Social context, diversity and risk among women who inject drugs in Vietnam: descriptive findings from a cross-sectional survey

Background Women who inject drugs (WWID) are neglected globally in research and programming yet may be likelier than males to practise sexual and injecting risks and be infected with HIV and more stigmatised but seek fewer services. Little is known about characteristics, practices and nexus between drugs and sex work of WWID in Vietnam, where unsafe injecting has driven HIV transmission, and commercial sex and inconsistent condom use are prevalent. This was the first quantitative investigation of Vietnamese WWID recruited as injecting drug users. This article summarises descriptive findings. Findings A cross-sectional survey was conducted among WWID in Hanoi (n = 203) and Ho Chi Minh City (HCMC) (n = 200) recruited using respondent-driven sampling. Characteristics varied within and between sites. Twenty-two percent in Hanoi and 47.5 % in HCMC had never sold sex. Almost all commenced with smoking heroin, some as children. Most injected frequently, usually alone, although 8 % (Hanoi) and 18 % (HCMC) shared equipment in the previous month. Some had sex—and sold it—as children; most had multiple partners. Condom use was high with clients but very low with intimate partners, often injecting drug users. HIV knowledge was uneven, and large minorities were not tested recently (or ever) for HIV. Nearly all perceived intense gender-related stigma, especially for drug use. Conclusion This ground-breaking study challenges assumptions about characteristics and risks based on anecdotal evidence and studies among men. Most WWID were vulnerable to sexual HIV transmission from intimate partners. Interventions should incorporate broader sociocultural context to protect this highly stigmatised population.


Introduction
Vietnam reported in 1991 the first case of human immunodeficiency virus (HIV) infection; 25 years later, 227,114 people were living with HIV and 74,442 people had died of AIDS [1]. The epidemic has been slowed down in recent years from the annual new infections of over 30,000 in 2006-2007, but there were still 12,500 people newly diagnosed in 2013 [2].
The HIV epidemic in Vietnam was triggered and driven by drug injection. In the early 1990s, the annual proportion of newly diagnosed HIV cases among people who inject drugs (PWID) was as high as 87 % [3]. By the early 2000s, HIV prevalence among injecting drug users in Vietnam peaked at around 30 % before slowly and steadily reducing to around 10 % in 2014 as harm reduction was introduced and scaled up [4].
Women who inject drugs (WWID) tend to progress faster than males to dependence; inject more frequently; have intimate partners who inject, acquire and die from HIV/AIDS; and have greater combined risks, partly because many sell sex to purchase drugs [5][6][7][8]. Stigma may be greater than towards men who inject drugs (MWID) because 'injecting drug use is often seen as contrary to the socially derived roles of women as mothers, partners and caretakers' ( [9], p. 19). Epidemic data shows that the share of drug injection as a mode of transmission has been reduced significantly from over 80 % in the 1990s to 35.4 % in the first 6 months of 2015 [4]. This indicates the ongoing significance of drug injection but also the increasing importance of sexual transmission. In such context, WWID as an HIV 'bridge' through cross-over of injecting and sex work (SW) is of epidemiological importance where commercial sex and inconsistent condom use are prevalent [10][11][12][13][14][15]. There is a dearth of research on WWID and of interventions that encompass drug use and wider health needs [7].
Little is known about characteristics, usage patterns, extent of sex work and HIV risks among WWID in Vietnam. Sentinel surveillance among PWID excludes females; most data on WWID is about SWs who inject [5,12,16,17]. This paper reports descriptive findings from a cross-sectional survey, aimed to inform policy makers and programme managers about characteristics of WWID in the two major cities of Vietnam and their HIV-related behaviours so that policies and programmes can be adapted to produce stronger impacts on the HIV epidemic in Vietnam. The research, conducted in Hanoi and Ho Chi Minh city in 2010-2011, was funded through an Australian Development Research Award.

Methods
An advisory group-consisted of representatives of WWID, HIV programme managers and public security officials-was set up to guide the study. Per advice of the group, participants were recruited from Hanoi and Ho Chi Minh city-the two largest cities with the highest numbers of people who inject drugs and also the highest concentration of WWID.
Women aged 18+ who injected at least once in the previous 6 months were recruited using respondentdriven sampling (RDS) [15,[17][18][19]. The sample size of 200/site was based on the assumed prevalence of 50 % for key responses (which would yield the biggest sample size), 95 % confidence interval, 8.5 % margin of error and design effect of 1.5. In each city, the recruitment started with nine 'seeds' , balanced between age groups, HIV status and sex work involvement. Each participant was given three coupons to recruit others. Data collection was done at a drug user organisation's office. Core members of the organisations provided information about the study; screened potential participants for eligibility, especially by checking injection marks and asked questions about injection practice; and monitored recruitment to avoid repeated participation. Interviewers were social researchers experienced in and comfortable with interacting with WWID. Participants got compensation of 150,000 Vietnam dongs (around 8

Knowledge and testing
Knowledge about HIV transmission through tattoos and breastfeeding was inadequate, and one fifth in Hanoi and 40 % in HCMC believed they could identify an

Gender and perceived stigma
Our sample perceived WWID (especially) and SWs as intensely stigmatised. The vast majority felt drug use or selling sex inhibited finding a non-injecting partner ( Table 3).   Sex work is worse than female drug use 11.5 (5.8-16.5) 11.6 (7.
Over one fifth in Hanoi and nearly half in HCMC reported they had never sold sex. Among those who had, two thirds were using drugs before they first sold sex (Fig. 2).
Some sold sex as children. Mean age of first SW was 25  in Hanoi and 20  in HCMC, where almost 20 % sold sex before age 17 (3.3 % in Hanoi). Substantial minorities sold sex to buy drugs for partners. Over one third claimed SW was their decision. In  were found mainly in public places, rather than brothels or bars. Condom use was high with clients. However, 76 % of sexually active women in Hanoi and 83 % in HCMC had ≥1 'husband/boyfriend' , many/most of whom used drugs; just 17.5 % in Hanoi and 32 % in HCMC used condoms the last time.

Limitations
Reporting of certain behaviours may be influenced by recall and social desirability bias. RDS recruits through peer networks; hence, some types of WWID, e.g. those who rarely interact with others, may not be sampled. Fears of facing the police (for doing sex work or using drugs) might have prevented some WWID to participate. Also, lack of a known sampling frame precludes certainty about generalisability.

Discussion
Participants' demographic data reflects the diversity of WWID (age range, socio-economic status, living arrangement, etc.), and the North-South differences imply different strategies are needed to reach and to deliver interventions to them.
However, common issues (and needs) of WWID were identified through the study: being single mothers, had sex or sold sex as a child, heavily dependent on drugs with a high frequency of injection, not using condom with intimate partners-multiple of them-most/all injectors with high probability of having HIV, inadequate knowledge on HIV transmission, suboptimal access to HIV testing, lacking knowledge on STI and viral hepatitis and high perceived stigma from society. Programmes to prevent blood-borne infections should be intensified among WWID. Psychological support, counselling, family planning and parenting skills are among interventions needed to address their different immediate needs.
From these WWID, we learn that drug use led some of them to sex work. Drug-dependent treatment would be an important intervention strategy to prevent this. But we also learn that not all WWID sell sex, so programmes targeting sex workers would not reach many of the WWID.   Among those who had regular clients --Unable to generate confidence interval in RDSAT