Using synthetic controls to estimate the population-level effects of Ontario’s recently implemented overdose prevention sites and consumption and treatment services

Abstract

BackgroundOntario recently implemented overdose prevention sites and consumption and treatment services (OPS/CTS) to stem the harms of the opioid epidemic. We tested whether operating any site improved local opioid-related health service use and mortality rates.MethodsWe used monthly counts of all opioid-related emergency department (ED) visits, hospitalizations, and deaths between January 2014 and December 2020 for our outcomes. For each public health unit (PHU) that implemented any OPS/CTS, we created a synthetic control as a weighted combination of unexposed PHU. We estimated the population-level effects of operating any site using controlled interrupted time series with segmented regression and adjusted for time varying confounders (i.e. OPS/CTS capacity, naloxone kits distributed, and persons receiving opioid agonist treatment per 100,000 population). We repeated the analysis using a multiple baseline approach to estimate province-wide effects.ResultsBetween 2017 and 2020, nine out of 34 PHU implemented at least one OPS/CTS. ED visit (RR=0.96, 95%CI: 0.92 – 0.99) and hospitalization (RR=0.95, 95%CI: 0.92 – 0.98) trends declined faster among treated units. Improvements in local ED visit rate trends were observed for the majority of treated units. Hospitalization rates declined faster for London (RR=0.97, 95%CI: 0.95 – 0.99) and Niagara (RR=0.95, 95%CI: 0.92 – 0.98); while mortality rates declined faster for Hamilton (RR=0.93, 95%CI: 0.90 – 0.96), Niagara (RR=0.97, 95%CI: 0.94 – 0.99) and Guelph (RR=0.94, 95%CI: 0.88 – 1.00).ConclusionAlthough OPS/CTS are not sufficient to stem the harms of the opioid overdose epidemic on their own, they play a critical role in local harm reduction strategies.