Unit Costs of Needle and Syringe Program Provision: A Global Systematic Review and Cost Extrapolation

Author: Jordan Killion Christopher Magana Javier Cepeda Anh Vo Maricris Hernandez Cassandra Cyr Karen Heskett David P Wilson Peter Vickerman Fern Terris-Prestholt Adriane Wynn Natasha Martin Peter

Theme: Epidemiology & Public Health Research Year: 2022

Needle and syringe programs (NSP) are effective at preventing HIV and HCV among PWID, yet global
coverage is low- in part because governments lack data on cost and cost-effectiveness of NSP in their
countries. We conducted a global systematic review of unit costs of NSP provision, and developed
regression models for extrapolating costs in countries without data.
From January to October 2020, we conducted searches of 11 peer-reviewed literature databases and 5
grey literature sources using both economic and intervention-specific terms, with no geographic, date,
or language restrictions. The outcome of interest was the cost per syringe distributed (converted and
inflated to 2020 USD). A series of linear mixed-effects models were built to assess associations between
the log unit cost per syringe distributed and country-level (per capita GDP, WHO Health Systems Ranking
Index (HSRI), number of syringes distributed per PWID) and program-level (age, number of intervention
components, and inclusion of ancillary services) predictors. Predictive accuracy was examined using a
sequential ‘leave one out’ procedure to predict costs in countries with available data. Using the best
performing model, unit NSP costs were extrapolated for countries without data.
We identified 55 cost per syringe distributed estimates from 14 countries. The majority of estimates
were from high-income countries (n=43), but estimates covered 4 high-income and 10 middle-income
countries. There were no low-income country estimates. From the regression, a substantial portion of
the variability in unit NSP costs were explained by WHO HSRI, inclusion of ancillary services, and
program age. For the best performing model, all 14 country estimates fell within the prediction
intervals; we extrapolated costs to 137 countries.
Our review identified key gaps in NSP costs for low-income settings. Regression models are useful for
estimating costs in countries without data to inform HIV/HCV prevention programming.
Disclosure of Interest Statement:
Nothing to disclose.

Download abstract Download poster