. (1995, August). What I have learned about prevention research (so far!). Paper presented at the meeting of the American Psychological Association, New York. Weissberg, R. P., & Greenberg, M. T. (1998). Prevention science and collaborative community action research: Combining the best from both perspectives. Journal of Mental Health, 7, 479–492.