Outcome predictors in substance use disorders

Identifying predictors of treatment response in substance abuse is complicated not only by the heterogeneity of substance use disorders, but also by the various treatments available and the research community's choice of which predictor variables to study. Research in the area reflects the contrasting approaches of those who look for unifying principles underlying certain illnesses, and those who seek finer delineation of subgroups for whom specific treatments may be individualized. As a consequence, the search for predictors serves both purposes, by describing shared factors associated with recovery across different substance use disorders, while also characterizing clinically relevant patient subgroups. Also complicating studies of predictors is the large number of effective treatment options available in substance abuse. A characteristic associated with a positive response to a psychosocial treatment may not have the same relationship to a medication therapy, and predictors vary even among various psychosocial interventions and different pharmacotherapies. The selection of predictors in studies influences the interpretation of data and the clinical relevance of findings. To some degree, investigators are slaves to their technology and expertise. As a consequence, very few truly integrative biopsychosocial studies exist. The field is then left to interpret whether different predictors are linked, whether some are merely epiphenomena, and whether they have practical or heuristic value. Does a specific biologic marker that is associated with poor response merely reflect greater use of the substance, or does it provide insights into the underlying process of disease development or response to a specific therapy? Is low self-efficacy a consequence of poor coping mechanisms, or the logical conclusion of a person who has a relatively intractable form of addiction? The following overview should be read with these limitations in mind. For simplicity it is organized by drug class, although some of the studies address additional drug and psychiatric comorbidities. For each drug of abuse biologic and psychosocial factors are discussed; however, there are few studies examining the interaction. Given the practical limitations on the length of this article, alcohol dependence is focused on in greatest detail, and other substances in less detail.