This application proposes a secondary data analysis of outpatient adolescent substance abuse treatment outcome research based on data from six prior studies. A master data set will be developed from randomized trials conducted at Oregon Research Institute's Center for Family and Adolescent Research, the Cannabis Youth Treatment study, and the University of Miami's Center for Family Studies. Each of these studies involved a form of family treatment (FT), and five studies included one or more versions of cognitive behavior therapy (CBT). The analyses will focus on identifying outcome predictors for CBT and FT. Because these studies have employed similar methodologies and have included many of the same or comparable assessment measures, including outcome measures and risk and resilience variables, integration of the data sets will provide the exceptional and unique opportunity to identify key predictors of treatment success and failure. The analyses will specifically address two consistent and heretofore unexplained findings from the adolescent substance abuse treatment outcome literature: the variability in treatment response within interventions and the existence of clusters of individuals who fit identifiable response profiles. Once the data sets are properly combined, latent class longitudinal growth modeling procedures will be used to identify different classes of individuals who share common trajectory estimates of substance use change across assessment points. These analyses will focus upon marijuana use as the primary dependent variable. The sample size available from this master data set (1075 males and 225 females);will permit key analyses on gender and race/ethnicity variables. The master data set will include approximately 25% African Americans, 25% Hispanics, and 50% non Hispanic Caucasians. Pooled data will have relatively few individuals of Asian or Native American backgrounds (4%). The research will examine the role of gender and race/ethnicity as possible moderator effects of the latent classes within each of the major treatment modalities. The proposed secondary analyses also examine the joint and independent effects of internalizing, externalizing, family dysfunction, individual coping skills, and exposure to deviant peers as predictors of change profiles. We hypothesize that externalizing problem behaviors are associated with improvements in FT interventions, while internalizing disorders are associated with improvements in CBT. Taken together, the proposed analyses will provide important insight into the variability in treatment outcomes that characterizes the adolescent substance abuse treatment field. The identification of distinct profiles or trajectories of change for treated youth and the risk and protective factors that predict these change profiles has critical implications for improving adolescent treatment outcomes. Specifically, if researchers can identify predictors of profiles of change, treatments can be developed or refined to better meet the needs of individual adolescents, and treatment resources can be allocated more efficiently to specific subgroups of adolescents.
| Brody, Janet L; Turner, Charles W; Annett, Robert D et al. (2012) Predicting adolescent asthma research participation decisions from a structural equations model of protocol factors. J Adolesc Health 51:252-8 |
| Sexton, Thomas; Turner, Charles W (2010) The effectiveness of functional family therapy for youth with behavioral problems in a community practice setting. J Fam Psychol 24:339-48 |
