Informing health-friendly drug policy in Africa: A comment on the challenge of drug consumption data

Csete, Joanne Article Informing health-friendly drug policy in Africa: A comment on the challenge of drug consumption data Journal of Illicit Economies and Development (JIED) Provided in Cooperation with: LSE Press, London School of Economics and Political Science (LSE) Suggested Citation: Csete, Joanne (2024) : Informing health-friendly drug policy in Africa: A comment on the challenge of drug consumption data, Journal of Illicit Economies and Development (JIED), ISSN 2516-7227, LSE Press, London, Vol. 5, Iss. 3, pp. 27-36, https://doi.org/10.31389/jied.158 This Version is available at: https://hdl.handle.net/10419/320684 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ POLICY COMMENTARY CORRESPONDING AUTHOR: Joanne Csete Columbia University Mailman School of Public Health, US [email protected] KEYWORDS: Africa; African Union; drug consumption; data; bio-behavioural surveys TO CITE THIS ARTICLE: Csete, J. 2024. Informing Health-friendly Drug Policy in Africa: A Comment on the Challenge of Drug Consumption Data. Journal of Illicit Economies and Development, 5(3): pp. 27–36. DOI: https://doi.org/10.31389/ jied.158 Informing Health-friendly Drug Policy in Africa: A Comment on the Challenge of Drug Consumption Data JOANNE CSETE ABSTRACT Considerable resources have been spent on estimating the size of populations of people who use drugs (PWUD) and people who inject drugs (PWID) in Africa. Precise estimates are elusive, not least because of stigma and criminalization faced by these populations. Bio-behavioral surveys focused on injection drug use can be useful but are expensive and have not always been designed and implemented with meaningful participation of PWID themselves. A pan-African effort of the African Union (AU) to collect drug-use data from drug treatment facilities is undermined by the dearth of facilities addressing the drug problems posing the greatest morbidity and mortality risks. Efforts that have involved PWID meaningfully and respectfully in designing and implementing data collection have shown some success in informing health service programming for PWID as well as sustained regular monitoring of drug use. But meaningful participation in data collection may be contingent on both concerted anti-stigma measures and some form of decriminalization of minor drug offenses to reduce the fear of arrest and incarceration experienced by people who use drugs, a change as yet little realized in Africa. 28Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 INTRODUCTION In high-level United Nations resolutions in 2016 and 2019, member states unanimously espoused ‘evidence-based’ policies to address the ‘world drug problem’ that would be consistent with human rights standards and with ’promoting and protecting the health, safety and well-being’ of everyone (UN General Assembly 2016; UN Commission on Narcotic Drugs 2019). At the General Assembly Special Session on drugs in 2016, numerous countries declared their intent to treat drug consumption as a challenge for health and social policy rather than principally as a law enforcement matter (Csete & Wolfe 2017). While punitive prohibitionist drug policies have been the norm in much of Africa, some African countries have begun to join the call for more health-focused policies on drug use (Eligh 2019). The continent is also seeing some growth of specialized health services for people who use drugs, including HIV-related harm reduction services (Harm Reduction International 2021). Africa still lags behind other regions in this regard in spite of being the most HIV-affected part of the world. The AU Plan of Action on Drug Control and Crime Prevention for 2019–2023 urges AU member states to adopt drug policy that includes comprehensive harm reduction programs (African Union 2019a). A commitment to ensuring good health practices in drug-control policy in Africa raises, among other issues, the question of exactly what evidence would inform the design and implementation of evidence-based, rights-friendly, public health-friendly drug policies on the continent. What data are most needed to inform policy that aims to promote and preserve health, and what are the constraints to collecting these data? How especially will data collection on drug consumption and the health needs of people who consume drugs be collected in places where even minor drug infractions are criminalized and highly stigmatized? This article seeks to comment on some existing data sources on drug consumption and drug-related health problems with an eye to identifying feasible ways to generate the evidence needed for an era of health-friendly drug policy and programs in Africa. A full review of drug-related data in Africa is beyond the scope of this comment, which rather considers mostly data-related strategies of the AU and other African multilateral bodies. WHAT IS RIGHTS-BASED, HEALTH-FRIENDLY DRUG POLICY? To guide thinking about evidence to inform drug policy, it is worth noting how both member state UN resolutions and UN technical agencies define health-friendly drug policy. The General Assembly resolution of 2016 suggests that drug policy consistent with the right to health would facilitate (1) services to minimize the harms of controlled drugs and medicines; (2) a full range of prevention and treatment services for drug use disorders; and (3) health services for people in prison and other state custody (UN General Assembly 2016). In more specific terms, the World Health Organization (WHO) endorses a ‘comprehensive package’ of services for people who inject drugs, which includes needle and syringe programs, opioid substitution therapy, and prevention and treatment services for viral hepatitis, HIV and tuberculosis (WHO 2022). WHO and the UN Office on Drugs and Crime (2020) have also established standards for evidencebased treatment of drug use disorders (DUD). These include eliminating abusive, unethical and scientifically unsound practices, which are important elements as drug treatment is highly unregulated in many countries. The agencies also note that the ’criminal justice system should collaborate closely with the health and social systems to ensure that treatment for drug use disorders in the health care system takes precedence over criminal prosecution or imprisonment’ (ibid.: 10). Guidance for gender-appropriate treatment is also included. While many of the standards of public health-friendly drug policy were inspired by HIVand hepatitis-related harm reduction, other health concerns should also figure in drug policymaking. Drug-related mortality from overdose is a growing problem as fentanyl and its analogs have found thriving markets worldwide (World Drug Report 2021: booklet 3). The 2016 General Assembly resolution also emphasizes the need to ensure that all people have access to controlled medicines ’for medical and scientific purposes, including the relief of pain and suffering’ (UN General Assembly 2016). In most low-income countries, including in sub-Saharan Africa, access to morphine for cancer and end-of-life pain relief is dramatically inadequate (Ju et al. 2022). 29Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 EXISTING DATA ON DRUG CONSUMPTION The World Drug Report (WDR) includes estimates of the number of people country-by-country who used drugs in a recent year and those who suffered from drug use disorders (DUD), defined as use that is ’harmful to the point where [people who use] may experience drug dependence and/or require treatment’ (WDR 2021: booklet 2). The 2021 report estimated that in 2018 some 269 million people in the world had used drugs in the previous year, with a confidence interval of 166 to 373 million (ibid.). The estimate for Africa was about 60 million (CI: 35–81 million). The global estimate for people living with DUD was 36.5 million (CI: 19.6–53.0 million). There was no DUD estimate for Africa, but the estimated number of persons who inject drugs on the continent, including North Africa, was 950,000 (CI: 590,000–1,760,000). The 2021 WDR predicted drug use trends, emphasizing that the greatest regional growth in drug use from 2020 to 2030 was likely to be in Africa, based largely on the projected rate of general population growth (ibid.). The report notes that the most rapid increases since 2010 in Africa were in consumption of cannabis and opioids, including heroin and tramadol (especially in West and North Africa). The United Nations Office on Drugs and Crime (UNODC) collects data on drug use through responses to the annual report questionnaire (ARQ) that UN member states agreed to submit each year. Of 54 countries in Africa, however, only 10 submitted data – in six cases only partially – for the section of the ARQ used to estimate the prevalence of drug use for the 2021 report (WDR statistical annex 2021). For other countries, UNODC either estimated prevalence from earlier data or had no estimate. The AU drug-control plan for 2019–2023 includes as an objective to increase the number of member states submitting the ARQ each year (African Union 2019a). Under the best of circumstances, arriving at national estimates of people who use drugs is challenging, as the large confidence intervals around the WDR figures attest. The 2021 WDR notes that national-level general population surveys, usually government-managed, are the main source of the report’s estimates, but some countries do not undertake these surveys and there are relatively few of these in sub-Saharan Africa, for example. National surveys also tend to exclude people in prisons and other detention settings, and in residential drug treatment facilities, and they may not to capture people with problematic drug use or who use the most highly criminalized and stigmatized drugs (WDR; booklet 2 2021). Because these surveys rely on self-reported behavior, stigma and criminalization may reasonably be expected to keep significant numbers of people from reporting drug use. In addition, some countries make estimates based on entries in drug registries, which are likely to yield gross underestimates of drug use prevalence. In addition, definitions of drug use are not uniform from country to country, and protocols for surveys—including who is excluded and in what settings they are administered—also differ among countries (WDR 2020). In WDR figures, general population surveys are sometimes complemented by, for example, respondent-driven methods whereby a relatively small number of ’seed’ persons identified as people who inject drugs are asked to identify others who inject drugs, who in turn follow suit and identify still others until an appropriate sample size is achieved (WDR: booklet 2 2021). WDR 2021 also cites ’capture-recapture’ methods, which are used to estimate the completeness of usually more than one routine data source, such as a methadone patient registry, and extrapolate from there. None of these methods is perfect, and all raise concerns about biases and missing what may be the population segment of greatest interest – the one most fearful of being discovered. DATA ON INJECTION DRUG USE Estimating the size of the population of people who inject drugs (PWID)—an important subset of people who use drugs—is of particular interest because of concerns linked to bloodborne diseases such as HIV and hepatitis C as well as overdose injury and mortality. Major international HIV donors, including the US President’s Emergency Program for AIDS Relief (PEPFAR) and the Global Fund to Fight AIDS, TB and Malaria, have invested considerably in surveys to estimate the size of PWID and other key populations and in building capacity to conduct these surveys in lowand middle-income countries (Viswasam et al. 2020). Some NGOs at least perceive that 30Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 PEPFAR bases its funding decisions on size estimates of ‘key populations’ such as PWID and has been too willing to accept low estimates to justify low funding levels (MPact 2021). The quality of data on PWID and their HIV status has been the subject of considerable debate over the years. In 2013, for example, the WDR’s estimate of 1.6 million people globally who inject drugs and live with HIV represented a 46% decline from the figure used by UN agencies at the time (Harm Reduction International 2013). Harm Reduction International (2013) and the International Drug Policy Consortium (2013) asserted that these data, based on ARQ and other sources not publicly available, gave an erroneous impression that HIV prevention efforts among PWID had enjoyed dramatic success and might thus discourage further investment from donors. These advocates criticized the lack of transparency of the process by which these data were developed and the lack of meaningful consultation of affected communities of PWUD in this process. Since 2014, UNODC, UNAIDS, WHO and the World Bank have reportedly worked together with a ’broad group of experts from academia’ and a range of civil society organizations to consider all available data on prevalence of injection drug use (WDR 2020). Estimates derived statistically from what may be several data sources in a given country are sent to the countries for their review before publication. The 2020 WDR notes that for prevalence of injection drug use, 122 countries had information, of which 41% had high-quality information from recent surveys. Of those, almost none were in Africa, which UNODC characterized still as having very limited information on PWID (ibid.). On the matter of HIV prevalence among people who inject drugs, 36% of countries were judged to have high-quality information, almost all of them outside Africa. Addressing controversies in this area, UNODC noted that HIV prevalence estimates should be regarded as the best updates of the most reliable data reported previously, and from these data ’[t]here is no intention to imply that there has been an actual change in the prevalence of injecting drug use or HIV among PWID at the regional or global level’ (ibid.: 33). Collecting data on injection drug use, including population size estimates, has received a boost since about 2000 with the proliferation of bio-behavioral surveillance (BBS) in lowand middleincome countries (sometimes referred to as integrated biological and behavioral surveillance – IBBS.) These surveys, focusing largely on ’hard to count’ HIV key populations such as PWID, sex workers and men who have sex with men (MSM), are designed to assess HIV burden and behavioral risk factors and usually population size as well (WHO et al. 2017). As noted by WHO and other technical agencies, there rarely exist ’ready-made sampling frames’ for these highly stigmatized populations (ibid.: p 1). Thus, BBS investigations have used a variety of sampling strategies, often non-random, depending on formative assessment of such factors as whether the population is well networked among itself or whether members of it can be found in easily identifiable locations (ibid.: p 71). BBS studies have been conducted for various key populations in Angola, the Democratic Republic of Congo, Ethiopia, Ghana, Kenya, Mauritius, Mozambique, Namibia, Nigeria, Senegal, South Africa and Uganda, among others (all available online). WHO, with the US Centers for Disease Control and Prevention (CDC) and others, has established guidelines for BBS and recommends them as integral to planning national HIV strategies (ibid.). Among the elements captured in the most complete BBS surveys on PWID, such as that of Mauritius (Republic of Mauritius 2021), are HIV and hepatitis C prevalence among PWID disaggregated by sex, awareness and utilization of needle exchange services, substances injected and frequently of injection, needle sharing practices, risky sexual practices, and experiences of stigma. The usefulness of this kind of information for programming planning and implementation is obvious. In the case of Mauritius, a small island nation, a population size estimate for PWID was part of BBS and relied on a respondent-driven method. Not all BBS studies of PWID are as complete as that of Mauritius, however, though most include some attempt at population size estimates. The WHO/CDC guidance on BBS is over 200 pages of detailed recommendations on the challenges of finding and counting key populations, which may be hidden, as well as procedures for HIV testing and for asking marginalized and criminalized persons about their risk behaviors (WHO et al. 2017). WHO and CDC recommend strongly that BBS should include population size estimates, but their guidance notes that estimates of ’hard-to-count populations’ are very challenging to obtain, and there is no ’gold standard’ method on which to base them (ibid.: p 31Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 109). BBS surveys have been largely donor-funded and have often been supported technically on the ground by university teams or public health agencies from high-income countries. BBS studies are a promising avenue for getting programmatically useful information on drug use and injection drug use, but they are costly and unlikely to be repeated frequently enough to inform dynamic program planning (Hakim et al. 2018). Reliance on donor funding and political unwillingness to prioritize assessments among criminalized or highly stigmatized persons are also barriers to undertaking BBS (ibid.). A number of African countries have conducted BBS on sex workers or clients of sex workers but not on PWID, MSM or transgender persons, for example (ibid.). A 2020 review concluded that, even with all the BBS and other studies that have been undertaken, there were only 21 PWID size estimates from the 54 countries in Africa (some now quite dated), as opposed to 70 for female sex workers, for example (Viswasam 2020). Remarkably, in spite of the considerable investment made by donors in size estimates of PWID and other HIV ‘key populations’, these estimates have been little used by donors and policymakers in their HIV strategies and programming (ibid.). Viswasam and colleagues conclude that the non-involvement of local stakeholders, including key population members themselves, in the design and implementation of some of these surveys contributes to their non-use in program planning (ibid.). As noted in more detail below, BBS surveys conducted with meaningful involvement of people who use drugs themselves may have positive collateral benefits. Programmatically important information other than population size that BBS surveys have the potential to gather is still lacking in much of Africa for PWID. According to UNAIDS’ data repository, only eight sub-Saharan countries have estimates of the per-capita number of needles available to PWID. Eight have estimates of the degree to which needles are shared and four have any information on availability of medication for opioid use disorders (methadone and buprenorphine) in spite of concerns about growing opioid use disorders on the continent (UNAIDS 2022). A PAN-AFRICAN DATA EFFORT BY THE AFRICAN UNION The AU has an initiative to improve data for health-related decision making in drug policy. The Pan-African Epidemiology Network on Drug Use (PAENDU) responds to an AU ministerial directive to ’address the lack of information on drug use and its effects on the continent’ (African Union 2019b). PAENDU collects data from centers providing treatment for drug use disorders as well as data from law enforcement agencies. The nine countries included are Angola, Botswana, Cameroon, Ghana, Guinea, Tanzania, Togo, Uganda and Zambia. Others with ’pre-existing surveillance programs’ are Kenya, Liberia, Mauritius, Nigeria, Senegal and South Africa and are noted as collaborating with PAENDU (ibid.). Treatment centers provide data on type of treatment, how the patient was referred to treatment, the main drug used, other health problems of patients and socio-demographic information. PAENDU was supported in its first years by the Bureau of International Narcotics and Law Enforcement of the US Department of State (ibid.). A PAENDU progress report for 2016-17 concludes that the main ’substances of abuse’ for people seeking DUD treatment in the PAENDU countries were cannabis and alcohol (ibid.). Opiates were the ’secondary substance of abuse’ in Nigeria; cocaine was the secondary drug in Guinea. The treatment centers were largely outpatient facilities except for psychiatric hospitals in Nigeria. There were relatively few treatment centers in most countries, but data from South Africa came from 83 distinct facilities. Most of the people for whom data was recorded by PAENDU were young men ’referred’ by family members or seeking treatment on their own. PAENDU was inspired partly by the South African Community Epidemiology Network on Drug Use (SACENDU), which has collected data from drug treatment centers since 1996 (South African Medical Research Council, undated), and the more recently established West African Network (WENDU) (ECOWAS and UNODC 2019). The well-established SACENDU not only reports on trends from arguably the largest network of drug treatment facilities per capita of any sub-Saharan country, but also serves as a platform for treatment providers and policymakers to share experiences and discuss trends. Perhaps most importantly with respect to drugrelated mortality, SACENDU’s regular reports since 2019 have also included information from 32Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 community-based harm reduction services (see, for example, South African Medical Research Council 2022). This crucial information on utilization of syringe exchange and methadone programs not only highlights characteristics of users of those services, but also shows where those services are sparse or absent. WENDU, the West African network supported by the European Community (EC), reported that for the period 2014–2017, as with PAENDU, cannabis and alcohol were the most frequently indicated primary substance used among people presenting for treatment, based on data from 15 countries (ECOWAS and UNODC 2019). Though cannabis and alcohol were declared as the principal problems by people seeking treatment, the same report indicates that cocaine, opioids (heroin and tramadol, a legal medicine) and amphetamine-type substances figured most prominently among seizures by drug authorities in the region. It is unsurprising that harm reduction services are not included in this report since they are so rare in the region. The WENDU report asserts that data from treatment facilities include only a small proportion of PWUD, particularly given the inaccessibility of treatment in many countries of the region (ibid.: 24). The report recommends expansion of community-based treatment facilities for drug use disorders as a policy priority, including ’increased advocacy for opioid replacement therapy’ (ibid.: 11). It is understandable that the AU would look to an apparently thriving system such as SACENDU as a model for improving its health-related drug information. But there are obvious disadvantages to relying on data from treatment facilities when specialized facilities, especially those offering scientifically sound care, are so rare in most of sub-Saharan Africa. The AU itself estimated in 2019 that only about 5% of people in need of treatment for drug use disorders in its member states actually obtained it, versus about 16% worldwide (African Union 2019b: 6). Moreover, the preponderance of treatment facilities reporting cannabis and alcohol as the principal problems they address will not help the AU focus on dealing with injection drug use, or the growing consumption of opioids with its attendant overdose risk or with cocaine and amphetamine-type substances, which are reportedly on the rise in Africa (WDR 2021: booklet 2). It is reasonable to wonder whether those presenting for treatment at many of these facilities are polydrug consumers but indicate cannabis or alcohol as their preferred substances because those are less stigmatized than heroin, tramadol or stimulants. OTHER DATA COLLECTION APPROACHES Given the limitations of general population surveys, BBS studies and information from treatment facilities, what might be other means of generating data that will be helpful to reach the most vulnerable and stigmatized people who use drugs? There are some experiences from Africa that may be useful models. None of these is perfect or suitable for all settings, but they may hold some lessons for sustainable data collection and monitoring. These examples do not represent an exhaustive list and do not result from a comprehensive search of the literature, but rather serve to illustrate some promising approaches. A ’program mapping’ approach with the active participation of PWID peer informants was used over a four-month period in South Africa to identify injecting locations in several provinces at which interviews with PWID were conducted (Scheibe et al. 2017). This effort assessed injecting equipment and condom availability, other health service needs, disposal of injecting equipment, mobility of the drug-using population and engagement with police. Population size estimates were made at a very local level using several triangulated rapid methods, relying partly on estimates made by PWID themselves. Population size was estimated to assess the level of need for services in the identified localities rather than to contribute to an exact province-level or national-level figure. An obvious key to the rapidity and apparently credible results of this effort was the meaningful participation of PWID and their organizations at all stages of the work. (The experience of PWID and their organizations of the reality of service provision is obviously crucial to inform a realistic program map.) PWID community advisory groups were convened in all locations, and PWID acted as data collectors and were important to the identification of study sites and estimation of service needs. In Tanzania, a rapid assessment using qualitative methods was designed to assess the level and nature of drug use outside Dar es Salaam and Zanzibar, where much was already known (Tiberio et al. 2018). The primary key informants were people who used drugs, with a particular 33Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 effort to include women about whom the extent of drug use was little understood. Observations of PWUD were supplemented by a range of secondary informants including health service providers, NGO personnel and law enforcement officials. Over 400 PWUD, identified largely by peers, provided information on what drugs were available and how they were being consumed, as well as information about health and social service gaps. Key informants were assured of anonymity and confidentiality of their participation. According to the authors, treating informants with respect and without judgment enabled rich information to be gleaned. The triangulation of data from several sources in this effort enabled detailed insight as to how drug use moved around the country and what kinds of services would be needed to address it. Rather than attempting exact population size estimates, this effort had enough information to classify localities as having ’nascent’, ’established’ or ’pervasive’ drug use, categories that corresponded to distinct conclusions about health service priorities. Some countries have built on their BBS studies in programmatically useful ways. Based on results from a 2011 BBS supplemented by some smaller key population surveys, Kenya developed a fairly comprehensive key populations strategy earlier than some countries. This initiative included forming a Key Populations Technical Working Group with representation of key population-led organizations, which has been a ’safe space’ for discussing service expansion strategies and even possibilities for reforming punitive laws (Musyoki et al. 2021). An important effort has been routine behavioral monitoring at local levels by peer educators and outreach workers, which has also enabled population size estimates based on local data to be updated every five years. Progress in translating these activities to better program coverage has been greater for sex workers and MSM who had their own organizations at an earlier stage than people who inject drugs. But the building of methadone services since 2014, and of syringe programs over a longer period, has helped to make marginalized users of opioids visible for monitoring systems (Mbogo et al. 2022). In Mozambique, similarly, BBS efforts helped to shape ongoing national monitoring and program planning (Semá Baltazar et al. 2021). The BBS studies in Mozambique established intersectoral collaboration, including meaningful involvement of civil society actors, in gathering bio-behavioral evidence. In the case of PWID, collaborations born from the BBS efforts were instrumental in developing a national harm reduction strategy. Semá Baltazar and colleagues (ibid.) assert that the BBS studies brought key population organizations out of the shadows because of their meaningful participation in the design, implementation and dissemination of the surveys, which also built the capacity of these organizations to participate in ongoing regular monitoring efforts. DISCUSSION AND CONCLUSION While there is undoubtedly some value to estimating the size of the populations that use or inject drugs in African countries, it is clear after numerous costly efforts toward that end that precise nationwide estimates are elusive. Policymakers and donor agencies should not condition support for basic services for PWUD on having precise population size estimates. BBS surveys focusing on key populations hold promise for delivering information useful to planning and implementing health services for PWID. They seem, however, to be most useful where people who use drugs are included meaningfully in planning the surveys and deciding how to use the results. Meaningful participation of people who use drugs is most likely in places where they are allowed to form organizations that can be recognized and registered as NGOs. Unfortunately, that is still not the case in all African countries. Organizations of PWUD, when allowed a seat at the table in a meaningful way, can be the most effective advocates for ensuring that bio-behavioral data are collected in the most relevant locations, then used most effectively to inform program planning and implementation and the design of continued monitoring. Making the most of BBS studies is also more likely when governments start from a point of openness to HIV prevention and other harm reduction services for PWUD. As WHO and UNAIDS (2010: 6) emphasize in their population size estimation guidelines, ’If there is no commitment to provide services, do not waste resources with repeated size estimation studies’. Countries such as South Africa, Tanzania and Kenya have, without the benefit of perfect information, established methadone programs and needle exchanges in at least some cities and are able 34Csete Journal of Illicit Economies and Development DOI: 10.31389/jied.158 through those programs to discern matters of acceptability of services and how to overcome barriers to expansion of services. Of course, any effort to include patients in surveys or ask them to identify others for inclusion in data collection efforts must ensure confidentiality and seek informed consent. With the increase of heroin injection in many parts of Africa (Eligh 2019) and the predicted dramatic increase in drug use in Africa through 2030 (WDR 2021: Booklet 2), the AU would do well to concern itself with addressing the morbidity and mortality of injection drug use as a top priority. PAENDU’s data collection strategy will plainly not illuminate these issues when contributions to its database are dominated by facilities working on cannabis and alcohol. Again, the preponderance of people claiming that their greatest treatment needs are related to cannabis and alcohol underscores the stigmatization faced by people who consume ’hard drugs’ in Africa and all over the world (UNAIDS, 2019). Active measures to reduce stigma and discrimination related to drug use should figure prominently in AU strategies and action plans. A useful activity for the AU would be to share lessons learned from the countries that have established syringe programs, opioid agonist therapies and overdose prevention measures, including the obstacles they overcame, the way they worked with law enforcement and criminal justice authorities, and whether and how they needed population size estimates and other survey data. The Dutch-funded NGO Love Alliance, for example, has amassed a database for 10 African countries of data from BBS surveys but also the state of harm reduction and other services for PWID, the state of the law on minor drug infractions, the situation of drug-related health services in prisons and pretrial detention, and perhaps most importantly, the degree to which people who use drugs are meaningfully involved in program and policy decisions affecting them (Love Alliance 2022). It would be useful for the AU to complete a database of this kind for all its member states. In this regard, the work of the European Union’s European Centre for Monitoring Drugs and Drug Addiction (EMCDDA), which compiles extensive data on health and social services for people who use drugs (EMCDDA, undated-A), would be illustrative. EMCDDA’s focus on health and social services as best practices in responding to drug use disorders might be expected in a region where, in most countries, people are not incarcerated for minor drug offenses that may be linked to DUD (EMCDDA, undated-B). Whatever data collection methods are brought to bear about the situation of people who use or inject drugs in Africa, where minor drug infractions are criminalized and often harshly penalized, it must be realized that neither surveys nor services will easily reach those who rightly and constantly fear arrest and incarceration. Other strategies may be needed. For example, with donor support, many African countries have mobilized peer paralegals from sex worker, MSM and PWUD populations to help their peers who are arrested or harassed by police with minor cases and to refer major cases to back-up lawyers (Wirya et al. 2020). In some settings, community-based peer paralegals may be important sources of information about the need for health and social services for PWUD, though they are not included as informants in standard survey protocols. Confidentiality protections would need to be rigorous if they were to be included as informants. In the end, the most important strategy for collecting programmatically useful information for health-friendly drug policy would be decriminalization (or effective decriminalization) of minor drug offenses, including possession of quantities of drugs for individual use. Even without major changes in penal codes, some countries around the world are finding ways to ensure that the first response to minor drug infractions is not arrest and criminal charges but rather some form of referral to health and social services. This kind of approach – which would also greatly reduce stigma faced by PWUD – is unlikely to come easily in much of Africa, but it should be seen as a pillar of rights-based, health-friendly drug policy. COMPETING INTERESTS The author has no competing interests to declare. AUTHOR AFFILIATIONS Joanne Csete Columbia University Mailman School of Public Health, US