Male-to-male sex is the major route of HIV transmission in high-income countries, and the emerging data from low-income and middle-income countries indicate that men who have sex with men (MSM) bear a substantial burden of HIV epidemics there as well.1 2

It is estimated that 1.5 (range 1.1–1.9) million people were living with HIV in the countries of central and eastern Europe (CCEE) in 2007, with an estimated adult prevalence (15–49 years old) of 0.8% (range 0.6%–1.1%).3 Injecting drug use remains the main mode of HIV transmission in the 15 countries of the former Soviet Union, though there has been an increase in the number of reported heterosexually acquired infections.4 In the other countries of eastern Europe, the main modes of transmission are sexual: both heterosexual and through male-to-male sex.5 MSM remain the group at the greatest risk of HIV in western Europe, because of high levels of HIV-related risk behaviours and increasing incidence of sexually transmitted infections (STIs).6 7

In the CCEE, data on the epidemiology of HIV among MSM are less available compared with other vulnerable groups, which has partly been attributed to specific social conditions and stigmatisation of homosexuality in post-communist, transitional societies.8 9 In the recently published assessment of the quality of HIV sero-surveillance in low-income and middle-income countries, only one country of eastern Europe and central Asia (Ukraine) was assessed as having a fully functioning surveillance system.10 The key weaknesses identified were the over-reliance on HIV and AIDS case reporting in longer-term tracking of the epidemic and few studies conducted among at-risk groups other than injecting drug users. The concerns about insufficient quality of data on population health and their limited usefulness for informed policy decisions, coupled with weak infrastructure for communicable disease control, have been well recognised by experts.11

To address the apparent gap in knowledge about the level of HIV and risk behaviours among MSM in eastern Europe, in 2008 the World Health Organization Regional Office for Europe (WHO/Europe) commissioned a study that aimed to review the available data on HIV and STIs among MSM in the CCEE, the response to the epidemic and the evidence on stigma and discrimination against MSM. This paper aims to present the results of this study.

METHODS

We used the classification of the European Centre for the Epidemiological Monitoring of HIV/AIDS (EuroHIV), which groups the CCEE into a central region (Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Hungary, the former Yugoslav Republic of Macedonia, Poland, Romania, Serbia and Montenegro, Slovakia, Slovenia, Turkey), and an eastern region (Armenia, Azerbaijan, Belarus, Estonia, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Republic of Moldova, Russian Federation, Tajikistan, Turkmenistan, Ukraine, Uzbekistan).

Data collection

In each country, researchers proposed by WHO/Europe were asked to collect the literature on HIV epidemiology among MSM, prevention responses and stigma and discrimination in the local language and/or in English, published between 1 January 2000 and 1 February 2008. Where data were not available after 2000, we asked country collaborators to send us the latest available data. The literature included published papers and reports on HIV, STIs and behavioural surveillance surveys, research studies carried out among MSM, published abstracts, conference reports and “grey” literature available in the country. In order to receive comparable reports, all collaborators were sent a template of a country report and a list of 17 biological and behavioural indicators with specific time frames. If these data were not available, collaborators in the countries were asked to provide existing data and state the time frame for each indicator.

Additional literature searches were conducted at the WHO Collaborating Centre for Capacity Building in HIV Surveillance in Zagreb using Pubmed and Embase. The following medical subject heading (MeSH) terms were used for PubMed, while the same terms were used as keywords in Embase: “Homosexual, Men” OR “Homosexual” which were cross-referenced with the keyword (AND) “HIV” OR the MeSH term “Human Immunodeficiency Virus” and the keywords “Eastern Europe”, “prevention” and “stigmatization” and limited to reports in the English language, published since 1 January 2000.

The EuroHIV report was used to assess the HIV cases reported among MSM in the CCEE.5 Data on coverage with HIV prevention interventions were collected from the most recent proposals for the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM) grants as this was considered the source that could provide the most comparable data and because it was not possible to find good quality data on the prevention responses.

RESULTS

The country reports were received from all countries contacted except Azerbaijan and Lithuania.

Estimates of the frequency of male same-sex behaviour

Data on the prevalence of male same-sex sexual activity were available only in a few countries. They were assessed in surveys using different indicators, which limits the comparability. One per cent of men in Slovenia12 (age group 18–49) reported ever having homosexual intercourse while 4% of men in the survey in Latvia reported this (age group 15–49).13 In Serbia, 2.4% of men aged 15–49 reported having sexual intercourse with another man in the past 12 months.14

HIV case reporting

From the start of reporting to EuroHIV, until the end of 2006, 3101 newly diagnosed HIV infections in MSM were reported in central Europe and 1828 in eastern Europe, out of 18 253 and 343 047 infections diagnosed among men, respectively. This is compared with 84 561 cases among MSM in the 23 countries of Western Europe, out of 179 135 HIV infections diagnosed among men.5

Table 1 shows the proportional contribution of key HIV transmission groups in the total number of newly diagnosed HIV cases reported in 2006 by EuroHIV.5 A high proportion of cases of homosexual transmission were found in the countries of central Europe, particularly in Slovenia, the Czech Republic, Croatia, Slovakia, Hungary and Serbia. A low proportion of cases attributable to homosexual transmission was reported in eastern Europe.

HIV and STI prevalence

The most recent data on HIV and STI prevalence among MSM are shown in table 2 and presented according to the robustness of the surveillance methods, starting with respondent-driven sampling (RDS) and time-location sampling (TLS) to convenience sampling. No HIV and STI prevalence data among MSM are available in Bosnia and Herzegovina, Cyprus, the Czech Republic, Hungary, the former Yugoslav Republic of Macedonia, Romania, Serbia, Tajikistan or Turkmenistan. As shown in table 2, results of HIV prevalence studies vary considerably. In the cities of four countries with data from RDS surveys, HIV prevalence is close to or higher than 5% (Croatia,15 Estonia,16 Georgia17 and Ukraine18). No cases of HIV were detected in surveys only using convenience sampling in Belarus,19 Bulgaria20 and Kazakhstan21 and the Russian city Tomsk.22 The highest levels of HIV were found in convenience sample surveys in Tashkent, Uzbekistan (10.8%)23 and in Odessa, Ukraine (23.2%).18

Several countries such as Bulgaria,20 Croatia,15 Georgia,17 Kyrgyzstan,24 the Russian Federation (Yekaterinburg)22 and Turkey25 have high lifetime syphilis prevalence among MSM (table 2). Prevalence data on herpes simplex virus type 2 (HSV-2) among MSM are only available from a study in Croatia, indicating a prevalence of 9.4%.15 Data on gonococcal infection among MSM were found in only two countries: Croatia, rectal gonorrhoea, 13.2%15 (n = 360) and Turkey, 3.0% (n = 166).25 Prevalence of rectal Chlamydia trachomatis infection was measured only in Croatia, and was found to be 9.0% (n = 360).15 Data on the prevalence of antibodies to hepatitis B surface antigen (table 2) suggest that a vast majority of MSM are susceptible to hepatitis B. The prevalence of hepatitis C infection in MSM in several countries (Croatia,15 the Republic of Moldova,26 the Russian Federation27 and Uzbekistan23) was higher compared to results among non-injecting MSM in studies carried out in the United Kingdom (0.9%)28 and the United States (1.5%),29 which suggests transmission other than sexual.

More than one wave of HIV surveillance surveys is available in six countries, though small sample sizes and short time intervals limit the comparisons of data over time. HIV prevalence among MSM has increased in Yerevan, Armenia, from no cases found in a survey in 2002 (n = 50) to 2% in 2007 (n = 100).32 In Minsk, Belarus, none tested positive in 2005 (n = 170) and 2007 (n = 279), while HIV prevalence in 2006 was 0.2% (n = 372).19 In Latvia, a survey carried out in Riga in 1997 found a prevalence of HIV of 2% (n = 207), while in the survey carried out in 1998 in Riga and Liepaja HIV prevalence was 5.4% (n = 242).33 In the Republic of Moldova, the first HIV survey among MSM conducted in 2003 (n = 120) found HIV prevalence of 1.7%.26 In the subsequent surveys in 2004 (n = 120), HIV prevalence was 2.5%, while in 2007 (n = 94) it further increased to 4.8%.26 In surveys carried out in Kazakhstan in 2006 (n = 215) and in 2007 (n = 450), no cases of HIV were found.21 HIV prevalence monitoring among MSM in Slovenia has been carried out annually since 1996 by small-scale one-day surveys using convenience sampling. The sample size ranged from 82 to 136, with prevalence rates at 2.4% in 1996, 3.7% in 2005 and 2.1% in 2006.36

HIV-related behaviours

Table 3 shows some behavioural indicators and data on self-reported testing on HIV from the most recent surveys, and the time frames used to assess these indicators.

No behavioural data are available in Bosnia and Herzegovina, Cyprus, Latvia, Romania, Tajikistan and Turkmenistan. Surveys that assessed only behaviours and not HIV or STI prevalence were carried out in the Czech Republic (2003),37 Hungary (2001),38 Serbia (2006)39 and the former Yugoslav Republic of Macedonia (2005).40

Use of condoms at last anal intercourse ranges from 36.0% in Turkey25 to 87.7% in Moscow27 (table 3). Recent commercial sexual contacts were frequently reported in Belarus,19 Hungary,38 the Republic of Moldova,26 the Russian Federation31 and Ukraine18 (table 3). Reporting of ever-injecting drugs was the highest in Albania (27.1%),30 the Republic of Moldova (16.0%)26 and Nizhniy Novgorod (11.1%)31 in the Russian Federation. Current partnerships with women were frequently reported in surveys throughout the region, particularly in Albania (93.1% in the last six months),30 Armenia (30% in the last year),32 Bulgaria (45.7% in the last six months),20 Czech Republic (17% in the last six months),37 Hungary (26.0% in the last year),38 the Russian Federation (ranging from 17.2% in the last year in Moscow to 44.8% in the last year in Krasnoyarsk),27 31 Turkey (35.5% in the last year),25 Ukraine (22% in the last year)18 and Uzbekistan (38% in the last six months).23

HIV prevention interventions among MSM in eastern Europe

As described in the country reports, most HIV prevention services are currently provided by community-based organisations that have a small number of staff and lack capacities and funds to implement programmes more frequently and on a larger scale. More comprehensive prevention responses started at the beginning of the 2000s, particularly in the 16 countries that received funding from the GFATM. HIV-related prevention services among MSM consist of condom distribution, outreach work, internet-based counselling, face-to-face counselling and the distribution of educational materials. Country collaborators often mentioned the lack of staff, inadequate financial resources and unsupportive legal environments as important barriers to work. STI services are reportedly available in a few countries, though it is unclear the kind of tests and specimens that are used to make a diagnosis. Self-reported testing on HIV (table 3) is low, although in the last couple of years there has been an increase in voluntary counselling and testing (VCT).

The data on the coverage with prevention activities and the impact of prevention on HIV transmission were difficult to find. To describe the estimated coverage of MSM with prevention services we therefore used the most recent proposals for the GFATM grants because of the presumed comparability of such proposals and the lack of other data sources (table 4).41^–56 In countries that are not eligible for GFATM funding, data on the coverage with HIV prevention interventions were not available. The coverage in the GFATM proposals is described either as an absolute number of MSM or a proportion of the estimated MSM population reached with prevention activities. In the majority of countries, HIV interventions include behavioural change communication and the provision of condoms, lubricants, educational materials and injecting equipment during outreach. The baseline coverage at the time of grants’ submission was very low, averaging around 1–2%, with the exception of Romania and Serbia, where it was slightly higher (table 4).

Stigma and discrimination against MSM

In a number of the countries reviewed, homosexual relationships were punishable by law until the mid-1990s, or even later, while in Turkmenistan and Uzbekistan homosexuality is still illegal and punishable by imprisonment. Available data from surveillance surveys or special studies show high levels of discrimination and violence towards gay people. Physical violence as a result of sexual orientation ranged from being reported by 10% of MSM in Georgia17 to 23% in Turkey.25 Psychological violence is more common, and ranges from 17% as reported by MSM in Hungary57 to 70% in Serbia.58 In some instances, the utilisation of healthcare and psychological services by MSM is hindered by a lack of healthcare professionals’ knowledge of the sexual health needs of MSM.59 60

DISCUSSION

Current evidence on the size of the HIV epidemic among MSM shows that HIV surveillance needs to focus on this group and embrace methods beyond case reporting. Continuing support from the GFATM and other donors for surveillance and prevention is essential, but has to be accompanied with destigmatisation (legal and societal) and changes in public health responses in order to meet the challenge of the evolving epidemics.

There is a substantial diversity between and within the CCEE with regard to the dynamics of HIV transmission among MSM. This is possibly due to differing times when HIV first occurred in the populations and the interplay of societal factors, injecting drug use, mobility and sex work.

Available data on the frequency of male-to-male sexual behaviour should be interpreted taking into account possible biases in reporting such behaviours. Low estimates of homosexual behaviour in Slovenia12 were explained by participation bias and men’s fear of disclosing homosexual behaviour, while the country report in Latvia13 mentions that homosexuality is stigmatised and respondents might not have reported truthfully.

HIV case reporting is the mainstay of HIV surveillance in the CCEE. However, its coverage and completeness vary among the countries, depending on the existence and accessibility of VCT services, reporting practices and the extent to which MSM are willing to disclose the mode of transmission when diagnosed with HIV. Most of the prevalence surveys used convenience sampling, which limits the generalisability of these data. Few clinic-based prevalence data are available, including data from VCT centres and STI clinics. STI clinics that serve MSM population are rarely available in the CCEE, which explains the poorer availability of HIV and STI data from these sites, as well as rarely available prevalence estimates of STIs, such as gonorrhoea, Chlamydia trachomatis and HSV-2 from population-based surveys.

An important finding from this review is that there is a substantial discrepancy between the cumulative number of HIV cases reported among MSM in the national case reporting systems and the HIV prevalence reported in surveys, particularly in eastern Europe (Armenia, Georgia, the Russian Federation, Ukraine and Uzbekistan). For example, Ukraine has a reported prevalence in four cities in the range of 4.4%–23.2%,18 and the total number of MSM cases reported since 1987 is 110.5 In the context of such under-reporting, population-based HIV surveys have better public health utility in estimating the burden of HIV than the case reporting system, and can reach those parts of high-risk groups that have poor access to services. Public health benefit of surveillance would be maximised by having multiple data collection methods, in particular targeted population-based surveys and clinic-based sentinel surveillance.61

No studies were found that provided estimates of recent HIV infections. Nor were data found on serosorting and disclosure of HIV status. Male sex work remains unexplored in the region in spite of the high prevalence of commercial sexual contacts, particularly in the countries of eastern Europe. It is almost impossible to interpret STI epidemiology among MSM in the CCEE as there is a striking lack of data. Without knowing the extent of these problems, it is challenging to address them in the programmatic responses.

Barriers to implementation of HIV prevention programmes

The key barriers to implementation of prevention are the stigmatisation and discrimination of MSM and the subsequent difficult access to this community, coupled with weak surveillance on which effective prevention interventions should be based. The review of the GFATM proposals was undertaken because it was not possible to find well described activities of the programmatic responses and their results. The data on the estimated levels of coverage with HIV prevention from the most recent GFATM proposals, which are all written in the early and mid-2000s, were either low or not available. Underestimating the coverage for MSM and modest funding in many countries cannot make an impact on the HIV epidemic among MSM. Though stigma and discrimination are mentioned as the main obstacles to prevention among MSM, we were not able to find a comprehensive approach to addressing stigma and discrimination. Another key challenge to implementation of more comprehensive HIV prevention is the almost complete lack of evaluation of HIV prevention interventions in experimental studies in the CCEE. Only two randomised controlled trials evaluating the effect of HIV interventions done in the region were carried out, in the Russian Federation and Bulgaria.62 63

The lack of MSM-specific STI screening services contributes to STIs being undiagnosed, particularly in the light of the high proportions of asymptomatic rectal infections.64 Along with the development of STI services for MSM, there is a need to scale up behavioural interventions and increase HIV testing and hepatitis B vaccination uptake.65 66

Needs for further research

Further research should determine the size and characteristics of the MSM population, more robustly characterise the nature of the MSM HIV and STI epidemics and describe the nature of sexual networks of bisexual men and MSM who inject drugs, and assess their potential impact on population transmission of HIV. Research is also needed to describe the policy and legislative context within which these epidemics are occurring. Data on the extent and characteristics of discrimination and stigmatisation should be collected in special surveys or in surveillance surveys. Operational research is needed to explore the policy and programmatic issues that hinder or enable the implementation of HIV prevention services.

Limitations of data sources

Interpretation of survey data in the CCEE is limited by small sample sizes, short time-intervals of surveillance and the fact that population-based surveys have only recently started to be implemented. In the context of high stigmatisation of MSM, measurement, selection and participation bias and the social desirability bias—that is, when respondents reply in a manner that will be viewed favourably by others, have to be taken into account when interpreting findings. Ability to extrapolate the findings across the CCEE is limited because surveys are geographically limited to urban areas.

In conclusion, the HIV epidemic among MSM in the countries of central and eastern Europe, as in many other parts of the world, is hidden and prevention responses and surveillance are underfunded. The epidemic potential differs, underscoring the strategic importance of focusing prevention programmes on those areas where the potential for the expansion of the epidemic is more severe. This is particularly relevant for countries such as Croatia, Estonia, Latvia, the Republic of Moldova, Poland, Ukraine, Uzbekistan and some cities in the Russian Federation where the HIV epidemic among MSM is already well established. Programmatic efforts should improve HIV and STI surveillance and ensure that HIV prevention programmes are evidence-based and appropriately funded.