Prediction of relapse among heroin users treated in Spanish therapeutic communities. A comparison of different models.

The roles of selected heroin use "relapse" theoretical models were studied by analyzing the composition of each predictive model, comparing their predictive capacity on relapse and abstinence, and integrating them into an overall model. This study is based on two groups of heroin misusers (one "relapsed," n = 130; one "abstinent," n = 130) treated in 15 Spanish therapeutic communities. Data were collected by using a structured interview and urine analysis. A logistic regression model was used to build up and cross-validate six different predictive models. This study resulted in several partial models with a predictive capability ranging between 47 and 78% of abstinence and relapses, while the overall model reached 85 and 89%, respectively. The authors conclude that it is advisable to integrate several models in order to prevent relapse more effectively: motivational-conditioned, situational, cognitive-affective, and coping skills models.