HCV in Republic of Georgia & in Russia
Download the PDF here
Download the PDF here
Prevalence of hepatitis B and C among HIV positive patients in Georgia and its associated risk factors
Georgian Med News. 2008 Dec;(165):54-60.
Badridze N, Chkhartishvili N, Abutidze A, Gatserelia L, Sharvadze L.
The aim of the study was to determine the prevalence of hepatitis B virus (HBV) and hepatitis C virus (HCV) co-infection among HIV positive patients, to identify most relevant risk factors of co-infection and develop preventive interventions. Study participants were voluntary individuals 18 years of age or older recruited from AIDS Center VCT unit in Tbilisi, Georgia. Eligibility criteria of participants were: HIV positive result confirmed by western blot; age; and voluntary participation. Total 175 patients undergo interview with specially designed questionnaires. Most of the participants were male (71.4%), age range of HIV positives varied from 20 to 77 years old. Prevalence of HCV among HIV positive patients is high. Almost half (48.57%) HIV positive patients are co-infected with HCV. Men were more likely than women co-infected with HCV (60.80% and 18% accordingly). Major risk factor of male co-infection was related to drug use, needle and injection equipment sharing. Prevalence of HCV among injecting drug users was (73.40%). Drug users had 3.25 times more risk (PR 3.25; 95%CI; CL--1.89-5.26; p<0.01) to be infected with HCV compare non IDUs.
Prevalence of being infected with HBV (Anti-HBc) among HIV positives was 43.42% (76/175) and the prevalence of Chronic HBV (HBsAg positive) was 6.86% (12/175). Prevalence rate of HBsAg among IDUs was 8.51% and among non IDU participants 5.26%. Triple infection (HIV, Hepatitis C and chronic form of Hepatitis B--HBsAg) was among 9 patients (5.14%). Infections were associated with injection drug use (88.88%) and mostly were related to share of needles/syringes and other injecting medical equipment. Transmission of HBV and HCV by sexual contact was not observed among those 9 participants. High risk behavior among HIV positive participants mostly related to drug use and unprotected sex with non regular partners. Other risk factors for Hepatitis transmission were associated with invasive medical manipulations, blood transfusion, surgery, abortions and etc. None of cases of HIV, or Hepatitis (B, C) transmission through medical manipulations can be documentary proved based on those research data.
Outside of the government, more reliable hepatitis C estimates are available from surveys and studies of varying breadth and quality over the past decade. They indicate that hepatitis C prevalence in Georgia is especially high among injecting drug users (up to 70%4) and people living with HIV (48.6%5).
Prevalence of HIV, hepatitis C and syphilis among injecting drug users in Russia: a multi-city study
Addiction February 2006
Objectives To estimate the prevalence of HIV, hepatitis C virus (HCV) and syphilis in injecting drug users (IDUs) in Russia.
Methods Unlinked anonymous cross-sectional survey of 1473 IDUs recruited from non-treatment settings in Moscow, Volgograd and Barnaul (Siberia), with oral fluid sample collection for HIV, HCV antibody (anti-HIV, anti-HCV) and syphilis testing.
Results Prevalence of antibody to HIV was 14% in Moscow, 3% in Volgograd and 9% in Barnaul. HCV prevalence was 67% in Moscow, 70% in Volgograd and 54% in Barnaul. Prevalence of positive syphilis serology was 8% in Moscow, 20% in Volgograd and 6% in Barnaul. Half of those HIV positive and a third of those HCV positive were unaware of their positive status. Common risk factors associated with HIV and HCV infection across the cities included both direct and indirect sharing of injecting equipment and injection of home-produced drugs. Among environmental risk factors, we found increased odds of anti-HIV associated with being in prison in Moscow, and some association between official registration as a drug user and anti-HIV and anti-HCV. No associations were found between sexual risk behaviours and anti-HIV in any city.
Conclusions HIV prevalence among IDUs was markedly higher than city routine surveillance data suggests and at potentially critical levels in terms of HIV prevention in two cities. HCV prevalence was high in all cities. Syphilis prevalence highlights the potential for sexual risk and sexual HIV transmission. Despite large-scale testing programmes, knowledge of positive status was poor. The scaling-up of harm reduction for IDUs in Russia, including sexual risk reduction, is an urgent priority.
Despite indication of levelling out in the number of new HIV cases reported nationally to the Ministry of Health, the Russian HIV epidemic is one of the fastest growing in the world. Russia contributes approximately 70% of cumulative HIV cases in the eastern, central, south-eastern and central Asian region . Over 80% of HIV cases reported since 1996 have been associated with injecting drug use (IDU), and the evidence indicates that multiple HIV outbreaks have occurred among IDUs in different cities, over a decade after HIV transmission peaked in western Europe . In one earlier survey of community-recruited IDUs in Togliatti City, we found 56% (234/418) HIV positive, half of whom were under 25 years, and 87% (357/411) HCV positive .
Furthermore, there is high background prevalence of sexually transmitted infections (STIs) in the Russian population, with syphilis prevalence peaking in 1997 at 275 per 100 000 population and remaining high by 2002 at 120 per 100 000 population . Little is known about the prevalence of syphilis among Russian IDUs. Recent suggestion of increasing sexual HIV transmission in Russia, and high levels of engagement in sex work by female IDUs, raises the concern that sexual mixing between IDUs and their sexual partners might facilitate a shift towards a more generalized HIV epidemic .
There is a strong emphasis on population screening in Russia, with around 20 million HIV screening tests conducted annually. Despite this there are few reliable city estimates of HIV, hepatitis C virus (HCV) and syphilis prevalence among IDUs, especially in populations sampled outside drug treatment or health service settings. Recognizing the urgent need to foster evidence-based prevention of blood-borne and STIs in Russia and the eastern European region more broadly, we undertook the largest community-recruited survey to date of IDUs in Russia to measure HIV, HCV and syphilis prevalence.
We conducted an unlinked anonymous cross-sectional survey of 1473 IDUs recruited from non-treatment settings in Moscow (n = 455), Volgograd (n = 517) and Barnaul, Siberia (n = 501) in September and October 2003. All participants had injected drugs in the last 4 weeks. IDUs were recruited in multiple sites in community settings by a team of trained 'indigenous fieldworkers' who also administered the interview-based survey [6,7]. Community recruitment settings included street locations, respondents' homes and cafes. Indigenous fieldworkers were defined as interviewers who were current or former drug users, had established professional experience in the substance misuse field or had privileged access to injecting drug user networks. Indigenous fieldworkers may act to reduce the bias arising from socially desirable responses in behavioural data [6,7]. As in similar studies , measures to ensure data quality and minimize network bias included: limiting the number of interviews per fieldworker to two per day and a total of 40 interviews throughout the fieldwork; random spot-checks in the field; and validation follow-up interviews with 10% of participants. IDUs received HIV prevention materials (including needles and syringes) as well as chocolates and cigarettes for their participation. The study had ethical approval from the Riverside Local Research Ethics Committee and the support of the Russian Ministry of Health National Scientific Centre for Research on Addictions.
Anti-HCV, anti-HIV and syphilis antibody testing
Oral fluid specimens were obtained using the OraSure device (Epitope Inc., OR, USA) and screened for antibodies to HCV (anti-HCV) and HIV (anti-HIV). Specimens were screened for anti-HCV using a UK Health Protection Agency (HPA) modified Ortho HCV 3.0 enhanced SAVe enzyme-linked immunosorbent assay (ELISA) . For anti-HIV, the Oral Fluid Vironostika HIV-1 Microelisa System (BioMerieux, Inc., NC, USA) was used [9,10]. Specimens reactive on initial testing were subject to confirmatory testing using Detect HIV (Biostat Diagnostics, UK) and HIV blot 2.2 Western blot assay (AbbotMurex, UK). These blot assays employ a modified protocol developed for use with oral fluid by the UK HPA . Antibodies to Treponema pallidum were tested using a commercial test Murex ICE Syphilis (AbbotMurex) modified for use with oral fluid by the UK HPA (J. Parry, personal communication, 2005).
Associations between antibodies to HIV and hepatitis C and covariates were explored univariably and by multiple logistic regression. The outcomes were positive antibodies to HIV (anti-HIV) and hepatitis C (anti-HCV), and odds ratios associated with infection. Risk factors associated with positive T. pallidum cases are reported elsewhere . Intra-cluster correlation coefficients were calculated to measure the degree to which observations from individuals recruited by the same interviewer were correlated and general estimating equations (GEE) were used to adjust for any correlation. All multivariable analysis followed a conceptual framework approach . This involved classifying variables into the five groups displayed in Tables 1-4, with the analysis conducted in three stages: (i) separate univariable models explored each of the variables alone with the outcome; (ii) variables associated with the outcome in univariable analysis to a significance level of P = 0.2 were included in separate multivariable models for each group; and (iii) variables reaching a significance level of P = 0.2 in each of the five multivariable models were then included in one overall multivariable model. In addition, variables excluded at the first stage were added at the second- and third-stage models to assess their association with the outcome variable in the presence of other variables.
A total of 1473 IDUs were recruited into the study. Two-thirds (70.5%) were male, half (50%) were aged under 25 years, one-fifth (20%) had injected for 2 years or less and most (81%) injected less than daily. The most commonly injected drug in the last 4 weeks was heroin (72%). Approximately a quarter (26%) had injected home-produced drugs in the last 4 weeks, including 'hanka' or 'mak' (liquid derivatives of opium poppy straw) or 'vint' (a liquid methamphetamine).
In the last 4 weeks, approximately one-sixth (14%) reported injecting with needles and syringes used previously by others, and 84% reported injecting with shared injecting paraphernalia other than needles or syringes, including 'frontloading' (whereby a drug solute is squirted from a donor syringe into another by removing the needle), injecting with filters previously used by others, filling their syringe from a 'working syringe' used by multiple people to distribute the drug solute or injecting with prefilled syringes (usually prefilled prior to purchase). The majority of the sample (82%) reported injecting less than daily. More than half (59%) reported inconsistent condom use during vaginal sex with a non-paying sexual partner in the last 4 weeks. Among female IDUs, 24% (105/433) had exchanged sex for money, drugs or goods in the last 4 weeks.
The prevalence of anti-HIV was 14% (55/403, 95% CI 10.3-17.0%) in Moscow, 3% (13/477, 95% CI 1.3-4.2%) in Volgograd and 9% (44/499, 95% CI 6.3-11.3%) in Barnaul (Table 5). A third (34%) of those HIV antibody positive in the total sample self-reported as such. Between half (52% and 53% in Moscow and Barnaul, respectively) and three-quarters (74.5% in Volgograd) of IDUs had been HIV tested in the last year. Almost all (92%) of those anti-HIV positive were also anti-HCV positive.
Table 1 shows prevalence of anti-HIV in the survey sample by key characteristics of participants in Moscow. In the univariable analysis, prevalence and odds were higher among IDU who reported that on the last day of injection they only injected once (OR = 1.6, 95% CI 1.38-5.02) and among those who had ever injected with a used needle or syringe (OR = 3.1, 95% CI 1.23-7.77). Of the environmental risk factors, odds of being positive to anti-HIV was higher among those who had ever been in prison (OR = 1.9, 95% CI 1.46-2.53) and among those who were registered as a drug user at the narcology service (OR = 2.4, 95% CI 1.26-4.65). Odds of being anti-HIV positive were lower among those who reported ever having a sexually transmitted infection (STI) (OR = 0.5, 95% CI 0.25-0.85). Prevalence and odds also increased by duration of injection.
After adjustment, five variables remained associated with anti-HIV in both standard logistic regression and GEE-adjusted models. There was no substantial difference in the GEE and logistic regression models indicating that there was limited clustering of risk behaviours related to fieldworkers, so only one set of figures are presented below (GEE estimates available on request). IDUs who reported that they had ever injected with used needles/syringes were three times as likely to be anti-HIV positive than those who had not (OR 3.5, 95% CI 1.43-8.36). IDUs who reported a history of prison were twice as likely to be found positive to anti-HIV than those who had never been in prison (OR = 2.2, 95% CI 1.0-4.65), and those officially registered at narcology (drug treatment) services were twice as likely to test positive to anti-HIV (OR = 2.3, 95% CI 1.12-5.05). IDUs who reported ever having an STI were less likely to test positive for anti-HIV (OR = 0.3, 95% CI 0.15-0.70). Those reporting on the last day that they injected they had injected once compared to twice were more likely to be positive to anti-HIV (OR = 2.4, 95% CI 1.12-5.05).
Table 2 shows univariable and multivariable risk factors for antibodies to HIV among IDU in Volgograd. The same variables were significant in both adjusted and unadjusted analyses. In the adjusted model, IDUs reporting that they had injected with a used needle/syringe of a sex partner in the last year were almost 10 times more likely to be found anti-HIV positive than those who had not (OR = 9.6, 95% CI 1.95-47.0). IDUs who reported injecting daily had increased odds of testing positive for anti-HIV (OR = 6.9, 95% CI 1.91-25.1) compared to those injecting less than daily. Finally, those who reported not having a regular job were less likely (OR = 0.2, 95% CI 0.05-0.75) to be positive for anti-HIV than those with a regular job. The small number of HIV cases in this city prevented the use of the GEE model.
Table 3 shows univariable and multivariable risk factors for antibodies to HIV among IDU in Barnaul. In the univariable analysis the following risk factors were associated with testing positive for antibodies to HIV: injecting heroin compared to home-made drugs (OR = 4.0 95% CI 1.82-8.81); use of communal spoon (OR = 3.7, 95% CI 1.93-6.98); having one or no sex partners in the last 12 months compared to having two or more (OR = 2.3, 95% CI 1.23-4.44); and never having injected home-produced drugs (OR = 2.4 95% CI 1.07-5.28). Reduced odds of anti-HIV were associated with injecting twice on the last day of injection compared to once (OR = 0.4, 95% CI 0.18-0.81) and having been in prison (OR = 0.4, 95% CI 0.18-0.87). After adjustment, the logistic regression model indicated increased odds associated with anti-HIV for the following risk factors: female IDUs were less likely to be positive for anti-HIV than men (OR = 0.3, 95% CI 0.13-0.72); and participants who had been in prison were less likely to be positive to anti-HIV than those with no prison experience (OR = 0.2, 95% CI 0.09-0.57). Increased odds of HIV were associated with injecting heroin compared to vint or mak (OR = 3.2, 95% CI 1.40-7.37), use of a communal spoon (OR = 4.7, 95% CI 2.29-9.62) and injecting from a communal working syringe (OR = 5.0, 95% CI 1.75-14.1).
In Barnaul, there was strong evidence of clustering of HIV by fieldworkers. All HIV cases were found among respondents interviewed by two fieldworkers and all these respondents reported heroin as their main drug. This might suggest that the cases were concentrated in one or two networks recruited by fieldworkers. After adjustment with the GEE model, only two risk factors remained associated with being positive to anti-HIV. First, those who reported that on the last day of injection they had injected twice were slightly less likely be anti-HIV positive than those who had injected once (OR = 0.7, 95% CI 0.58-0.93). Secondly, IDUs who reported sharing communal spoons for the preparation of drugs were more likely to be anti-HIV positive than those who had not (OR = 1.7, 95% CI 1.1-2.56).
The prevalence of anti-HCV was 68% (296/434, 95% CI 63.8-72.6%) in Moscow, 70% (353/507, 95% CI 65.6-73.6%) in Volgograd and 54% (265/491, 95% CI 49.5-58.4%) in Barnaul (Table 5). Half (48%) of those found HCV antibody positive in the total sample self-reported as such. Of those with a history of HCV testing and reporting their last test to be antibody negative (n = 454), 52% were anti-HCV positive. While most IDUs were HCV tested in the last year in Volgograd (73%), this was the case for only a third (35%) in Barnaul and a half (47%) in Moscow).
Table 4 summarizes univariable analysis of risk factors associated with anti-HCV for each of the three cities. For the Moscow sample, key univariable associations indicating increased odds of infection with anti-HCV included ever having injected with a used needle or syringe (OR = 2.5, 95% CI 1.61-3.89), a history of drug treatment (OR = 2.3, 95% CI 1.43-3.62) and being officially registered as a drug user at a narcology service (OR = 2.5, 95% CI 1.34-4.47). Decreased odds were associated with sex work (OR = 0.3, 95% CI 0.15-0.67) and not having a regular source of income (OR = 0.6, 95% CI 0.41-0.97).
In Volgograd, odds and prevalence of anti-HCV increased with age and duration of injection. IDUs who reported ever having injected home-produced drugs were more likely to test positive for anti-HCV (OR = 2.8, 95% CI 1.9-4.3). Additionally, increased odds were found among those who had injected using a filter (usually a cotton) used previously by someone else (OR = 1.8, 95% CI 1.2-2.81) and those who reported ever having injected with a used needle/syringe (OR = 2.3, 95% CI 1.54-3.38). Increased odds were also associated with a history of drug treatment (OR = 2.3, 95% CI 1.17-2.57) and being officially registered as an IDU at a narcology service (OR = 3.0, 95% CI 1.78-5.05). Decreased odds were associated with sex work (OR = 0.1, 95% CI 0.06-0.28).
In Barnaul, odds and prevalence of anti-HCV increased with age and duration of injection. Participants who reported ever having injected home-produced drugs were 4.2 times as likely to be anti-HCV positive than those who had not (95% CI 2.2-8.0). Increased odds were also associated with frontloading (OR = 3.6, 95% CI 1.6-8.1), using with a filter that someone else had used previously (OR = 1.7, 95% CI 1.03-2.7) and reuse of the same needle more than once (OR = 2.1, 95% CI 1.4-3.1). Of the environmental variables, increased odds were associated with ever having been in prison (OR = 2.9, 95% CI 1.93-4.3), a history of drug treatment (OR = 2.9, 95% CI 1.9-4.4) and being officially registered as a drug user at a narcology service (OR = 2.6 95% CI 1.6-4.2).
Table 6 summarizes the risk factors associated with anti-HCV in each site in the multivariable analysis. In Moscow, the adjusted GEE model showed that prevalence and odds of anti-HCV was lower among female IDUs involved in sex work than men (OR = 0.2, 95% CI 0.08-0.50). IDUs who reported ever having injected with used needles/syringes were almost three times as likely to be anti-HCV positive than those who had not (OR = 2.5, 95% CI 1.57-3.85) and those who reported being officially registered at a narcology service were almost twice as likely to be anti-HCV positive than those who were not (OR = 1.9, 95% CI 1.12-3.13). There was little difference between the adjusted GEE and the standard logistic regression model other than the logistic regression model indicated an association with history of drug treatment (OR = 1.8, 95% CI 1.5-2.98) rather than registration as an IDU and an association with ever having injected home-produced drugs (OR = 3.2, 95% CI 1.48-6.72).
In Volgograd, the adjusted GEE model again showed that female IDUs (OR = 0.7, 95% CI 0.51-0.95) and female IDU sex workers (OR = 0.3, 95% CI 0.11-0.66) were less likely to be anti-HCV positive than male IDUs. Prevalence and odds of testing positive for anti-HCV increased by duration of injection, participants who reported injecting between 3 and 5 years were 2.5 times more likely be found anti-HCV positive (OR = 2.4, 95% CI 1.55-3.66) and those who reported injecting between 6 and 9 years had almost three times the odds of testing positive for anti-HCV (OR 2.7, 95% CI 1.37-5.41). There were no differences in significant risk factors between the GEE and logistic regression models.
In Barnaul, the adjusted GEE model indicated that IDUs who had attended higher education were less likely to be anti-HCV positive (OR = 0.5, 95% CI 0.3-0.97) than those who had not. As with Volgograd, prevalence and odds of anti-HCV increased with duration of injection, with those who reported injecting longer than 10 years being over three times as likely to test be anti-HCV positive than recent initiates into injecting (OR = 3.3, 95% CI 1.33-7.98). Also similar to Volgograd, those who reported ever having injected home-produced drugs were twice as likely to be anti-HCV positive than those who had not (OR = 2.0, 95% CI 1.07-3.65). Finally, IDUs in Barnaul who reported injecting with used needles/syringes in the last 4 weeks were almost twice as likely to be anti-HCV positive (OR = 1.8, 95% CI 1.05-2.97) than those who had not. The adjusted logistic regression model indicated a much stronger association between testing positive for anti-HCV and duration of injection than the GEE model. It also indicated increased odds associated with reuse of needles more than twice (OR = 2.1, 95% CI 1.39-3.13), whereas no association was found with the GEE model (OR = 1.0, 95% CI 0.58-1.57). Conversely, no association was found for ever having injected with used needles/syringes in the adjusted logistic regression model (OR = 1.4, 95% CI 0.77-2.44) but was for the GEE model (OR = 1.8, 95% CI 1.05-2.97).
The prevalence of syphilis was 8% (32/414, 95% CI 5.1-10.3%) in Moscow, 20% (93/438, 95% CI 16.2-23.5%) in Volgograd and 6% (32/494, 95% CI 4.3-8.7%) in Barnaul (Table 5). Over half (85/156, 54%) of those found positive to T. pallidum (which indicates past or current infection) reported never having had an STI. The mean (SD) age of those testing positive to syphilis was 26 (5.9) years. Three-quarters (74%) of those positive to syphilis were HCV positive and 10% HIV positive.
This study indicates varied prevalence of HIV (at around 10% or more in two cities), high prevalence of HCV and varied prevalence of syphilis (at 20% in one city) among the largest community-recruited sample of IDUs in Russia. Two-thirds of IDUs were unaware of their positive HIV status, and half were unaware that they had had syphilis. A third of IDUs were unaware of their positive HCV status, and over a half whose last test was reportedly HCV negative were found to be HCV positive. Despite large-scale investment in blood-borne virus screening programmes targeting populations at risk in Russia, significant proportions of IDUs (including HIV- and HCV-positive IDUs) had not been tested recently for HIV and HCV. Findings point to a large burden of blood-borne and sexually transmitted infection, of which a considerable proportion is undiagnosed, with major public health implications.
Findings underscore the need to foster regular access of IDUs to HIV, HCV and STI screening while maximzsing their self-awareness of HIV, HCV and syphilis status. Targeted behavioural surveillance in community settings can serve to highlight the limits of city screening and testing programmes among hard to reach populations at risk. Of 2 126 958 HIV screening tests undertaken in Moscow in the year of the survey (2003), only 0.3% (6397) were among IDUs, of which 3.2% (202) were HIV positive. Similarly, in Volgograd and Barnaul, HIV tests among IDU represent only 1.7% (5214) and 0.5% (2710) of the total screening tests conducted in 2003 (298 810 in Volgograd and 509 273 in Barnaul). Under 3% of screening tests among IDU in each city were positive (2.6%; 136 in Volgograd; and 2.7%; 75 in Barnaul). Our study findings thus estimate HIV prevalence among IDUs to be over four times higher than estimates derived from screening programmes in Moscow, over three times higher in Barnaul but roughly equivalent in Volgograd, where significantly higher proportions of IDUs reported having had recent HIV tests.
The majority of HIV case reports among injecting drug users are compiled from narcology clinics where HIV testing is obligatory. Case reports are also compiled from clinics for sexually transmitted infections and AIDS centres, when testing can be conducted anonymously for a fee. Anonymous tests are not included in national case reports. Evidence suggests that attendance at narcology clinics has declined in recent years, with greater proportions of IDUs favouring private treatment or self-medication . The low coverage of drug treatment services and access to anonymous testing may introduce bias into the routine screening system as well as explain the difference in estimated prevalence between our findings and city case reports. These findings emphasize the critical importance of second generation surveillance and targeted surveys of HIV prevalence among community-recruited samples of injecting drug users as an adjunct to current large-scale HIV screening programmes.
Our findings suggest while risk factors associated with anti-HIV varied by city, there were some common risk behaviours across the cities that require urgent attention to prevent further HIV transmission. In Moscow and Volgograd increased odds of HIV were associated with injecting with used needles and syringes. In Barnaul, increased odds of HCV were associated with the use of communal spoons or glasses used for mixing and cooking the drug solute. Some studies point to a relationship between injecting paraphernalia sharing and HCV transmission risk, although the validity of this association has been questioned [15-17]. None the less, our findings highlight an urgent need to reduce levels of direct needle and syringe sharing as well as indirect sharing of injecting paraphernalia among Russian IDUs.
In Moscow we found increased odds of HIV among participants who had been in prison. According to the Russian Ministry of Justice, around three-quarters of prisoners in Russia have a serious disease such as HIV infection and tuberculosis, almost all of whom have experience of drug use . In our view, this highlights a need for introducing the provision of sterile needles and syringes within prisons.
In all cities we found some evidence of increased odds of HCV associated with having injected home-produced drugs. Similar associations have been observed in other Russian cities [2,3]. Although less common since the diffusion of heroin powder into most Russian cities, a combination of anecdotal and qualitative research evidence suggests that a communal pattern of drug preparation and use was, and still remains, a feature of home-produced drug use, and that a pre-existing culture of group use thus also shapes how heroin is used. Group drug-using situations offer greater potential for injecting equipment sharing. We also found some association between official registration as a drug user and anti-HIV and anti-HCV. There is some evidence to suggest that official registration with narcology services can serve to entrench further the social marginalization of IDUs, affecting their capacity to obtain legal employment as well as marking them out as targets for police surveillance [14,19]. Such factors may in part explain the increased odds associated with registration and HIV in Moscow and HCV in Volgograd, and are illustrative of how environmental factors contribute to the social structural production of HIV risk .
We found no association between sexual risk behaviours and anti-HIV in any city, suggesting that currently the predominant source of transmission risk is injecting drug use. The reduced odds of HCV among sex workers in Moscow and Volgograd implies that sex workers may be engaging in less risky injecting practices, thereby reducing their chances of acquiring HCV. HIV prevalence was no less among sex workers, perhaps suggesting the potential role of sexual transmission among the population. Additionally, reported condom use was low among all IDUs, highlighting the need for interventions targeting sexual risk reduction to prevent onwards transmission of HIV and syphilis to non-injecting populations.
As the data are drawn from a cross-sectional survey and reported behavioural findings are based on self-reports, any inferences about causality are limited. Using a fieldwork team of current or former drug users with privileged access to the target population may have limited potential bias associated with socially desirable responses. A further limitation inherent in methods of community recruitment of hidden populations is that there is no established sampling frame from which a measure of representativeness can be obtained. We attempted to minimize potential geographic and network bias by ensuring multi-site and multi-network recruitment and by the use of general estimating equation models in the analysis. Few HIV cases in Volgograd led to weaker odds ratios and wide confidence intervals in the HIV risk factor analysis, and these results should therefore be interpreted with greater caution.
This study indicates the urgent need to scale-up HCV and HIV prevention, as well as sexual risk reduction, for IDUs in Russia. It has been suggested that 10% HIV prevalence can be a critical threshold in the efficient containment of HIV epidemics among IDU, as after this point far greater resources and intervention coverage are required to bring about epidemic containment or reversal . STIs, including untreated syphilis, can accelerate sexual HIV transmission . Additionally, high levels of HIV and HCV coinfection can complicate medical management, including relating to antiretroviral HIV treatment . An estimated 10% of HIV positive IDUs in Russia have access to HIV combination therapy, and under 15% in HIV treatment are IDUs (who make up 80% of all HIV infections) .
Moreover, epidemics of HIV, HCV and syphilis among IDUs coincide with wider social and economic transitions in Russia, creating 'risk environments' conducive to the spread of blood-borne and sexually transmitted infections [25,26]. Rapid diffusions in drug injecting have coincided with drug trafficking, and population migration and mixing, in the context of economic transition and restructuring, which has increased unemployment and poverty, of which the growth of informal economies is a feature (including drug and sex markets) . At the same time, decline in health and welfare status have coincided with increasing incidence of communicable disease and weakening public health infrastructures which have produced vulnerability in public health capacity and response . Prevention of transmission of blood-borne viruses as well as HIV and HCV treatment coverage for IDUs is woefully inadequate. International support for local 'harm reduction' has resulted in approximately 70 syringe exchange projects throughout the Federation, but crude estimates suggest these reach under 1%, and at best around 16%, of local IDU populations .
Public health-orientated approaches to 'harm reduction' in Russia remain in tension with detoxification treatment and law enforcement oriented approaches to tackling drug problems. The linked epidemics of HCV, HIV and syphilis among IDU in Russia are of potential major significance to the public health. The scaling-up of harm reduction, including sexual risk reduction, remains an urgent priority.
We are grateful for the support of the UK Department for International Development, who supported this study through programme grants, and to the UK Department of Health who provide core funding to the Centre for Research on Drugs and Health Behaviour. We thank the study participants and the following: Sergei Belikh, Dimitry Blagov, Elvira Demyanyenk, Alexander Fillipov, Nadezhda Gorshkova, Ali Judd, Elena Kudravtseva, Olga Mikhailova, Andrei Rylkov, Anya Sarang, Nelly Savelevna, Lenar Sultanov, Grigoryev Svyatoslav, Mikhail Tichonov, Venyamin Volnov and Konstantin Vyshinsky.
Declaration of interest
This study was undertaken with ethical approval from the Riverside Ethics Committee and with the support of the Russian Ministry of Health National Scientific Centre for Research on Addictions. In addition, there are no conflicts of interest to declare.