Spatial Analytic Approaches to Explainingthe Trends and Patterns of Drug Overdose Deaths

摘要

To effectively utilize and interpret spatial analyses, substance use re-searchers, public health practitioners and policy makers should be familiar withsome of the available data analytic technique, each of which comes with advan-tages and drawbacks. In this chapter we first discuss three cluster detection toolsand their associated software applications. We then present a Bayesian hierarchicalapproach, briefly reviewing its theoretical underpinnings, commonly used models,and how inferences may be drawn from a sample-based posterior distribution. Wedemonstrate the use of each approach on a set of substance abuse mortality data,comparing the results across the four tools. Our empiric illustration considers therole of neighborhood-level socioeconomic status (SES) in explaining opiate-relatedoverdose deaths in New York City. We end with a discussion of the implications ofthe choice of technique and software on interpreting spatial analyses of substanceabuse and conclude that the choice of a method will be driven by the question tobe answered, data and software availability and the intended audience or context inwhich the research is being conducted.