Geographic and Temporal Trends in Fentanyl-Involved Overdose Deaths in Connecticut, 2009-2019

• Activity Number: 289 - Contributed Poster Presentations: Section on Statistics in Epidemiology
• Type: Contributed
• Date/Time: Tuesday, August 9, 2022 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Haidong Lu* and Forrest W. Crawford and Gregg Gonsalves and Thomas Thornhill and Lauretta Grau
• Companies: Yale University and Yale University and Yale University and Yale University and Yale University
• Keywords: Substance uses ; Bayesian statistics; Spatiotemporal analyses; Drug overdoses; Opioid epidemic

Abstract

Since 2012 fentanyl-related overdoses have risen from 4% of all overdoses in the state of CT to 82% of these deaths in 2019. Here we aim to investigate the geographic and temporal trends in fentanyl-involved overdose deaths in Connecticut during 2009-2019. We obtained data on the dates and locations of opioid overdose deaths that occurred during 2009-2019 from Connecticut Office of the Chief Medical Examiner. Using a Bayesian space-time regression model adjusted for demographic covariates, we estimated the spatial and temporal trends of fentanyl-involved overdose death. As a result, during 2009-2019, 6,632 opioid-involved overdose deaths were identified. Among these, 3,234 (49%) were fentanyl-involved, and 3,398 (51%) were non-fentanyl fatal overdoses. Significant spatial and temporal variation in the probability of fentanyl involvement given an opioid overdose death was identified from Bayesian space-time model. The spatial patterns suggest opioid-involved deaths in the north-eastern region of Connecticut had higher probability being fentanyl-involved. Our findings suggest that geographic variation exists in the probability being fentanyl-involved given an opioid overdose death.