Using social indicators to estimate county-level substance use intervention and treatment needs

This study describes the development of a model for estimating county-level substance use intervention and treatment needs using social indicators. Fo rty-five indicators conceptually related to substance misuse were reduced t hrough factor analysis. Logistic regression models were then developed usin g selected indicators to predict service needs. The following variables wer e significant predictors of substance use intervention and treatment needs: the percent of the population who is male aged 15 to 34, urbanicity, and p opulation density. The results suggest that substance misuse may be success fully modeled by a few easy to obtain and reliably measured variables descr ibing the population characteristics of local communities.