Population Analytics
We are public health scientists funded by the CDC to utilize data from the Tennessee Department of Health to provide targeted information to support prevention, education, and intervention locally and statewide to reduce opioid-related morbidity and mortality in Tennessee.
We apply best practices in public health and population science methodology (epidemiologic methods, statistics, and data management science) to support providing accurate and timely numbers and statistics.
These methods underlie the dashboards and data briefs and other reports and publications.
Data support data briefs, reports, dashboards, abstracts, presentations, and publications.
We have several ongoing analyses to address specific questions of interest. See below for more details.
Selected Ongoing CDC-funded work by Public Health Scientists on the Prescription Drug Overdose Team:
Project Title: Prescription Opioid Use during Pregnancy and Infant Birth Outcomes in Tennessee
Opioid use among women of reproductive age is on the rise in the US, with some of the highest use observed in southeastern states and among women of lower socioeconomic status. Exposure to opioids in utero is associated with adverse birth outcomes among infants including preterm birth, low birthweight, and neonatal abstinence syndrome (NAS), which can result in short and long-term health consequences and costly health expenditures. In 2016, 1,002 infants were diagnosed with neonatal abstinence syndrome in Tennessee. In 2014, the overall preterm birth prevalence was 10.8% in Tennessee, higher than the national average. We are conducting a comprehensive evaluation of the role prescription opioids and related risk factors in association birth outcomes using statewide linked Tennessee Department of Health data (TN statewide Birth Data, TN Hospital Discharge Data, and the TN Controlled Substances Monitoring Database). Results can inform prevention strategies to reduce infant and child morbidity due to the opioid epidemic.
PI/lead analyst: Sarah Nechuta (sarah.nechuta@tn.gov)
Funding source: Prevention for States Grant, 5 NU17CE002731-02-00
Project Title: A Predictive Model for Injury as a Gateway to Long-Term Opioid Use
Long-term opioid use is associated with increased medication-related side effects, the development of drug dependence and abuse, overdose, and, in some people, progression to unprescribed and illicit opioids. Like car crashes and athletic injuries, injuries at work may be a gateway to long-term opioid use. Previous research on opioid use by injured workers has relied on Workers’ Compensation insurance records, resulting in underestimation of prevalence and misclassification of long-term use. We are using the Controlled Substances Monitoring Database to comprehensively measure prescribed opioids to injured workers, and develop a predictive model for opioid-naïve workers developing long-term opioid use after injury.
PI/lead analyst: Melissa McPheeters (melissa.mcpheeters@tn.gov)/Zoe Durand (zoe.durand@tn.gov)
Funding source: Prevention for States Grant, 5 NU17CE002731-02-00
Project Title: Utilizing the CSMD and TN statewide data to identify populations at risk of overdose mortality by contributing substances and the role of prescribing patterns in mortality risk
Understanding how prescription opioid use, related prescribing patterns, and other risk factors differ by type of overdose drug death can provide targeted information to support prevention and education efforts to reduce morbidity and mortality in Tennessee. However, these data are lacking for Tennessee. To provide critically needed information, we have linked mortality and PDMP data to conduct analyses on the role of prescription opioid use in drug overdose deaths among TN adults and identify populations at higher risk of death by type of contributing drug(s). Example analyses include: 1) sociodemographic characteristics by type of overdose death, 2) type, dose, and timing of prescriptions filled prior to death by type of overdose death, and 3) characteristics of individuals who died with and without a prescription in the CSMD in the past year. Building on this work, we are designing a large epidemiologic cohort study to enable time-dependent and survival analyses to further identify populations at risk for opioid-related mortality.
PI/lead analyst: Sarah Nechuta (sarah.nechuta@tn.gov) and Melissa McPheeters (Melissa.mcpheeters@tn.gov)
Funding source: Prevention for States Grant, 5 NU17CE002731-02-00
Project Title: Enhanced Methodological Approaches for Understanding Opioid Prescribing, Use, and Overdose Outcomes in Tennessee
Opioid-related fatal and non-fatal overdoses have increased rapidly in Tennessee. Public health stakeholders require accurate indicators of the extent of opioid prescribing, use, and adverse outcomes in order to allocate resources when and where they are most needed. This project aims to provide accurate calculation of important indicators of overdose and risky prescribing practices using data from Tennessee’s vital records, Hospital Discharge Data System, and Controlled Substance Monitoring Database. We are improving the calculation of indicators developed by the CDC and creating state-specific indicators optimized for use with Tennessee data with stringent validation and quality control at all levels. Specific improvements to existing indicators include more sensitive detection of non-fatal opioid overdoses and comprehensive identification of patients who obtain opioids from multiple providers (“doctor shoppers”).
PI/Lead Analyst: Benjamin Tyndall (ben.tyndall@tn.gov)
Funding Source: Prevention for States Grant, 5 NU17CE002731-02-00
Project Title: Geospatial Data Analysis
Our team GIS epidemiologist is Sutapa Mukhopadhyay (Sutapa.Mukhopadhyay@tn.gov). She uses geocoding approaches to improve geographic region data estimates (e.g., county-level indicators on opioid prescribing in Tennessee). She generates maps by geographical areas of interest utilizing PDO team descriptive statistics on prescribing patterns, morbidity (e.g., drug overdose hospitalizations, opioid overdose hospitalizations) and mortality (e.g., drug overdose deaths) to identity populations to target for education and prevention.