Global coverage of interventions to prevent and manage drug-related harms among people who inject drugs: A systematic review

Abstract

Background:

Harm reduction and treatment programs are essential for reducing harms experienced by people who inject drugs (PWID). We aimed to update estimates from a 2017 review of global coverage of needle-syringe exchange programmes (NSPs), opioid agonist treatment (OAT), and other harm reduction services that target PWID (take-home naloxone [THN] programs, supervised consumption facilities, and drug checking services).

Methods:

We conducted a systematic review of available evidence from peer-reviewed and grey literature databases. Programmatic data were collected on the availability of services, as well as the number of sites, people accessing services, and equipment distributed in countries where there is evidence of injecting drug use. National estimates of coverage of OAT (i.e., number of people accessing OAT per 100 PWID) and NSPs (i.e., number of needle-syringes distributed per PWID per year) were generated where available, using the most recent data. Regional and global estimates were derived and compared to the World Health Organization indicators.

Findings:

We included 195 studies and found there were 90 countries implementing OAT (75% of PWID population) and 94 countries implementing NSPs (88% of the global PWID population). Only five countries (2% of the global PWID population) are providing high coverage of both services. Far fewer countries were found to be implementing THN programs (n=43), supervised consumption facilities (n=17), and drug checking services (n=26), with nine countries implementing all five services. Globally, we estimated there were 18 (95% uncertainty intervals [UI]=12–27) people accessing OAT per 100 PWID, and 35 (UI=24–52) needle-syringes being distributed per person who injects drugs per year. More countries reported high (OAT 24; NSP 10), moderate (OAT 8; NSP 15), and low (OAT 38; NSP 47) coverage of services compared to the previous review.

Interpretation:

Global coverage of OAT and NSPs has increased modestly in the past five years but remains low for most countries. Programmatic data on other key harm reduction interventions are scarce.

Introduction

People who inject drugs (PWID) experience increased risk of preventable morbidity and premature mortality compared to the general population.^1,2 It is recognised and recommended by multiple agencies (e.g., World Health Organization, United Nations) that a comprehensive combination of harm reduction interventions is required to reduce drug-related harm among PWID, and in the population, by addressing personal and structural risks and risk environments.³ There is strong evidence that opioid agonist treatment (OAT) for opioid dependence is effective in reducing blood borne virus transmission, mortality risk, and improving a wide range of other health outcomes.^4–6 Needle-syringe exchange programmes (NSPs) have been demonstrated to reduce HIV and potentially HCV transmission; ^5,6 other interventions that may reduce risk behaviour or improve structural factors include community distribution of naloxone, supervised consumption facilities, and drug checking services.⁷ Monitoring coverage of these, and other, interventions is critical for service planning and setting appropriate public health policies in response to drug-related harm.

In our updated series of systematic reviews, we have estimated that there are approximately 14.8 (10.0–21.7) million PWID globally, and that there was evidence of injecting in 190 countries.⁸ Our previous review, published in 2017, concluded that OAT and NSP coverage was poor in majority of these countries.⁹ Since publication, there have been disruptions to funding and capabilities of harm reduction and treatment services. For example, there is evidence of service closures, limited operating hours, and reduced client capacity due to the COVID-19 pandemic.¹⁰ Also, it is imperative to broaden the focus of the previous review beyond the prevention and treatment of blood borne viruses to include harm reduction services that target other drug-related harm.

The aim of this review is to identify current availability and estimate country-, regional-, and global-level coverage of OAT, NSPs, supervised consumption facilities, take-home naloxone (THN) programmes, and drug checking services for PWID.

Methods

This systematic review was conducted in accordance with the PRISMA and GATHER guidelines (Appendices 1-2). The review protocol was registered on PROSPERO (CRD42020173974).

Eligibility criteria

Studies were eligible for inclusion if they provided information on either: the availability of OAT, NSP, supervised consumption facilities, THN programs, and/or drug-checking services; or an indicator of service coverage (i.e., the number of people accessing these services, the number of visits these services received in any given timeframe, the number of service sites available, or the number of needle-syringes or THN kits distributed by the service). There were no language restrictions.

Data search, screen, and extraction

The multi-stage search strategy was adapted from previous systematic reviews of the global coverage of interventions for PWID and details are provided in Appendix 3.^9,11 The current strategy included searches of the peer-reviewed and grey literature to identify relevant data sources and consultation with international experts to seek additional relevant data. Three peer-reviewed databases were searched: Medline, PsycINFO and EMBASE. Systematic reviews identified with potentially relevant sources were hand-searched for relevant papers/reports.

Searches were conducted in June 2021 and then updated in May 2022 and limited results to publication from January 2017 (the previous review was conducted in 2017). Websites that provided relevant information in the previous review were searched in the current review for reports published after January 2017.⁹ These included reports published on international agency websites such as UNAIDS, UNODC, the Global Fund, and the European Monitoring Centre on Drugs and Drug Addiction (EMCDDA). Website-specific and Google advanced search functions were used to refine search results. In addition, international experts were contacted for publications, reports, or other data via email and social media posts. Our team consulted members of the WHO, UNODC, and UNAIDS for relevant data sources and to ensure best available estimates were considered for inclusion.

All identified documents were catalogued in an Endnote (version X.9) library for initial de-duplication of publications. Peer-reviewed journal articles were imported into Covidence for screening. All peer-reviewed and grey literature reports were screened by two independent reviewers for inclusion at the title and abstract stage (if available), and again at the full-text screening stage. Data sources that were identified after the formal search stage were reviewed in full for eligibility by one team member, and studies to be included were confirmed by a senior team member. Data sources that were not in English were translated through Google Translate if a lingual-specific team member or contact could not be sourced. Data selection process and criteria are documented in Appendix 4.

A database to extract information was built in Microsoft Access. Data were grouped by country of report and year of information. Countries and regional groupings were based upon those used by UNAIDS, WHO and UNODC. Information on service availability, site number, number of people accessing the service, and equipment distribution (NSPs and THN programs) were extracted by one team member and double checked by a second team member; disagreements were resolved by the full team. If national data were unavailable for an indicator, sub-national data were extracted where available. In the current review, all time-series information was extracted up to the year 2000. Where two estimates were provided for the same year, data quality was assessed through team and expert consultation (where necessary) to ensure the most accurate estimate was chosen (further detail provided in Appendix 5). We collected data on service provision reported by gender, where available.

Data synthesis

We collected data that identified availability of services (i.e. any implementation of NSP, OAT, supervised consumption facilities, THN programs, and drug checking services), including restrictions to access for PWID. For all services, each country was then categorised into one of the following: i) service available, programmatic data available (i.e., site and/or client data); ii) service available, programmatic data unavailable; iii) service unavailable; iv) service availability unknown.

The WHO/UNODC/UNAIDS Technical Guide for setting indicative targets for universal access to HIV prevention for PWID informed our interpretation of service coverage.³ Targets for low, moderate, and high coverage of OAT and NSPs, as well as descriptors for THN, supervised consumption facilities, and drug checking services, are indicated in Appendix 6. The methodology employed to estimate NSP and OAT coverage is consistent with the previous review and outlined in Appendix 7.⁹

In generating regional and global estimates of NSP coverage, the most recent, total number of needle-syringes distributed was the numerator, and the total number of PWID estimated for that region (where injecting was known to occur) was the denominator. For countries known to have NSPs, but with no estimate of needle-syringes distributed, the regional level of coverage was assumed.

We calculated OAT coverage with reference to two denominators: 1) the number of PWID, and 2) the number of PWID whose main drug injected is opioids. We assumed the same level of coverage as in other countries in the region for countries where OAT is known to be implemented, but no coverage estimate was available. Full detail on the calculation of national population estimates of PWID are provided elsewhere.⁸ Briefly, prevalence estimates of injecting drug use in each country were collated using the highest quality, most recent national estimates. Prevalence point, lower, and upper estimates were then multiplied by the UN 15 −64 adult Population Size estimates from 2021 to generate the number of PWID in each country.¹² Where national estimates were unavailable, regional or global estimates were imputed. To determine regional and global estimates of OAT coverage among PWID primarily injecting opioids, data were sourced from the same review on the epidemiology of injecting drug use.⁸ The review presented pooled estimates of the proportion of PWID who reported opioids as their main drug in a country. To estimate the denominator used in the current analysis, we multiplied the country population prevalence of injecting drug use by the proportion of people who primarily inject opioids. We multiplied this by the country population size, then 95% uncertainty intervals (UI) were estimated using Monte Carlo simulation taking 100 000 draws. We used a binomial distribution because the parameters of interest were proportions. The simulated UIs incorporated the uncertainty of both the injecting drug use prevalence and opioid injecting proportion estimates.

Role of funding source

The funder of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report.

Results

Eligible studies

In total, 177 studies and reports documented relevant information on OAT availability (n=72), NSP availability (n=88), and drug checking services (n=15), supervised consumption facilities (n=7), or THN programs (n=28) (Appendix 8).

Service availability

Table 1 presents harm reduction service availability by region. We identified 90 countries with OAT programmes (among 190 countries with evidence of injecting), equating to approximately 75% of the global population of PWID (Appendix 9). Methadone was available in 85 countries, buprenorphine in 57 countries, and both methadone and buprenorphine in 54 countries. Other OAT types (e.g. slow-release oral morphine, heroin-assisted treatment, naltrexone) were available in 25 countries. Of the remaining countries with evidence of injecting drug use, OAT was unavailable in 99 countries and was unknown in one. NSPs were available in 94 countries (88% of the estimated global PWID population; Table 1), with programmatic data on numbers of needle-syringes distributed available for 87 countries. NSPs were unavailable in 95 countries and availability was unknown in one (Appendix 10).

Table 1.

Evidence of implementation of harm reduction interventions by region

Region	Number of countries with evidence of injecting / total number of countries	Proportion of PWID population	Needle and syringe programmes				Opioid agonist treatment programs				Supervised consumption facilities	Naloxone distribution	Drug checking services
			Countries implementing		Countries with programmatic dataa		Countries implementing		Countries with programmatic datab		Countries implementing	Countries implementing	Countries implementing
	2022	2022	2017	2022	2017	2022	2017	2022	2017	2022	2022	2022	2022
Australasia	2/2	0.8%	2	2	2	2	2	2	2	2	1	2	2
Caribbean	8/15	0.6%	2	2	2	2	1	2	-	2	-	2	-
Central Asia	5/5	1.6%	4	4	4	4	3	3	3	3	-	3	-
East and southeast Asia	17/18	25.8%	10	10	6	8	8	9	8	9	-	4	-
Eastern Europe	17/17	15.4%	17	17	17	17	16	16	16	16	1	8	3
Latin America	19/20	4.1%	6	1	1	1	3	3	1	1	1	1	5
Middle East and North Africa	21/22	2.2%	8	9	8	9	7	6	4	5	-	1	-
North America	2/2	22.4%	2	2	2	2	2	2	-	2	2	2	2
Pacific Islands	15/17	0.2%	-	-	-	-	-	-	-	-	-	-	-
South Asia	9/9	11.8%	6	6	4	5	6	6	5	6	-	2	-
Sub-Saharan Africa	44/47	8.5%	7	13	7	11	8	12	3	8	-	2	-
Western Europe	31/33	6.7%	29	28	22	26	30	29	26	28	12	15	14
Global	190/207	100.0%	93	94	75	87	85	90	68	82	17	43	26

Note. Hong Kong was listed as part of China in the 2017 review but is listed as a separate country in this review.

PWID: people who inject drugs

^aNumber of needle-syringes distributed

^bNumber of people accessing the service.

Few countries had implemented supervised consumption facilities, THN programs, and drug checking services (Table 1 & Appendix 11). There was very sparse client or distribution data available for these three services. Evidence of supervised consumption facilities was observed in 17 countries (28% of PWID population) across five regions (in Australasia, Eastern Europe, Latin America, North America, and Western Europe), THN programs in 43 countries (71% of global PWID population) across all regions except Pacific Islands, and drug checking services in 26 countries (34% of global PWID population) across Australasia, Latin America, North America, and Western Europe (Appendices 11 & 12). It is important to note that most drug checking services were located at festivals or on short-term trials so the actual number of PWID accessing these services would be limited. There were nine countries with evidence of all five harm reduction interventions (Australia, Mexico, Canada, United States, France, Germany, Portugal, Spain, and Switzerland).

OAT and NSP coverage (WHO/UNODC/UNAIDS indicators)

Table 2 displays the estimated number of people accessing OAT per 100 PWID (with estimated UI) and the estimated number of needle-syringes distributed per PWID per year (UI) by region and globally. We estimated that there were 2.7 million people receiving OAT globally, equating to approximately 18 (12–27) people receiving OAT per 100 PWID and 22 (21–23) per 100 PWID primarily injecting opioids. Overall, 518 million needle-syringes were distributed through NSP sites within 12 months, equating to approximately 35 (24–52) needle-syringes distributed per PWID per year globally. For both interventions, this is classified as low coverage.

Table 2.

Regional and global estimates of coverage of opioid agonist treatment and needle syringe exchange programmes

Region	Estimated number of PWID (UI)	Opioid agonist treatment			Needle-syringe exchange programs
		Clients per 100 PWID (UI)	WHO Coverage	Clients per 100 PWID primarily injecting opioids (UI)	Needle-syringes distributed per PWID per year (UI)	WHO Coverage
Australasia	121,276 (85,888–155,020)	48.5 (38.0–68.5)	High	78.4 (74.4–82.8)	436.6 (341.6–616.5)	High
Caribbean	95,826 (65,264–140,501)	5.5 (3.8–8.1)	Low	6.7 (6.5–7.1)	1.9 (1.3–2.7)	Low
Central Asia	241,025 (161,988–338,607)	1.0 (0.7–1.5)	Low	1.2 (1.1–1.2)	122.7 (87.3–182.5)	Moderate
East and Southeast Asia	3,820,526 (2,945,400–4,701,235)	9.6 (7.8–12.4)	Low	9.4 (7.6–12.3)	20.8 (16.9–26.9)	Low
Eastern Europe	2,282,673 (1,634,524–3,031,379)	1.5 (1.2–2.1)	Low	1.8 (1.7–2.1)	20.3 (15.3–28.3)	Low
Latin America	605,859 (347,180–873,403)	2.6 (1.8–4.6)	Low	1.9 (1.8–2.1)	1.0 (0.7–1.7)	Low
Middle East and North Africa	319,929 (105,825–486,105)	4.7 (3.1–14.2)	Low	4.4 (4.4–4.5)	1.8 (1.2–5.4)	Low
North America	3,315,773 (1,746,780–6,369,251)	20.8 (10.8–39.5)	Moderate	27.8 (26.5–29.5)	41.5 (21.6–78.7)	Low
South Asia	1,749,118 (1,540,456–1,960,093)	44.7 (39.9–50.7)	High	50.4 (47.8–54.1)	29.4 (26.3–33.4)	Low
Sub-Saharan Africa	1,258,728 (662,344–2,238,422)	1.2 (0.7–2.3)	Low	1.2 (1.2–1.2)	4.7 (2.7–9.0)	Low
Western Europe	991,138 (708,010–1,332,357)	69.5 (52.1–98.1)	High	89.4 (80.7–104.3)	115.3 (85.8–161.4)	Moderate
Global	14,825,158 (10,019,592–21,660,383)	18.0 (12.4–26.7)	Low	21.5 (20.6–22.8)	35.0 (23.9–51.8)	Low

Note. Pacific Island states and territories is not included in this table as there was no evidence of either program in the region. UI: uncertainty intervals. WHO: World Health Organization. PWID: people who inject drugs.

^aCoverage indicated by people accessing opioid agonist treatment per 100 PWID (low: <20 people; moderate: 20–39 people; high: ≥40 people)

^bCoverage indicated by number of needle-syringes distributed per person who injects drugs per year (low: <100 needles; moderate: 100–199 needles; high: ≥200 needles)

At the regional level, OAT coverage was high in Western Europe, Australasia, and South Asia (70 [52–97], 49 [38–69], and 45 [40–51] people accessing OAT per 100 PWID, respectively). North America had moderate OAT coverage (21 [11–40] people per 100 PWID), and the remaining regions had low coverage. The Pacific Islands (no OAT available), Central Asia (1 [1–2] people per 100 PWID), and Sub-Saharan Africa (1 [1–2] people per 100 PWID) had the lowest OAT coverage. Australasia was the only region with high coverage of both OAT and NSP (437 [342–617] needle-syringes distributed per PWID per annum). There was moderate NSP coverage in Western Europe and Central Asia with approximately 115 (86–161) and 123 (87–183) needle-syringes distributed per PWID per year, respectively. The region with the lowest distribution of needle-syringes per PWID was Latin America (1 [1–2]), and there were no NSPs identified in the Pacific Islands.

National data on the number of people receiving OAT were available for 82 countries, 70 of which had an estimate of injecting drug use prevalence (Appendix 9). Of the 70 countries with an estimated IDU prevalence, 24 countries had high OAT coverage (with 8% of the PWID population; see Figure 1a), eight countries had moderate OAT coverage (5% of the PWID population), and 38 countries had low OAT coverage (59% of the PWID population). In Afghanistan, Kazakhstan, and Azerbaijan, there were fewer than one person receiving OAT per 100 PWID.

Figure 1. Global coverage of opioid agonist treatment (OAT) and needle-syringe exchange programmes (NSP) among people who inject drugs (PWID).

Of the 94 countries with evidence of NSP implementation, needle-syringe distribution data were available for 87 countries. Population prevalence data were available to estimate NSP coverage in 72 countries (Appendix 10). Of these, only 9 countries (with 2% of the global PWID population) had high NSP coverage (see Figure 1b). There were 15 countries with moderate coverage (with 5% of PWID population), 48 countries with low coverage (with 73% of the PWID population). Australia (508 [401–709]) and Finland (450 [309–504]) had the highest distribution of needle-syringes per PWID. There were 5 countries with both high NSP and OAT coverage (with 2% of the global PWID population), and 3 and 28 countries with moderate and low coverage of both services, respectively (see Figure 2).

Figure 2. Combination coverage of needle-syringe exchange programmes (NSPs) and opioid agonist treatment (OAT) for people who inject drugs.

Figure note. Only includes countries with a non-zero estimate for both NSP and OAT coverage

Comparison to previous review

Since the previous review was published five years ago, OAT provision has now been identified in a further five countries (90 compared to 85 in the previous review), NSP provision increased from 93 to 94 countries, reflecting increases in countries in Sub-Saharan Africa and the Middle East (Table 1). There were 11 additional countries with programmatic data for OAT, five of which were estimated to be low coverage (Figure 3). Similarly, there were 10 countries with new data for needle-syringe distribution (with Cambodia estimated to be moderate coverage) and nine countries (Malaysia, Bangladesh, Mozambique, Sierra Leone, Uganda, Albania, Denmark, Germany, and Serbia) categorised as low coverage. While the global change in coverage was modest between the previous review and this review, there were notable changes in coverage at the national level for many countries. Some changes were due to a genuine change in service provision in country, due to either significant improvements in service coverage (e.g., number of people accessing OAT in Romania) or reductions in service provision (e.g., NSP service closures in countries in Latin America; see Panel A). In other countries, updated estimates of injecting drug use prevalence between reviews have resulted in changes in estimated service coverage despite similar estimates of service delivery (e.g., India and Belgium). Finally, changes to programmatic data recording (including the availability of programmatic data where none previously existed) has contributed to changes observed in some countries (e.g. Scotland and Croatia; see Appendix 13 for percent changes in coverage level for OAT and NSP between reviews).

Figure 3. Proportion of the estimated number of people who inject drugs covered by opioid agonist treatment (OAT) and needle-syringe exchange programs (NSP), by coverage level.

Note. This figure does not include San Marino, where availability of OAT and NSPs was unknown.

There were very few reports documenting gender distribution of people accessing OAT and NSP services (Appendix 14), with no clear difference between estimated gender distribution of PWID compared to gender distribution of people accessing services overall.

Discussion

Global coverage of major harm reduction and treatment interventions for PWID remains low. While there have been clear improvements in data availability and evidence of substantial improvements in coverage in certain countries, most are failing to meet high or even moderate coverage targets set out by WHO, UNODC, and UNAIDS.³ Only five countries, with 2% of the global PWID population, are providing high coverage of both NSP and OAT. Our review also found that few countries had introduced any THN programs, supervised consumption facilities, and drug checking services; only nine countries had evidence of implementing all five interventions. Our findings illustrate global gaps in harm reduction service coverage to inform service planning and identifies countries and regions where increased advocacy is required to improve availability of services.

There has been a modest increase in estimated OAT coverage from 16 (UI=10–24) to 18 (UI=12–27) people per 100 PWID and an improvement in the number of countries with moderate or high coverage (from 26 to 32 countries).⁹ We identified 38 countries that reported more people accessing OAT in this review, and an overall increase of 250,000 people accessing OAT globally. In Russia, where approximately half of the 1.3 million PWID population are living with HIV,8 OAT is prohibited by federal law.¹³ A modelling study demonstrated that without additional interventions, the HIV prevalence would increase among PWID in Russia and OAT coverage would need to be scaled up to 50% to avert 35% of new HIV infections.¹⁴ Although other countries in Eastern Europe have implemented OAT, the relatively large PWID population means service provision is completely insufficient to meet demand (only 1 person accessing treatment per 100 PWID).

More than 518 million needle-syringes were recorded as being distributed annually in the current review, an increase of 20 million since the previous review.⁹ Yet there have been notable regressions in service provision due to changes in harm reduction policy in Latin America. Since the previous review, NSPs in Argentina, Brazil, Colombia, Paraguay, and Uruguay have ceased operation. In much of Latin America, provision of sterile injecting equipment is deemed unnecessary due to the small population of PWID and, it is likely that the more punitive drug policies that have been increasingly adopted have also contributed to service closures.^7,8 In Mexico, the only country in Latin America where NSPs are currently operating, lack of funding has severely limited service provision.¹⁵ Structural barriers and funding strains for NSPs can be further compounded by fear of arrest for drug use or possession of injecting equipment, stigma and discrimination, and lack of trust of government by injecting communities more broadly.^16–20

Supervised consumption facilities, THN programs, and drug checking services are currently inaccessible to most PWID, yet they contribute to mitigating individual- and structural-risk factors that contribute to drug-related harm. The need for scaling up these interventions within the community has been recognised within wider harm reduction strategies.^7,21 Even in countries with high coverage of OAT and NSP services, there are excess drug-related deaths and extensive morbidity associated with injecting drug use.^1,22,23 Supervised consumption facilities, in particular, offer a safe place to inject, sterile equipment, and immediate overdose response, among other benefits.²⁴ Widening the focus of global drug policy beyond blood borne virus transmission to include other critical harms experienced by PWID will likely improve the evidence on these interventions.

Monitoring the availability and coverage of harm reduction services can inform planning, prioritisation, and investment of funding resources to improve service delivery and is critical for managing competing public health priorities. The Global Fund’s Strategy 2017–2022: Investing to End Epidemics describes intensive support for harm reduction services in low- and middle-income countries that is consistent with increased coverage and greater programme data¹⁰ – but also recently noted that, for the first time in their 20 year history, access to HIV prevention and testing services have regressed due to the COVID-19 pandemic.¹⁰ This may be particularly detrimental to development of harm reduction services in Sub-Saharan Africa, where services for people who use drugs are unavailable in most countries or, where available, have very low coverage.

Limitations

There were additional countries with programmatic data located for the first time for OAT and NSPs, and nearly all were considered low coverage. These estimates likely represent shorter term and smaller scale programmes that are now collecting data, reducing regional coverage estimates. Therefore, comparisons to the previous review should be made with caution. Further, some of the changes between this and the previous review reflect changes in the estimates of population sizes of PWID as well as the for the numbers of harm reduction and treatment programs used (outlined in Panel A). However, we used the best and most updated available evidence in this review.

There was an absence of programmatic data from multiple countries. Despite exhaustive reviews of the literature and consultation with UN agencies and other experts in the field, we may have missed some evidence. We encourage ongoing contribution from national experts to improve the estimates provided and ask that those with insights into programme provision contact us with developments in future. We have noted previously that there are problems with estimating the prevalence of injecting drug use with a lack of external information or formal methods validating the estimates.^9,25 Our estimates of NSP and OAT coverage are uncertain, but we lack evidence on uncertainty around these estimates (particularly insofar as the data may not capture all service provision occurring in a country), potentially underestimating the measures of uncertainty once combined with population prevalence. Regional estimation was also calculated by imputing country-level data for six countries where needle-syringe distribution data were not available and eight countries where OAT client data were unavailable, which may not reflect the heterogeneity between countries in the same region. That said, this is unlikely to change the overall picture of low coverage. As an example, Nigeria would need to be distributing over 120 million needles to increase Sub-Saharan Africa from low to moderate NSP coverage.

Finally, national estimates of coverage may not represent sub-national accessibility of services. For example, access to supervised consumption facilities in most countries is restricted to a minority of PWID living in select urban settings.²⁶ Similarly, NSP and OAT coverage may vary geographically within a country. We also classified level coverage of OAT and NSPs according to definitions given in WHO, UNODC, and UNAIDS Technical Guidance and previous global reviews.^9,27 It is likely, however, that other features of coverage that are not yet readily measured in global reviews or incorporated into coverage recommendations are critical to reducing drug-related harm in PWID. For example, a combination of OAT and high-level coverage of NSPs (associated with reducing HCV and HIV transmission) refers to having at least every injection covered by a sterile needle-syringe.^4,5,28 Similarly, epidemiological and model evidence shows that OAT coverage, retention, and availability in prisons are all important for reducing mortality risk.^29–31

Conclusion

Globally, most PWID currently lack access to harm reduction services designed to reduce public health burden and improve quality of life. While the past five years has seen progress in certain countries toward improved programmatic data availability and service coverage, the available evidence suggests upscaling of services is required to respond to the clinical harms that many PWID face. This review has identified major regional disparities in harm reduction service availability and coverage, highlighting priority areas for funding provision. Increasing the amount and quality of evidence on the coverage and impacts of harm reduction interventions on reducing drug-related harm will increase power of advocacy, consequentially bolstering arguments for their provision. Ongoing evaluation of service coverage is also required to navigate resource allocation most efficiently and enhance service accessibility.

Research in context.

Evidence before this study

Global systematic reviews were conducted on the epidemiology of injecting drug use and the coverage of harm reduction, testing, and treatment interventions for people who inject drugs (PWID) in 2017. Harm Reduction International published their report The Global State of Harm Reduction 2020 that documents country-level availability of needle-syringe exchange programmes (NSP), opioid agonist treatment (OAT), drug consumption rooms, and peer distribution of naloxone; however, this report does not estimate the coverage of NSP or OAT among PWID and much of the epidemiological information informing this report was sourced from the previous reviews.

Added value of this study

In the five years since the previous review was published there has been substantial effort by the Global Fund and other public health agencies to improve accessibility of harm reduction programmes. This multi-stage systematic review updates national, regional, and global estimates of NSP and OAT coverage, and expands the focus to include other interventions that reduce drug-related harm. Specifically, this review examines availability of take-home naloxone (THN) programs, supervised consumption facilities, and drug checking services. Estimates of OAT and NSP coverage are crucial for informing the United Nation’s (UN) 2030 Agenda and Sustainable Development Goals on the prevention and treatment of substance abuse, and monitoring implementation and coverage of less focal harm reduction interventions is critical for service planning and setting appropriate public health policies in response to drug-related harm.

Implications of all the available evidence

Coverage of OAT and NSP are low in majority of countries and for majority of PWID. There is indication that coverage is improving; however, it is still insufficient when compared to the indicator targets set out by the World Health Organization (WHO), United Nations Joint Programme on HIV/AIDS (UNAIDS), and United Nations Office on Drugs and Crime (UNODC). This review highlights gaps in harm reduction service coverage to inform service planning and identifies countries and regions where increased advocacy is required to improve availability of services.

Panel A. Exploring reasons for changes in service provision between the previous review and the current review

Change in coverage due to policy/funding changes

Our findings indicate that, in certain countries, there have been both improvements and reductions in service provision that reflect changes in service coverage estimations between reviews.

Romania

Available data sources indicated that in the previous review, there were 547 people accessing OAT and, in the current review, 1772 people (still classified as low coverage). Despite Global Fund support ceasing in 2017, harm reduction activities have been expanding in Romania in alignment with their National Anti-Drug Strategy (2013–2020).

Argentina, Brazil, Colombia, Paraguay, and Uruguay

In the previous review there was evidence that NSPs were operating in these Latin American countries, although there were no data on the number of needle-syringes that had been distributed. Experts in country have advised us that since this time, there have been substantial political changes in the region and that these NSP sites have since ceased operation.

Change in coverage due to changes in estimated population prevalence

The estimated prevalence of injecting drug use in a country is used as a denominator to inform the estimated coverage of NSP and OAT services. Substantial changes in the estimate population prevalence can result in substantial changes in estimates of coverage, whether the service provision has changed or not.

India

While the estimated number of needle-syringes distributed nearly doubled between reviews (from 16,917,292 to 30,200,000 needle-syringes), a more recent estimate of the prevalence of injecting drug use was located that was higher than the older estimate (i.e., prevalence of 0.02 [0.01–0.03] in the previous review, and 0.09 [0.09–0.10] in the updated review). It Is difficult to know what is driving the difference (e.g., change in methodological approach or change in the population size); however, this has resulted in estimated coverage of NSPs to drop from 86 (63–133) to 34 (31–38) needle-syringes distributed per person who injects drugs per annum.

Belgium

In the previous review, the estimated population prevalence of injecting drug use was 0.35 (0.24–0.49); however, in this review the estimated prevalence was 0.10 (0.07–0.13). Thus, while there was a modest increase in needle-syringes distributed between reviews (approximately 229,000 more needle-syringes being distributed), the change in the population prevalence resulted in substantial changes in coverage. Previously, we estimated that there were 40 (28–57) needle-syringes distributed per person per annum, while currently we estimate that there are 179 (132–262) needle-syringes being distributed per person per annum.

Change in coverage due to changes in programmatic data availability

Scotland

The estimated number of OAT clients recorded in the previous review (N=3,686) reflected the number of initial assessments in 2012 measured at a single point in time and not the total number of patients receiving OAT in that year. Public Health Scotland has since released the estimated number of OAT clients from 2011 to 2021 (approximately 29,416 people accessing OAT in 2021). Although it is noted that these are likely underestimates of the true population of people accessing OAT, their reporting finds annual client numbers ranged from 27,475 in 2012/13 to 29,729 in 2017/18.

Croatia

In the 2017 European Drug Report published by the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA), the number of needle-syringes distributed in Croatia was recorded as 923,650 and consequently led to an estimate of 146 (112–179) needle-syringes distributed per person who injects drugs per year in the previous review. Since publication of this report, the EMCDDA have updated their annual estimates indicating that distribution equated to approximately 234,000 needle-syringes in 2014 and 341,900 in 2019 (the most recent year of estimate).