Spatio-temporal Methods for Surveillance of the Opioid Syndemic
The opioid epidemic has been an ongoing threat to public health for over 20 years and has recently resulted in a record number of over 100,000 overdose deaths per year in the United States. However, the impacts of the opioid epidemic are not limited to overdoses as it also forms a syndemic with HIV and HCV. Syndemic theory recognizes that health conditions occur and interact in specific social, temporal and geographic contexts. Opioid misuse is the key underlying driver of the syndemic and an effective response needs to address this population, but quantifying opioid misuse is extremely challenging, particularly at local levels. This project, supported by a NIDA R01, focuses on the development of novel multivariate spatio-temporal models within the Bayesian paradigm that leverage multiple sources of existing surveillance data at varying levels of spatio-temporal support. As part of the project, the team has developed methods to uncover the complex dependencies and synergy that drive the syndemic and adapted methods from statistical ecology that enable indirect estimation of the prevalence of opioid misuse at the county level. These methods have been applied to analyses in Ohio, New York and North Carolina. Future work is informed by conversations with an advisory board that consists of state and federal public health officials.
Team members: David Kline (MPI), Staci Hepler (MPI), Brian White (BDS) and Chenhui Qui
Other team members: Lance Waller (Emory), Bill Miller (UNC), Andrea Bonny (Ohio St) and Erin McKnight (Ohio St)
