Suicide is the second leading cause of death among youths 15-19 yrs old. It is estimated that 40-50% of high-school youth have considered suicide; that there are about 10 suicide attempts for every completion; and that co-occurring risk factors frequently include depression, drug abuse and school deviance. Most school-based prevention programs have not been scientifically evaluated; new research suggests some may even increase distress among vulnerable youth. Recommended prevention approaches embrace recognizing the risk factors; intervening in a broader context of mental health, coping skills, and social support; and targeting high-risk youth specifically. Accordingly, this prevention trial focuses on 2 promising prevention programs: 1) a MAPS (Measure of Adolescent Potential for Suicide) protocol, a 1:1 assessment interview and brief crisis-intervention that enhances social network connections with parents and school persons; and 2) the addition of CAST (Competence and Support Training)(a 10-hr life skills training program delivered in small groups) to the MAPS intervention. In line with prevention science, these interventions are aimed at reducing posited antecedent risk factors and enhancing protective factors. Overall, the goal is not only to test the efficacy of the 2 approaches, but also to compare the effects of varying """"""""dosage levels"""""""" of social support and skills training in reducing adolescent suicide potential. The key program elements proposed in this study are distilled from previous interventions which produced large effects (>1 SD above the M) for reducing suicide behaviors, depression and hopelessness.
The specific aims are: l) to test the hypothesis that a MAPS+CAST program will be more effective than MAPS alone, which in turn will be more effective than a CONTROL condition for decreasing depression, drug involvement, school deviance and suicide behaviors; 2) to test the efficacy of MAPS, MAPS+CAST, and CONTROL conditions for increasing 2 mediating factors, personal and social resources; and 3) to explore a causal model of theoretically predicted mechanisms and effects of the experimental prevention approaches. Differential effects for gender will be tested across aims. We will randomly assign (to MAPS, MAPS+CAST, and CONTROL conditions) 441 high-risk youth who screen in at suicide risk in 5 high schools. We expect to retain a total of 375 subjects, 125 in each condition, at 9 mos follow- up. Tests of Aims 1 & 2 will include latent growth modeling to specifically address the variance/covariance structure and change in levels of the 4 dependent variables and 2 mediating variables over time and among the 3 experimental conditions. To test Aim 3 we propose structural equation modeling. Thus, this prevention trial should determine if the 2 prevention approaches work, demonstrate their relative efficacy, and explicate the mechanisms by which social support and life skills training interventions work to reduce suicide potential among high-risk youth.
