Optimizing placement of public-access naloxone kits using geospatial analytics: a modelling study.

Background: More than 85 000 people die annually across North America from opioid poisoning; naloxone in the hands of the public is an effective intervention and saves lives. We compared the accessibility of different placement strategies for public-access naloxone kits.

Methods: We evaluated all opioid-poisoning incidents recorded by BC Emergency Health Services between December 2014 and August 2020 in Metro Vancouver, Canada. We determined the number of opioid poisonings "covered" (i.e., within a 3-minute walk) by 3 different coverage strategies: (1) existing locations participating in take-home naloxone programs; (2) blanket naloxone kit placement at chain businesses, pharmacies, and registered public-defibrillator locations; and (3) optimization-based strategic kit placement at transit stops based on historical poisonings.

Results: We included 14 089 opioid poisonings. Existing locations participating in take-home naloxone programs (647 locations) covered 4988 (35.4%) opioid poisonings. Chain businesses (10-233 locations) covered 6 (0.0%) to 1165 (8.3%) opioid poisonings, and chain business categories (12-810 locations), pharmacies (790 locations), and public-defibrillator locations (980 locations) covered 97 (0.7%) to 3152 (22.4%) opioid poisonings. Optimization-based strategic placement of naloxone kits at transit stops yielded generally higher coverage levels, ranging from 2907 (20.6%) opioid poisonings covered with 10 kit locations, to 7506 (53.3%) with 1000 kit locations.

Interpretation: Optimized placement of publicly accessible naloxone kits at transit locations was most effective at improving public accessibility of naloxone, and blanket placement at take-home naloxone program locations covered a substantial proportion of opioid poisonings. Public-access naloxone may improve community access to naloxone in response to opioid poisonings.