Citation

Irvine MA, Bardwell S, Williams S, Liu L, Ge W, Kinniburgh B, Coombs D, Buxton JA. Int. J. Drug Policy 2024; 128: e104454.

Copyright

(Copyright © 2024, Elsevier Publishing)

DOI

10.1016/j.drugpo.2024.104454

PMID

38788389

Abstract

BACKGROUND: British Columbia (BC) Canada has a large take-home naloxone (THN) program, implemented as part of the provincial response to the ongoing toxic unregulated drug supply emergency. Ascertaining the rate of use of THN kits is vital to understanding the full impact of the program. However, this is a challenging problem due to under-reporting of kit distribution. This study aims to estimate the total number of THN kits used based on the number of THN kits shipped, the number of THN kits reported as distributed, and the number of THN kits reported as used.

METHODS: We used BC THN shipment and distribution records (February 2015 to August 2023) to inform a simple Bayesian model of naloxone kit distribution and use. A logistic regression term by health region and distribution site type was incorporated to account for variable under-reporting, and a convolution term was incorporated to account for kit distribution.

RESULTS: We find the number of THN kits reported as used, and the number of total THN kits distributed, are largely under-reported. An estimated 1,500 (95 % CrI: 1,430 - 1,590) THN kits per 10,000 BC population were used, of which 288 per 10,000 had been reported as used. Of all the THN kits shipped, the model estimated that 43 % (95 % CrI: 41-45 %) of kits were used. We also found variation in both distribution and use by distribution site type, with kits distributed from overdose prevention sites having the highest rate of use (56 %; 95 % CrI: 53-59 %).

CONCLUSION: Across all sites, kit use is approximately five times higher than has been reported. Our framework can also be applied to other localities where THN programs operate, in order to better estimate the true reach and impact of take home naloxone distribution.

Language: en

Keywords

Bayesian statistics; Mathematical model; Naloxone; Opioid-related overdose; Take-home naloxone; Toxic drug supply emergency