Journal of Consulting and Clinical Psychology 2000, Vol. 68, No. 4, 684-696 Copyright 2000 by the American Psychological Association, Inc. 0022-006X/OW5.00 DOI: 10.1037//0022-006X.68.4.684 Adolescent Substance Abuse Treatment Outcome: The Role of Substance Abuse Problem Severity, Psychosocial, and Treatment Factors William W. Latimer University of Minnesota Michael Newcomb University of Southern California Ken C. Winters and Randy D. Stinchfield University of Minnesota A structural equation model incorporating substance abuse problem severity, psychosocial risk and protection, and treatment variables examined adolescent drug abuse treatment outcome pathways across 6- and 12-month follow-up points. Findings on resiliency factors and an empirical method adapted from previous research were used to select and assign 10 psychosocial factors to either a multiple protective factor index or a risk factor index. Gender, substance abuse problem severity, treatment modality, treatment length, and aftercare participation were also examined as outcome predictors. The findings suggest that treatment intensity decisions may be better informed by pretreatment psychosocial risk level rather than by substance abuse problem severity. The present study also suggests that drug-abusing adolescents who receive sufficiently long treatment, participate in aftercare, and possess at least 1 individual or interpersonal protective factor during their recovery process have the best chance to maintain gains made during treatment. Substance abuse problem severity and psychosocial outcomes among adolescents in treatment for substance abuse mirror the adult literature, with 1 in 2 youths abusing substances within 90 days posttreatment (Brown, Vik, & Creamer, 1989). With few exceptions, extant research has examined the impact of pretreat- ment psychosocial risk factors on adolescent treatment outcome. Two notable gaps characterize the current adolescent literature (Catalano, Hawkins, Wells, Miller, & Brewer, 1991). The paucity of research examining the influence of posttreatment factors on adolescent treatment outcome is noteworthy given that the adult literature suggests that posttreatment factors account for half of the variance in outcome (Moos, Finney, & Cronkite, 1990). Second, extant treatment outcome research has nearly excluded any exam- ination of psychosocial factors that protect against relapse. Current adolescent substance abuse treatment research reveals a dearth of findings (Catalano et al., 1991). The outcome predictors discussed below are organized according to demographic, sub- William W. Latimer, Institute on Community Integration, University of Minnesota; Michael Newcomb, Division of Counseling Psychology, Uni- versity of Southern California; Ken C. Winters and Randy D. Stinchfield, Department of Psychiatry, University of Minnesota. This research was supported by a Scientist Development Award (DA00254) and a FIRST Award (DA10777) from the National Institute on Drug Abuse (NIDA) and by NIDA Research Project Grants DA05104 and DA01070. Correspondence concerning this article should be addressed to William W. Latimer, who is now at the Department of Mental Hygiene, School of Hygiene and Public Health, Johns Hopkins University, 624 North Broad- way, 8th Floor, Baltimore, Maryland 21205. stance abuse problem severity, psychosocial risk and protection, and treatment domains. Extant research in the adolescent drug abuse treatment field has focused on the impact of gender and ethnicity on outcome (Cata- lano et al., 1991). In a majority of studies, ethnicity has not been associated with treatment completion or outcome (Jainchill, De Leon, & Yagelka, 1997). Findings on gender are mixed. Several studies on gender reveal no differences in outcome when treatment is sufficiently long (De Leon & Jainchill, 1991). Girls, however, may make greater gains on traits other than drug use, such as psychological well-being, in comparison with boys. A separate series of analyses conducted on the present sample found that girls exhibited less posttreatment alcohol use than boys; however, there was no gender effect for other substances (Latimer, Winters, Stinchfield, & Traver, in press). The present study incorporated gender in models as an intake control variable and predictor of prospective outcomes. The adult literature has consistently found that elevated pretreat- ment substance abuse problem severity predicts posttreatment re- lapse (Moos et al., 1990). There is surprisingly little known, however, about the extent to which greater substance abuse prob- lem severity at treatment entry predicts posttreatment substance use among adolescents. One study found that use of hard drugs other than marijuana predicts elevated posttreatment substance use (Friedman, Glickman, & Kovach, 1986). In the present study, relationships between levels of pre- and posttreatment substance abuse problem severity were examined. Relative to other treatment outcome predictors, there is a fairly broad base of research on psychosocial risk factors associated with adverse outcomes for recovering adolescents. Heightened pretreat- 684 This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.