Spatio-temporal Methods for Surveillance of the Opioid Syndemic

Wake Forest University Health Sciences, Winston-Salem NC

Investigators

Linked publications, trials & patents

Abstract

Project Summary/Abstract In 2019, the United States Department of Health and Human Services launched the Ending the HIV Epidemic (EHE) initiative to reduce incident infections by 75% within 5 years and 90% within 10 years. A priority population in the National HIV/AIDS Strategy is people who inject drugs. Concurrently, the overdose crisis remains a public health emergency. While the epidemiology of substance use is constantly evolving, the misuse of opioids, particularly synthetic opioids like fentanyl, continues to drive the crisis. The overdose crisis is part of a syndemic with HIV, hepatitis C (HCV), and opioid use disorder (OUD). North Carolina has been significantly impacted by the opioid syndemic with rates of HIV infections unabated from 2012-2022 and overdose death rates well above the national average. Understanding the opioid syndemic is of public health importance, but this is challenging because no single data source currently observed by the public health surveillance system fully characterizes opioid misuse at relevant spatial and temporal scales. Surveillance data is a valuable resource, but it typically consists of health care encounters for negative health outcomes (e.g. overdose) that only identify the portion of the population who had the event of interest recorded. These data are also increasingly prone to misclassification, or the notion that not everyone captured in the surveillance records belongs to the population who misuse opioids. For example, opioid overdose death counts may include people who use stimulants but overdosed due to fentanyl contamination. When allocating opioid- specific resources, it may not be of interest to include such misclassified individuals in the counts. Novel statistical methods are needed to better leverage existing data and integrate multiple surveillance outcomes while accounting for imperfect detection and misclassification. Additionally, while surveillance data are typically available at the county-level, evaluation of neighborhood accessibility of health services requires estimates at the sub-county level. Novel integration of electronic health record (EHR) data with surveillance data enables high resolution small area estimates of opioid misuse prevalence. There are several methodological challenges that will be overcome with achievement of the following aims: 1) Develop an integrated abundance model that accounts for misclassification in surveillance outcomes to estimate opioid misuse prevalence; 2) Develop methods for small area estimation of opioid misuse prevalence using data at multiple spatial scales including county-level surveillance data and address-level data from the EHR; 3) Assess the current allocation of health services in North Carolina neighborhoods to identify gaps and guide allocation of additional resources. Successful completion of these aims will enhance surveillance of the opioid syndemic by producing accurate small area estimates of opioid misuse prevalence that advance epidemiological understanding of the syndemic and guide public health planning and resource allocation to prevent HIV, HCV, and overdose.