Practice Profile: Interactive Programs for Preventing Marijuana Use in Middle School Students

Evidence Rating for Outcomes

Drugs & Substance Abuse | Marijuana

Date: This profile was posted on November 22, 2021

Practice Goals/Target Population

Interactive programs for preventing marijuana use in middle school students consists of skill-building and interaction-based activities (as opposed to lecture-style curricula) integrated into middle school-based programs (for grades 6–8) that are aimed at preventing marijuana use among adolescents ages 12–14.

Practice Theory

Research on risks related to marijuana use during adolescence has shown effects such as low school performance, addiction, impacts on working memory performance, and other risks that persist into adulthood. Focusing prevention efforts on middle school students could aid in changing expectations, beliefs, and behaviors related to marijuana use (Lize et al. 2017, p. 2).

Practice Components

Interactive programs can be delivered during school, after school, or a combination of both. These programs can be delivered by a teacher, a clinician, a police officer, a community volunteer, or a trained adult with a peer leader. The number of sessions varies between 3 and 40 and can include booster sessions 6 months to 1 year afterward.

Some examples of programs in this practice are

Across Ages, an intergenerational mentoring initiative designed to delay or reduce substance use by increasing the resiliency and protective factors of at-risk middle school youth. The interactive component of the program involves pairing older adult volunteers (55 and older) with students (10 to 13 years old) to create a special bonding relationship. The project also uses community service activities, provides a classroom-based life skills curriculum, and offers parent-training workshops.
LifeSkills® Training, a classroom-based, drug abuse–prevention program for upper elementary and junior high school students. This program aims to reduce substance use (and uptake, in particular) by increasing participants’ coping, refusal, and social skills and overall knowledge. The interactive program activities are based on cognitive–behavioral principles, including role-playing, modeling, immediate feedback, and reinforcement of positive behaviors.
Family Check-Up for Adolescents, a family-centered intervention designed to assist families with high-risk adolescents to reduce the growth of adolescents’ problem behaviors and substance abuse. Parents and adolescents are all involved in the interactive sessions. The program offers a menu of family-centered interventions that support family management practices and promote the well-being and improved behavior of children and adolescents.

Meta-Analysis Snapshot
	Literature Coverage Dates	Number of Studies	Number of Study Participants
Meta Analysis 1	1998-2015	30	29513

Meta Analysis 1

Lize and colleagues (2017) conducted a meta-analysis to examine the effectiveness of interactive, middle school–based drug prevention programs (in North America) on adolescent marijuana use. A systematic literature search was conducted to identify published and unpublished studies, using databases such as ProQuest, EBSCO, and Medline, for English-language studies from January 1998 to March 2014. Additional studies were identified from citations in prior reviews and eligible studies through April 2015. Studies included in the sample were randomized controlled trials and quasi-experimental designs of middle school–based (grades 6–8) substance use–prevention programs with interactive components (skills building and student engagement) that reported sufficient statistical information to calculate effect sizes of preventing cannabis use on adolescents ages 12–14. Studies were excluded if the sample included elementary or high school students, the primary program components were delivered in the home or in a community setting, it included program components described as extracurricular activities, it included comparison groups with another prevention program, it assessed only program completers, or attrition caused the sample groups to be nonequivalent.

Thirty studies yielding 23 independent samples were included. The studies included in the meta-analysis were conducted mainly in the United States, with one study from Canada. All but one study used random assignment of participants. Follow-up periods for the studies ranged from immediately posttreatment to 72 months, with the mean being 16 months. Sample sizes of the studies ranged from 42 to 5,756 participants, with the mean being 1,613 participants. Gender breakdown was approximately equal between boys and girls in study samples. Racial composition ranged from 97 percent White to 99 percent nonwhite samples. Studies were conducted in urban (9 studies), rural (4 studies), and mixed (10 studies) settings.

Study interventions were Project ALERT (5 studies), Life Skills Training (3 studies), All Stars (3 studies), and other program curricula (12 studies). Intervention sessions ranged from 3 to 40, with the mean being 17 sessions. Booster sessions were given in 10 studies between 6 months to 1 year postintervention. Interventions were delivered during school (17 studies), after school (2 studies), and a combination of during and after school activities (4 studies). Interventions were delivered by teachers (10 studies), a clinician (9 studies), a police officer (1 study), a trained adult with a peer leader (1 study), or a volunteer from the community (1 study). Comparison samples included students who received treatment as usual (11 studies), no programming (10 studies), or health information only (2 studies).

The meta-analysis looked at the following outcomes: posttreatment marijuana use, intent to use, and refusal skills. For the CrimeSolutions review of this meta-analysis, the outcome of interest was posttreatment marijuana use from random controlled designed studies (21 independent samples from 29 studies). Effect sizes (Cohen’s d) were calculated from continuous measures, and dichotomous measures were converted to the d index. Program effects were measured by computing the standardized mean difference effect size and weighted by the pooled standard deviations of the treatment and comparison groups. Individual effect sizes were combined using random effects models employing inverse variance weights with each effect size. Random effects models were used to account for possible variation other than sampling error among effect sizes as well as heterogeneity in programs and participants. To account for clustered studies, an effective sample size was estimated for each clustered sample to avoid inflating the effect size based on clustering.

Moderator analysis was conducted on the pooled effect size for any marijuana use with the program and study characteristic variables. For results of the moderator analysis, see Other Information.