Spatio-Temporal Predictive Modeling for Placement of Substance Use Disorder Treatment Facilities in the Midwestern U.S.

The inappropriate use of illegal and prescription drugs is an ongoing public health crisis across the United States and beyond. The demand for treatment services quickly outstrips the available supply, limiting access to care. Thus, a data-driven approach to assessing where new treatment facilities are to be built is an essential way to ensure new investments are strategically and optimally located. In this exploratory research, we report the findings of using 24 different public data sets to create three index variables used within a spatio-temporal modeling approach to predict what urban, suburban, and rural counties would most benefit from new substance use disorder treatments across the state of Indiana in the United States. Finally, we discuss the importance and potential limitations of taking this type of approach to develop policies that address complex societal issues.