PROJECT SUMMARY/ABSTRACT The identification of a valid and meaningful indicator of a successful outcome from the treatment of cocaine use disorders has been hindered by the lack of an association between reductions in cocaine use and a functional outcome measure that is sensitive to change. This R21/33 proposal submitted in response to NIDA Funding Opportunity PA-15-099 (Reductions in Illicit Drug Use and Functional Outcomes), seeks to address this issue by pooling data across 13 clinical trials of treatments for cocaine use disorders and using the dataset to empirically derive a new instrument for measuring functional outcomes that is sensitive to change. This new measure will be used to identify and validate changes in cocaine use that are associated with improvement in functional outcomes. The specific aims of each phase of this proposal are as follows: R21 Phase: 1. Create a large pooled database of more than 2,000 unique cocaine dependent individuals participating in one of 13 clinical trials conducted at Yale and the University of Connecticut. In all trials, measures of cocaine use were collected by both self-report and urine drug screens at least weekly during treatment, as well as during a 6- to 12-month follow-up period (>80% data availability). 2. Conduct an integrative data analysis to empirically derive a new measure of functioning covering a broad range of domains (e.g., mental health, physical health, criminality/legal problems, employment) that is psychometrically sensitive to change. R33 Phase: 1. We will determine the levels of change on the developed functional outcome measure from pre-treatment through follow-up period that correspond to cocaine use trajectories identified through latent growth mixture models. This will establish whether improvements in functioning are associated with `reduction'-based cocaine use patterns. 2. Validate the relationship between cocaine use patterns and functional outcomes using two reserved trials within the pooled data (n=442 and 358, respectively) to replicate the levels of change on functional outcomes corresponding to various cocaine use trajectories. This rich database will allow us to capitalize on advanced psychometric and statistical methods to produce a functional outcome measure that would be of high scientific and clinical utility, as well as provide a level of evidence to support non-abstinence measures of cocaine use as a meaningful outcome for clinical trials.