Addiction severity predictions using client network properties

By statistical analysis of client data it is shown how past or current network information together with other knowledge assesses treatment needs. The main findings are as follows. The client's previous exposure to addicts in the family has almost no influence on his or her present contacts with addicts in daily life. About 30% of the clients have experienced addicts both in family and among friends, about 30% have only family exposure, about 20% have neither kind of exposure, and about 20% have no family exposure but have current exposure to addicts. Exposure to addicts in family implies a higher risk of severe need for professional intervention than no such exposure. For the clients with no family exposure but with current exposure to addicts there is, somewhat surprisingly, a lower risk than for the other three categories of clients. This effect seems to be mainly because the psychiatric status of these clients is somewhat better than for those in the other three categories. The effect disappears if we look solely for drug addiction intervention needs; then there is a clear increase in relative treatment needs for the categories with previous or present addiction exposure compared to those without.