Prevention Programs

Urban Institute Project Report

Problem Behavior Prevention and School-Based Health Centers:
Programs and Prospects

The intervention programs identified and described in the previous chapter demonstrated positive impact on one or more risk-related behavioral outcomes in four major problem behavior areas of potential interest to SBHCs. The list of 51 programs embodies a set of "best outcomes" in these areas. It is intended to be analogous to the idea of compiling a list of "best practices" in a given discipline. We further refined this list for the research synthesis (Cook, et al., 1993) described in this chapter by limiting our in-depth context analysis to programs that met stricter criteria. This selection process culminated in a subset of 21 notable programs, which were examined to "deconstruct" the salient elements by applying a comprehensive classification/coding scheme to their main program content and delivery characteristics.

We then counted the overall frequency of each characteristic appearing in our subset of prevention programs. We were able to identify and compile the common program components and delivery methods into a reasonably comprehensive set of "best practices," (i.e., program elements associated with most or many of these effective interventions). Conceptually, this process is virtually the same one used in a standard meta-analysis, except that typically the meta-analyst includes the largest and most complete set of studies on a given topic, regardless of whether positive or negative effects are reported. In contrast, we have chosen to exclude programs with negative or no effects.

In practice, positive program findings, or best outcomes are correlated with the quality of program implementation, the rigor of the evaluation study design, and the availability of control variables (e.g., parental education or income level) that are used to statistically subtract out the effects of as many other contributing and limiting non-treatment factors as possible. Therefore, the logic of our approach requires us to separate program content and delivery features that are associated with successful interventions from major methodological or evaluation design features, particularly randomized controlled trial and matched group quasi experiments that are often associated with detecting strong behavioral impacts. In this sense, we are essentially holding methodological rigor constant in our compilations and research synthesis. Thus, we can then be reasonably confident that the success of these interventions can be attributed primarily to their program content and delivery mechanisms, not to their methodological rigor per se.

Methodology

Intensive Review Selection Criteria. The studies selected for intensive review include the 21 successful intervention programs that met the most stringent selection criteria listed in Chapter 2 (studies marked with an asterisk[*] in the program descriptions) and that we judged to have made especially significant contributions or been especially influential in their particular content area. Of these 21 programs, 16 were conducted using random assignment of individuals, classrooms, or in some studies schools or whole communities to treatment or control conditions. The remaining five programs were evaluated using the strongest matched-groups quasi-experimental designs available.

Coding Approach and Method. Thus, each of the 21 intervention evaluations was reviewed and salient program content and delivery characteristics were classified using a coding approach modified from Tobler’s (1992) school-based drug use prevention meta-analysis code book and further refined by Rooney & Murray (1996) for use in their meta-analysis of school-based tobacco prevention programs. Since the programs examined for this project cover not only substance use but also sexual behavior, conflict resolution and violence prevention, and mental health/conduct disorders, we made some additional modifications to Rooney & Murray’s coding scheme.

Based on the key published reports on program impacts, each program was coded for the following information: characteristics about the target population, curriculum content, program delivery methods, program elements, study methodology, and evaluation design. Two UI project staff members coded all applicable characteristics of each program independently. In a few instances when the needed information was not found in the literature, actual curriculum materials were consulted, if available. Nevertheless, there remained some characteristics that were not found and a "don’t know" code was used. When there was a disagreement between the two coders’ responses, the "correct" response was decided upon by the Project Director.

A Topographic Profile of Effective Interventions

Before reporting our synthesis and summary of specific program content and delivery mechanisms, we summarized the programs’ general characteristics or attributes. Of the 21 programs intensively reviewed:

Synthesis and Summary of Common Program Characteristics of Effective Interventions

Based on the intensive review and deconstruction of program content and delivery methods used in the 21 interventions, we conducted a research synthesis, finding that there are, indeed, important common programming elements among these successful programs. By knowing what characteristics are associated with the most successful programs across program content areas, we hoped to gain a greater understanding of what mechanisms are likely to work best in problem behavior prevention in general; and more specifically, what are the essential elements of effective programs that could be adapted for use in SBHC settings. Moreover, finding common elements across problem behavior content areas would be scientifically useful because there is strong evidence that problem behaviors tend to cluster together within individuals and most experts assume there is an underlying problem behavior syndrome contributing to the clustering (e.g., Dryfoos, 1990). Unfortunately, with only 10 sexuality and 8 substance use programs eligible for intensive review, the chances of identifying a large set of common elements were reasonably small.

Nevertheless, we identified six areas of substantial overlap:

• 16 programs (including seven sexuality and all eight substance use programs) use social learning and social cognitive theories to leverage behavior change;
• 14 programs (including eight sexuality and six substance use) have a set of narrowly and clearly articulated behavioral goals;
• 17 programs (eight sexuality and all eight substance use) have clear and skill-based program components;
• 20 programs (including nine sexuality and seven substance use) have a written curriculum and provide strong instructor training and feedback;
• 10 program (including four sexuality and three substance use) have relatively long duration and intensity; and
• 10 programs (including three sexuality and four substance use) have multiple component interventions

Interestingly, these six program elements, which are essentially the same as those noted by Kirby (1997) in his comprehensive review of sexuality education programs, are also applicable to substance use prevention programs. However, our synthesis does not lend equally strong support to two additional characteristics of successful sexuality programs discussed in Kirby’s review. We found less tangible indication that effective interventions have: (1) strategies tailored to the age, culture, and experience level of the target population (only seven programs, including three sexuality and one substance use); and (2) a strong peer support component (only eight programs actually used either same-age or older peers to present or help present intervention materials, but peer-led or co-led components were significantly more effective than traditional presentation methods or comparison group activities in six programs, including three sexuality and three substance use).

The following section provides more detailed contextual analysis of the major program elements enumerated above.

1. Many of the best prevention programs that we studied are theory-based and use social cognitive theories or behavior change theories to explain why and how their intervention will effectively change behavior. In our analysis, all but three programs stated that they were based in a specific theory, and eight were based in multiple theories. Four programs were based in principles of social learning theory; ten were based in social influence theory; and two were based in principles of the health belief model.
   
   
2. The most effective programs have a few specific and clearly articulated goals for health risk-taking behavior change. Sixteen programs highlight the negative consequences of the behavior being addressed. For example, Know Your Body, a smoking prevention intervention, demonstrates the immediate and long-term effects of smoking on the body as part of the curriculum. Becoming a Responsible Teen presents the negative consequences of sexual risk behavior by incorporating a session in which HIV-negative teens have a small discussion group with local HIV-positive teens.

   Eleven programs try to correct the expectation of high normative use of drugs or low use of condoms. Becoming a Responsible Teen has a component in which a small group of participants discuss reasons why adolescents do not use condoms. The final part of this session uses cognitive restructuring to teach youth to question counterproductive beliefs and replace them with statements more consistent with preventive behavior related to safe sex. Minnesota Heart and Health also has a session in which students are forced to question whether smoking is actually a normative behavior. Students’ expectations about how many teens smoke are compared with actual data showing that smoking prevalence is usually less than most teens think.

3. All of the selected programs use interactive student-to-student skill building methods to address the problem behaviors targeted in the interventions. These methods often include resistance/refusal and negotiation skills developed through the use of role-playing and rehearsal; use of behavioral modeling of communication and other skills by the trainers who also use guided practice and immediate feedback techniques to increase these kinds of behavioral skills; and other active student-to-teacher and student-to-student experiential, personalized instructional styles (e.g., students creating their own role play scripts).

   Seventeen programs provide examples and training on how to act out resistance skills and provide opportunities for guided practice and behavioral modeling. For example, several programs include role playing exercises or provide students with an opportunity to rehearse methods they learned to resist influence. Eight programs use methods such as problem solving, decision making, and self-control to address the specific risk behavior. Reducing the Risk, addresses the social pressures of having sex faced by adolescents by examining some of the common pressure tactics used by other teens. The intervention helps students to develop and practice effective strategies and skills to resist those pressures. Becoming a Responsible Teen also has a component in which role playing situations are conducted. During these role playing sessions, feedback and corrective suggestions are provided from the group leaders.

   Eighteen programs try to improve verbal and non-verbal communication skills. Safer Choices is an example of a program with a communication skill development component. Students learn characteristics of clear "no" statements and contexts in which they could use those skills. Thirteen programs teach general assertiveness skills.

   Sixteen programs focus on the social influences that encourage behavior, including peers and the media. Eight programs, mostly those addressing substance use behaviors, teach skills to resist advertising appeals. Project Shout, a tobacco prevention curriculum, includes a component that helped students become more aware of the social influences that promote tobacco use. As part of the program, students write letters to magazine editors and film producers about media portrayal of tobacco use. Students and parents are also given regular updates regarding tobacco legislation and research. Postponing Sexual Involvement uses the social inoculation model, which recognizes the societal influences of peers and other social forces and helps them to be educated about how to understand and resist the influences. The social inoculation approach uses role models-- and in the case of Postponing Sexual Involvement, slightly older peers-- to influence adolescent behavior related to sexual involvement.

4. Most of the successful programs examined in this analysis are based on a written curriculum that is presented by a person trained in the curriculum. In half of the programs, a classroom teacher presents the curriculum after he/she is appropriately trained. For the other components of the programs and for the other programs that do not use teachers, the other presenters are either health educators, same age peer leaders, older age peer leaders, parents, community members, or health professionals. For all the programs, a trained presenter teaches the curriculum. The training process varies among the programs, but for 20 of the programs the training involves both written materials and practice. The other program does not specify how the information is provided during the training. Sixteen programs include an observation of the presenter in the training and implementation process.
   
   
5. The most effective programs are generally more intensive in terms of the number of sessions and the length of intervention. Of the programs we examined, two-thirds include over 10 hours of intervention. Know Your Body and Seattle Social Development Program have over 100 hours of intervention. Half of the programs are completed over ten sessions, and a few include sessions during an entire school year or more.
   
   
6. Some of the most effective programs have multiple intervention components using multiple techniques and delivery mechanisms. Most of the multi-component programs are based in a school setting and have a classroom component to the intervention. A few programs only have a classroom component or another single component to the intervention, but many of the programs have more than one component. Some of the multiple component interventions include involvement by the community and/or the parents. Poder Latino, Project Northland, Midwestern Prevention Program as well as five other programs have multiple components to their intervention. Eight programs involve the community in some capacity, and seven programs have components involving parents.

Peer support, when used in conjunction with other delivery mechanisms, can be an effective part of many multi-component programs. Several of the multicomponent programs in our review include a strong peer education/support component. All the peer-led programs discussed here are based in social learning or social influence theories, which provide the theoretical underpinning for using peer education components. In four or five of the programs same-age peer leaders are recruited and in another four or five programs older-age peer leaders are recruited to present the message.

The process of selecting and reviewing the 21 programs presented here has been informative in discerning what works in preventing adolescent and childhood risk behaviors and what mechanisms help to make interventions successful. Because the standards we set for inclusion in our analysis were quite stringent, and because the core set of common features we identified so closely mirrored the features identified by Kirby and by Tobler, we feel that these common elements are central to the design and implementation of successful programs. The programs we reviewed not only represented programs that were methodologically superior and led to positive behavioral outcomes, but also they cut across several risk behaviors, including sexuality and substance abuse programs.