This review offers evidence that important and meaningful progress has been made in prevention research with children, families and schools during the last two decades. There have been advances in the theory, design, and evaluation of programs, and there are a growing number of programs with documented efficacy of beneficial impact on the reduction on psychiatric symptomology. These research findings have also influenced public policy as federal, state, and local governments are now calling for the utilization of empirically-validated, effective models of intervention for children and families.

Best Practices in Prevention Programming

Over time, researchers, practitioners, and policy makers have developed a more realistic perspective on the necessary intensity and comprehensiveness of programming to prevent psychopathology and promote positive development, especially with children and adolescents growing up in high-risk environments (Panel on High-Risk Youth, National Research Council, 1993). The following conclusions can be made regarding validated programs:

• Short-term preventive interventions produce time-limited benefits, at best, with at-risk groups whereas multi-year programs are more likely to foster enduring benefits.
  
• Although preventive interventions may effectively operate throughout childhood (when developmentally-appropriate risk and protective factors are targeted) given the resistance to treatment of serious conduct problems, ongoing intervention starting in the preschool and early elementary years may be necessary to reduce morbidity.
  
• Preventive interventions are best directed at risk and protective factors rather than at categorical problem behaviors. With this perspective, it is both feasible and cost-effective to target multiple negative outcomes in the context of a coordinated set of programs.
  
• Interventions should be aimed at multiple domains, changing institutions and environments as well as individuals.
  
• Prevention programs that focus independently on the child are not as effective as those that simultaneously "educate" the child and instill positive changes across both the school and home environments. The success of such programs is enhanced by focusing not only on the child's behavior, but also the teacher's and family's behavior, the relationship between the home and school, and the needs of schools and neighborhoods to support healthy norms and competent behavior.
  
• There is no single program component that can prevent multiple high-risk behaviors. A package of coordinated, collaborative strategies and programs is required in each community. For school-aged children, the school ecology should be a central focus of intervention.
  
• In order to link to other community care systems and create sustainability for prevention, prevention programs will need to be integrated with systems of treatment. In this way, communities can develop common conceptual models, common language, and procedures that maximize the effectiveness of programs at each level of need. Schools, in coordination with community providers, are a potential setting for the creation of such fully-integrated models. It is surprising that few comprehensive interventions have been developed and evaluated that combine school-wide primary prevention together with secondary prevention and treatment.

Future Directions

The past decade has brought to fruition well-designed studies that demonstrate the potential of preventive intervention in reducing harmful symptoms for children and youth. However, given the need for effective research in this field there are numerous issues for future research, policy, and practice.

• Few studies meet the criteria for fully-validated program models. Of most concern are the lack of replication of program effects by independent investigators and the absence of long-term follow-up to examine stability of program effects.
  
• One of the weaknesses in present research efforts is the lack of comprehensive follow-up data to chart the developmental processes of program participants in the years after receiving interventions. As a number of programs show stronger impacts at follow-up than they did at post-test, it is likely that the effects of prevention programs are underestimated at present; examining distal outcomes is critical.
  
• There has been greater attention to preventive interventions focused on externalizing disorders. As such, we still know less regarding effective prevention models for internalizing disorders. Further, as many children show risk for, or co-morbidity of internalizing and externalizing problems, intervention projects should examine the differential effects that interventions might have on those that have risks or early symptoms of co-morbidity. Further, outcome measures should include assessment of both externalizing and internalizing symptoms.
  
• A broader point is that there is significant inter-individual variability in program effects. There has been little focus on what factors in the child (e.g., gender, ethnicity) or environment (e.g., quality of home or school environment) might moderate the impact of intervention. It is necessary to know more regarding for whom specific programs are most likely to be effective.
  
• With few exceptions, there has been little exploration of how the quality of implementation affects outcomes. There is a need for greater attention to both the measurement of dosage as well as the quality and fidelity of the intervention delivery, especially as empirically-validated prevention programs begin to "go to scale."
  
• Due in part to the categorical nature of funding, programs often assess quite narrow outcomes (e.g., only substance abuse, psychological symptoms, positive adaptation). As programs often focus the intervention on modifying common risk factors for multiple problem behaviors as well as promoting competence, measures of multiple dimensions of outcome are necessary.

In spite of substantial gains in prevention research during the last two decades, it is important to acknowledge that considerable progress is needed to affect more tangibly the lives of American children and families. Only a small group of researchers have designed and evaluated multi-year, multi-component programs that target multiple mental health outcomes. Few successful efficacy trials have been replicated by independent investigators, and there have been even fewer attempts to evaluate the implementation process and impact of widely disseminated program models. In summary, although a solid scientific base is being created, the most important knowledge regarding preventive interventions will come from the next generation of prevention researchers!