An interrupted time series analysis of fentanyl, naloxone, and opioid-involved deaths in five counties in Eastern Missouri

Abstract

Introduction: Rates of opioid overdose deaths (OOD) have increased since the introduction of illicitly manufactured fentanyl in the U.S. drug supply. Though community-based naloxone distribution efforts have been found to effectively reduce OOD, no studies to date have examined their effects during the fentanyl era, nor in the epicenter of a state's overdose crisis. Thus, the current study assessed the impacts of both fentanyl and of subsequent grant-funded community-based naloxone distribution on OOD across Missouri's St. Louis region.

Methods: Medical examiner data includes individuals who died within one of the five study counties (St. Louis City, St. Louis County, and Franklin, Jefferson, St. Charles counties [combined as "Collar Counties"]) between 2011 and 2022 due to an opioid overdose (N = 6,799). Naloxone distribution date and location data came from Missouri's university-based naloxone distribution team. We conducted a controlled interrupted time series using an autoregressive model via proc ARIMA to examine changes over time in the rate of OOD associated with the introduction of fentanyl (defined to be present in 25% of opioid overdose deaths; June 2015) and the start of naloxone distribution for each location (August 2017, October 2017, and January 2018, respectively).

Results: The introduction of fentanyl into the drug supply of St. Louis City was associated with an increased rate of OOD over time (p < .001). Naloxone distribution in the City was associated with an immediate decrease in OOD (p < .001) followed by a slowed increase in OOD (p < .001). These findings were not replicated in St. Louis County nor the Collar Counties.

Conclusions: The effects of fentanyl and naloxone on OOD varied by region in Eastern Missouri. Specifically, fentanyl and naloxone effects were only found in a high-need urban area with existing harm reduction and street outreach programs, but not in surrounding suburban or rural areas. State-level naloxone distribution decision-makers should prioritize funding harm reduction services in both urban and non-urban settings to help increase targeted naloxone distribution to those actively using drugs. Additionally, interventions tailored for people who engage in solitary drug use - such as safe consumption spaces and overdose detection technologies - should receive increased investment and implementation, particularly in rural areas where harm reduction infrastructure is more scarce.

Keywords: Community-based naloxone distribution; Fentanyl; Harm reduction; Opioid response; Overdose.