Until 1995, central and eastern European countries including the Asian republics of the former Soviet Union, did not seem to be confronted with a major HIV threat. The total number of infections for the entire region with its more than 450 million inhabitants , was estimated at less than 30000 [2,3]; this is compared with 474000 individuals living with HIV in western Europe, and 12.9 million, 4.2 million and 1.2 million in Africa, Asia and Latin America , respectively. [The United Nation (UN)AIDS eastern European region includes the following countries: Albania, Armenia, Azerbaijan, Belarus, Bulgaria, Croatia, Czech Republic, Estonia, Georgia, Hungary, Kazakstan, Kyrgystan, Latvia, Lithuania, Former Republic Yugoslav of Macedonia, Republic of Moldova, Poland, Romania, Russian Federation, Slovakia, Slovenia, Tajikistan, Turkey, Ukraine, Uzbekistan, Yugoslavia (Serbia and Montenegro).] Within Europe, the 15 western European countries of the European Union accounted for an overwhelming 92.8% (149403 out of 160982) of all AIDS cases reported by December 1995 .
Two years later, the situation had changed dramatically. In December 1997, with about 190000 infections, UNAIDS and WHO estimated the number of persons living with HIV/AIDS in eastern Europe to have increased more than fivefold. Two-thirds of the total number of infections was believed to have occurred during the last 12 months . While the number of reported AIDS cases in western Europe had started to decrease for the first time, this was not true for eastern Europe where dramatic increases have to be expected .
This paper describes the development of the eastern European HIV epidemic in 1996 and 1997 in more detail. Subregional commonalities and differentials are described, and possible determinants of the epidemics mentioned. The potential implications of the epidemiological findings for policy-making, both within the region and by the international community, are discussed.
Background and methods
The former communist countries of eastern Europe have inherited an extensive HIV (and syphilis) testing and reporting infrastructure, and policies aiming to control HIV through case finding and surveillance of those infected. Mass HIV screening was introduced by 1987 in most countries, and has traditionally involved both low risk groups (e.g., pregnant women, blood donors, occupational groups, hospital patients) and particularly vulnerable groups [e.g., injecting drug users (IDU), prisoners, sexually transmitted disease (STD) patients]. Since the mid-1990s, testing policies have gradually changed in that pregnant women and specific occupational groups are no longer routinely tested. The total number of tests has decreased only very slightly in most countries, however, as many health workers still perceive mandatory screening as a main HIV prevention and control tool. Exceptions are countries such as Azerbaijan and Tajikistan, where the number of tests has declined significantly due to shortages of test kits. The routine screening of high risk groups including STD patients, prisoners and IDU in contact with health or police authorities is continued throughout the region.
In most countries, including the 15 successor states of the Soviet Union, HIV case reporting traditionally involves two stages, the recording of screening test results and the referral to a health institution for history-taking, advice and official registration. The clinic visit is also important because, on the basis of the interviews there, patients are re-classified according to transmission category. For instance, prison inmates, occupational groups and pregnant women are re- classified as IDU or as infected via homo- or heterosexual intercourse. Except for the Ukraine, which switched to laboratory reporting during 1996 and 1997, HIV reporting is usually based on cases registered by physicians. Syphilis screening and reporting follows similar procedures.
Given the relative uniformity and stability of case testing and case reporting procedures in the region, an analysis of the number of HIV infections detected through case finding and screening, and the proportions of HIV-positive cases among groups routinely tested was expected to provide a reasonable insight into patterns and trends of the epidemic.
In early 1998, national AIDS programme managers and epidemiologists were contacted and requested to provide national HIV surveillance data, including breakdowns of reported HIV cases by transmission category and year of reporting. In addition, UNAIDS/WHO country fact sheets were reviewed and analysed, and this information was supplemented further with published HIV prevalence and syphilis case reporting data and information from unpublished UNAIDS travel reports and expert evaluation.
National HIV case reporting data were obtained from all countries in the region except for Romania and Yugoslavia (Serbia and Montenegro) for which only AIDS data were available. HIV data on transmission categories were received from all countries except Croatia, the former Yugoslav Republic of Macedonia, and Hungary. The cumulative number of HIV cases reported from the region had increased more than fivefold between 1995 and 1997, from 9111 to 46573, with IDU accounting for more than 50% of the cumulative total. The cumulative rates of reported HIV cases by 100000 population and proportions of IDU among HIV cases by individual country are given in Table 1. Syphilis data obtained from the WHO Regional Office for Europe in Copenhagen and UNAIDS fact sheets showed large differences between the countries of the former Soviet Union (where case reports had been rising sharply) and central Europe (where cases had remained few).
Western newly independent states and Russian federation
In the Ukraine, since 1988 and until 1994, only 40-80 new cases of HIV-infections were registered annually; these were mainly foreigners infected through sexual contacts. From March to April 1995, more than 1000 IDU in the southern cities of Odessa and Nikolayev were found to be infected. One year later, HIV infection among IDU was reported from all 25 regional capitals, with Odessa, Nikolayev, Crimea, Donetsk and Dnipopetrov being the most affected cities/regions . The total number of diagnosed HIV infections increased to more than 12000 in 1996, and more than 25000 in 1997  - more than one-half of the regional cumulative total. More than one-half of all new HIV cases were registered IDU.
By the end of 1995, HIV infection among IDU in contact with narcology services and police in Odessa and Nikolayev reached 31% and 57%, respectively . Nationwide, seropositivity among registered IDU increased from virtually zero before 1995 to 2% in 1995 and 7% in 1996 . the prevalence of HIV among male STD patients increased from 0.005% in 1993 to 0.027% in 1995, and 0.37% in 1996. Prevalence among prisoners and military personnel in the Ukraine (as well as in Russia) also increased in 1996 and 1997; the vast majority of those testing HIV- positive were IDU.
Prevalence among pregnant women in Ukraine increased from 0.0002% in 1993, before the injecting drug use-associated outbreaks, to 0.002% in 1995 and 0.05% in 1996. It was highest in Nikolayev, with 0.24% of all pregnant women testing HIV-positive. Prevalence in blood donors increased sharply, from 0.05 per 100000 in 1993 to 0.12 in 1994, 2.09 in 1995 and 37.60 per 100000 in 1996 .
Few HIV infections had been reported in Belarus until May 1996 when mass screening among IDU in Svertlogorsk in the southern Gomel district revealed that 632 (almost 50% of those tested) were infected by HIV. Over 70 cases of HIV infection were also detected in a nearby town, Zhlobin. By the end of that year, HIV infections among IDU were reported from all administrative regions of Belarus . By the end of 1997, the total number of HIV infections had increased to almost 1800, the vast majority of these were IDU (Fig. 1).
In Russia, the HIV epidemic went through several distinct phases, starting with the detection of the first infected person who had contracted the virus in a high-prevalence African country in 1987 and initiated a chain of STD. In 1989, an outbreak of nosocomial HIV infections occurred among about 250 children in the southern Russian republic of Kalmykia, which was followed by a predominantly ‚gay‚ epidemic with around 40% of all cases attributed to homosexual transmission between 1991 and 1995 . At the end of 1995, out of a total of 1062 reported cases of HIV-infection, only seven were IDU . Starting from 1996, the number of case reports among IDU increased dramatically. In 1996, a total of 1546 new HIV infections were reported, and in 1997 this had increased to 4399, about 60% of them IDU. By 1 January 1998, the total number of cases had increased to more than 7000 (Table 1). Prevalence among pregnant women increased from 0.02 per 100000 in 1993 to 0.15 in 1995 and 0.52 in 1996 .
With more than 1500 reportedcases, 79% of them IDU, reported by 1 July 1997, the northwestern enclave of Kaliningrad was by far the most affected region of Russia . Out of a small sample of 103 sex workers arrested in the streets of the city of Kaliningrad by officials of the Ministry of Interior, 33 (32%) were HIV-positive IDU. According to the same source, 82% of female HIV-infected patients treated at the Regional AIDS Centre make their living from commercial sex (SP Krishtopa, in a letter to UNAIDS, 11 September, 1997). Other Russian cities with rapidly spreading HIV epidemics (with more than 100 injecting drug use- associated cases) are Krasnodar, Nizniyi-Novgorod, Rostov, Tver, and Saratov in central and southern Russia. Smaller numbers were reported from several other cities and most other regions including Siberia and the far north (V. Prokrovsky, Russian AIDS Center, personal communication, 1997).
Until late 1996, Moldova had reported between one and seven HIV-positive cases among its citizens, annually. Only one case of HIV infection among IDU was reported by the end of 1995. In 1996, a total of 55 new infections was reported, mainly in the capital Chisinau, 38 of them among IDU . In 1997, the cumulative total increased to 471, of which 362 were confirmed IDU.
Sexually transmitted diseases
In virtually all parts of the former Soviet Union, including Russia, Ukraine, Moldova, Central Asia and the Baltic States [10-12] reported syphilis rates have been increasing rapidly since 1991, reaching rates of up to 100 per 100000 in the Baltic states and more than 500 per 100000 population in Kaliningrad and the Russian far east (A. Gromyko, WHO-EURO, personal communication, 1998). Gonorrhoea rates have not increased to the same extent, presumably because many patients avoided attending the under-resourced public services and have resorted to semi-private care or self-medication. In the three Caucasus countries, STD seem to be spreading rapidly, too [13-15] (Table 2).
The Baltic states, the Caucasus republics and central Asia
With the only exception of Kazakstan, the other eight newly independent states in the Caucasus and in central Asia, and the three Baltic states, have all reported less than 100 cumulative HIV-infections each. Forty-eight and 38 cumulative cases were reported from Georgia and Azerbaijan, respectively, and 65 cases from Armenia. Most cases were due to heterosexual or unknown mode of transmission, but the reported number of infected IDU was starting to increase. In Georgia, virtually all new infections (except for one blood recipient infected much earlier) since 1996 were detected among IDU. In Azerbaijan, HIV infection among drug users had been unknown before 1995, but by mid-1997, accounted for more than a third of the cumulative total. In Armenia where the total number of HIV cases increased from three in 1996 to 65 in 1997, 14 infections were reported among IDU  (Table 1).
In the three Baltic States, most HIV infections had been detected in homosexual men, before infections among IDU started to occur. In Lithuania, the total number of cases of HIV infection reported as of January 1998 was 83. Whereas before 1997 it was mostly homosexual men and sailors who were affected, IDU have since accounted for 30 cases. In Latvia, a total of 76 persons including only one IDU had been diagnosed as HIV-infected by September 1997, but by the end of that year 12 new cases among IDU were reported. From Estonia, with a 1.5 million population, the smallest country in the region, no injecting drug use-associated HIV infection has yet been reported (Table 1).
In Central Asia, Kazakstan is the most affected country. A total of 69 cases of HIV infection had been reported by mid-1996, when following the detection of HIV infection among imprisoned IDU in the city of Temirtau in the northern Karagandy oblast, testing among IDU was intensified. By 31 December 1997, 447 HIV-positive persons had been newly registered, 422 (94%) of them IDU, resulting in a cumulative total of 516 reported HIV/AIDS cases.
By the end of 1997, Turkmenistan had registered only one HIV-positive citizen, and Tajikistan and Kyrgystan four and two, respectively. Foreigners, including citizens of neighbouring newly independent states are registered separately. In Uzbekistan, despite the routine testing of 22 different population groups and the conduct of several million tests each year, only a total of 19 HIV infections among Uzbek citizens was reported. In addition, 26 foreigners were found to be HIV-infected and were deported.
Except for Poland, between 20 and 120 new HIV infections were reported from the former communist countries of central Europe between 1995 and 1997. Homosexual and bisexual men continued to be the most severely affected population group. Between 49% and 57% of all HIV cases reported from Croatia, the Czech Republic, Hungary, and Slovenia were due to homosexual transmission, whereas in Slovakia 66% (36 out of 55) local HIV-infected citizens, were homosexual men.
Only two IDU were registered as HIV-positive each in Hungary and Slovakia, 10 in the Czech Republic, and nine and 11 in Slovenia and Croatia, respectively. Very low HIV prevalence of 0 and 0.1% among IDU was confirmed in surveys in Slovenia, among methadone-maintenance patients, and in the Czech Republic, among long-term addicts [17,18]. In Croatia, among 742 IDU tested, none were found to be HIV-positive. In the Slovak Republic, one out of 1400 IDU tested HIV-positive. HIV prevalence among other population groups, both of high and low risk, have also been low in central Europe [6,19,20].
In Poland, IDU account for the majority of HIV cases. The cumulative total number of HIV infections reported by the end of 1997 was 4990, 3267 (66%) of them IDU. The number of reported HIV infections among IDU increased from one in 1988 to 411 in 1989, to a peak of 653 in 1990. Infections then decreased to annual numbers between 205 in 1993, and 342 in 1996. There is anecdotal evidence that HIV-prevalence among IDU varies widely according to care institution and town, with up to 50% of IDU (29 out of 58) reported as HIV-positive in a study in Szeczecin and 60% in an equally small convenient sample in Lublin [21-24].
All analyses of HIV data from Polish population groups other than IDU have confirmed very low and stable prevalence rates [22,25,26]. Between 1992 and 1996, between only 12 and 19 out of approximately one million blood donors annually were found to be infected. Few studies among sex workers have been carried out. In 1995, among 539 female non-IDU who had stated having worked in the sex industry, one was found to be infected . One out of 44 sex workers tested in a small survey in Katowice was infected (M. Beniowski, personal communication, 1998).
The number of syphilis cases reported from central European countries remained low, with annual rates per 100000 population varying between less than one in Croatia, to 1.8 in Slovenia, 2.1 in Slovakia and four in Poland.
The Balkan countries differ from the rest of eastern Europe in that heterosexual transmission is more frequently reported. Except for Yugoslavia (Serbia and Montenegro), injecting drug use-associated HIV infection has remained the exception.
In Bulgaria, 217 HIV infections had been reported as of 1 September 1997, 169 (78%) of them due to heterosexual transmission. The annually reported number of HIV cases appeared to be increasing, from 11 and 12 in 1992 and 1993; to 18 and 14 in 1994 and 1995; and 34 and 30 in 1996 and 1997. The number of newly-reported syphilis cases in 1996 was 2272, a rate of 27 per 100000.
As of June 1997, Romania had reported a total of 4725 AIDS cases, 4226 of them among children. In 1988 and 1989, outbreaks including more than 4000 HIV infections among children in orphanages were reported. Epidemiological investigations showed that the majority of ‚nosocomial‚ infections were due to therapeutic ‚microinfusions‚ and multiple injections with non-sterile equipment. During the early 1990s, the yearly number of newly-diagnosed paediatric cases rapidly decreased to less than 10 per year of children born in 1993 and later, although the risk of HIV transmission in health care settings persisted and children living in institutions continued to be subjected to more than 100 injections during their first years of life .
In contrast, the number of reported adults diagnosed with AIDS in Romania steadily increased from less than 10 per year during 1985-1987 to 100 in 1995 and 113 in 1996. Heterosexual transmission accounted for 56% of all reported cases. Only one case of HIV infection among IDU had been reported by 1997; among STD patients in Bucharest, 0.1% had tested HIV- positive .
In Turkey, annually reported HIV infection appears to be slowly rising, from around 40 infections per year between 1987 and 1990 to between 60 and 90 infections in 1992 and 1995, 118 in 1996, and 145 in 1997. Forty-four per cent of all cases were reportedly due to heterosexual transmission, 11% were IDU, mainly from Istanbul; and 9% homosexual and bisexual men, with most of the remainder unknowns. Among an estimated 2000 unregistered non-Turkish (mainly Romanian, Ukrainian and Russian) sex workers in Istanbul, four were found to be HIV-positive, and among the routinely tested 3000 registered (Turkish) sex workers, no infection had been detected . In a small sample of 44 HIV-infected persons in Ankara, one-half were either Turkish citizens abroad or stated that they had had sexual contacts with tourists or foreign prostitutes .
In Albania, the first HIV infection was only detected in May 1993. Since then a total of 33 infections have been reported by the end of 1997, most of them believed to be due to heterosexual transmission. HIV infection among IDU had not yet been described. Two out of three of registered HIV-infected persons had reportedly contracted HIV whilst outside the country.
From three of the successor states of the former Yugoslavia [Bosnia-Herzegovina, the Former Yugoslav Republic of Macedonia and Yugoslavia, (Serbia and Montenegro)] little HIV data were available for this review. Between 1987 and 1992, 44% of IDU tested in Belgrade had been found to be HIV-positive . As of December 1996, among 605 (adult) AIDS cases, 52% were due to infections among IDU. In Macedonia, a total of 39 HIV infections were reported as of December 1997. Among 21 AIDS cases, five were reportedly due to injecting drug use.
Traditionally, only AIDS (and not HIV) cases are routinely reported to the European Centre for the Monitoring of AIDS, WHO, and UNAIDS. The European Centre also maintains the European HIV prevalence database. In order to describe current patterns and trends of HIV in the eastern European region, this paper has drawn on both existing reporting mechanisms and published data as well as on unpublished UNAIDS mission reports, expert situation assessments and national HIV case reporting data obtained through personal communication with national epidemiologists. Prevalence data from anonymous unlinked testing are scarce in the region and AIDS case reporting does not grasp fully the nascent nature and the rapid dynamics of the epidemic. HIV case reporting is therefore a useful additional monitoring tool, and it has been recommended to establish HIV case reporting at the European level in addition to the well-established AIDS reporting system .
The data presented suggest subregional differences in the magnitude of the HIV epidemic and in the importance of drug use for the spread of the virus. HIV is spreading rapidly among IDU in the eastern part of the region, in Ukraine, Russia, Belarus and Moldova, with a similar pattern emerging in the adjacent Baltic States and in the Caucasus. With the exception of Kazakstan, the number of cases reported is still very low in central Asia. Central Europe has so far escaped both the more extensive spread among homo/bisexual men and among IDU as seen in western Europe during the late 1980s, as well as the more recent drug use-driven epidemic in the east. Homosexual transmission accounts for a large proportion of cases in this subregion (with the exception of Poland), whereas in the Balkan countries, heterosexual transmission is reported relatively frequently. Table 2 summarizes the main epidemiological features of the various subregions.
Although mass screening has a low cost-benefit ratio and is ethically controversial, we believe it has allowed the emerging epidemics among IDU in Russia, Ukraine, Belarus and Kazakhstan to be detected at a relatively early stage. The available data suggest that the introduction of HIV into the IDU population of the southern Ukrainian port cities of Odessa and Nikolayev may have occurred as recently as 1994 or early 1995. In Russia in 1996 and 1997, researchers were able to confirm recent seroconversions in many of the persons newly diagnosed as HIV-positive (V. Prokrovsky, Russian AIDS Center, personal communication, 1998). The findings that AIDS cases were still generally few, and the proportion of IDU among AIDS cases, which had been virtually zero in 1995, increased more slowly than among HIV cases  also confirmed the recent dynamics of the epidemic among IDU.
Nevertheless, case reporting has probably not made visible the full extent of the epidemic spread among IDU in the former Soviet Union. An unknown proportion of IDU is not in contact with narcology services and police, and except in the Ukraine, where a system of reporting based on laboratory diagnosis was adopted, cases detected through screening are unlikely to be all officially registered. Policies concerning the screening of high-risk populations have hardly changed, however, and are unlikely to have affected overall reporting patterns and trends.
Several factors seem to have been fuelling the HIV epidemic associated with injecting drug use in the former Soviet Union. They include migration, increased drug supply and demand, and specific local drug production and consumption patterns. The importance of transport and economic links within the region, for both the trafficking of drugs and the spread of HIV has not yet been systematically examined. The Ukrainian (and Russian and Georgian) Black Sea coast is a popular summer holiday destination for the entire newly independent states, and several case histories of Russian persons testing HIV-positive have suggested a possible link between their infection and recreational periods in southern Ukraine. Krasnodar, where the first major scale injecting drug use-associated HIV outbreak in Russia was reported, is a northern Caucasus city with good transportation links with several Black Sea harbours. In Tumen, a medium sized town in western Siberia, all detected HIV-infected IDU were temporary workers from the Ukraine (V. Prokrosvsky, Russian AIDS Center, personal communication, 1998).
There has been a massive increase in the trafficking and use of illicit drugs throughout the former Soviet Union. Cultivation of opium poppies is traditional in the central Asian republics, Ukraine and parts of Russia. In Afghanistan and three of the central Asian republics bordering it (Uzbekistan, Tajikistan and Turkmenistan), poppy crop areas and opium production have expanded rapidly in recent years. Seizures of drugs other than raw opium, including heroin, cocaine and amphetamines have also been reported by national authorities in Russia and other countries in the region . Despite shrinking resources, the Ukrainian and Russian narcology and police services have registered increasing numbers of illicit drug users in recent years.
Specific drug use behaviours facilitate the spread of HIV. Sharing of needles and syringes has been described as common in reports from Moscow, Kaliningrad, Odessa and Poltava, among other cities ( A. A. Protopopov, O. L. Zapopozhets, unpublished data). Furthermore, many IDU re-fill their small syringes by front-loading from the dealers‚ syringes, many of them IDU themselves who intermittently inject from the same solution (G. M. Pavsky, unpublished data). HIV might also be introduced into the injecting drug during the preparation process itself. There are anecdotal reports from several cities including Kaliningrad, Moscow and Odessa, that blood is intentionally added by the dealers (often IDU themselves) to the drug solution during the preparation process as a ‚cleansing‚ substance, as it is believed that the red blood cells neutralize toxic reagents that are used during the production process.
Unlike Ukraine and Russia where several thousand HIV infections among IDU have occurred within a few months, Poland has been spared from an explosive spread of HIV. One possible explanation would appear to lie in drug demand patterns and their underlying societal determinants, including the degree of social stress experienced by the various societies. The few data available would suggest that Polish (and other central European) cities have significantly smaller IDU populations than Ukrainian and Russian cities with comparable population size [7,18,24,34,35]. Local drug preparation and consumption patterns may also differ.
Further projections of the epidemic in the region depend on the link between the currently mainly injecting drug use-driven HIV epidemic and its potential spread, via heterosexual intercourse, into the general population. Some data suggested a beginning spread of HIV into low risk populations, such as pregnant women and blood users, whereas others were more difficult to interpret. Increases in seroprevalence among pregnant women and blood donors were reported from both Russia and the Ukraine, for instance, but in the Ukraine, the number of pregnant women actually tested annually declined, and it cannot be excluded that testing has become more targeted towards women at higher risk for HIV. Furthermore, it appears that those testing HIV-positive were mainly IDU (Y. Kruglov, personal communication, 1998). The selection of blood donors has changed considerably throughout the Soviet Union. The formerly semi-voluntary system, where for instance, large numbers of factory workers were encouraged to donate blood, has not yet been replaced by a coherent system of voluntary donations; and currently, a substantial proportion of donors are paid. Rising trends among male STD patients and high prevalence among drug-injecting sex workers might, however, indicate an increased risk of HIV spreading from IDU to persons engaging in high-risk sexual behaviour.
The epidemiological findings presented here might be combined with vulnerability assessments to develop a tentative classification of countries and subregions, to assist the international community in deciding where to focus assistance. Clearly, Russia as the largest, and the Ukraine as by far the most affected country in the region, would appear to deserve the strongest possible support from the international community in their efforts to prevent further spread. Nevertheless, confronted with the double epidemics of drug use and HIV, policy-makers face difficult decisions with regards to the choice of type and mix of interventions and allocation of resources. To concentrate solely on drug supply and/or drug demand reduction - as has been the declared policy of most national governments throughout the region until very recently - might imply an unacceptably high risk of an epidemic spreading unabatedly. UNAIDS together with its co-sponsors and partner agencies has therefore embarked on an initiative to promote a harm minimization approach to HIV prevention among IDU in eastern Europe, in order to balance the existing policy and programme mix . An overview of attempts to operationalize harm reduction approaches in developing and transitional countries is provided elsewhere .
Another dilemma is posed by uncertainties about the future course of the epidemic. Non-repressive approaches to HIV prevention among IDU face considerable resistance, and may not become effective sufficiently fast to prevent a substantial spread of HIV outside of this group. A much more comprehensive approach therefore needs to be adopted. This should comprise not only HIV prevention among the most vulnerable segments of the population, such as IDU, sex workers and men who have sex with men, but also programmes to sensitize the population at large and decision-makers, in order to create an enabling environment. Such a broader strategy must also encompass measures of improved management and prevention of STD other than HIV.
In contrast to the newly independent states, the former communist countries of central Europe seem to be much less vulnerable to a large scale spread of HIV. Economic performances are better and fewer young people are unemployed, inject drugs or work in the sex industry. STD rates have remained at low levels. In addition, prevention programmes appear to be more sophisticated than those in eastern Europe. For instance, presentations made during internationally-attended conferences in Warsaw and Prague in 1996 revealed that several countries including the Czech Republic, Hungary and Poland had already started implementing outreach and peer programmes to prevent HIV among vulnerable populations .
The vulnerability to HIV of the Balkan countries, some of them among the least developed countries in Europe, needs further assessment. The Balkans are increasingly becoming drug traffic routes between Asia and western Europe. Although heterosexual transmission appears to be the most common mode of transmission in this subregion, available case reporting data do not allow one to ascertain that this pattern reflects reality. Homosexuality is still severely stigmatized, probably much more than in central Europe, and is often legally restricted. Homosexual transmission is therefore likely to be underreported. Most HIV-infected women seem likely to be partners of men who have lived in higher prevalence areas (e.g. Turkish men in Germany or Albanian men in Italy) or of men who had sex with men. Together with prevention among IDU, programmes and projects aiming to prevent infection among migrant populations and men who have sex with men should therefore become a priority in this subregion.
A considerable amount of experience with the HIV epidemic has been accumulated during the past 15 years worldwide. Eastern Europe is perhaps the last world region to be confronted with rapidly spreading HIV epidemics. Governments and the international community should not allow another disaster to happen.