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. Forty-five indicators conceptually related to substance misuse were reduced through factor analysis. Logistic regression models were then developed using selected indicators to predict service needs. The following variables were significant predictors of substance use intervention and treatment needs: the percent of the population who is male aged 15 to 34, urbanicity, and population density. The results suggest that substance misuse may be successfully modeled by a few easy to obtain and reliably measured variables describing the population characteristics of local communities.