Global Overview

The total number of people living with the human immunodeficiency virus (HIV) rose in 2004 to reach its highest level ever: an estimated 39.4 million [35.9 million–44.3 million] people are living with the virus. This figure includes the 4.9 million [4.3 million–6.4 million] people who acquired HIV in 2004. The global AIDS epidemic killed 3.1 million [2.8 million–3.5 million] people in the past year.

The number of people living with HIV has been rising in every region, compared with two years ago, with the steepest increases occurring in East Asia, and in Eastern Europe and Central Asia . The number of people living with HIV in East Asia rose by almost 50% between 2002 and 2004, an increase that is attributable largely to China’s swiftly growing epidemic. In Eastern Europe and Central Asia, there were 40% more people living with HIV in 2004 than in 2002. Accounting for much of that trend is Ukraine’s resurgent epidemic and the ever-growing number of people living with HIV in the Russian Federation. Sub-Saharan Africa remains by far the worst-affected region, with 25.4 million [23.4 million–28.4 million] people living with HIV at the end of 2004, compared to 24.4 million [22.5 million–27.3 million] in 2002. Just under two thirds (64%) of all people living with HIV are in sub-Saharan Africa, as are more than three quarters (76%) of all women living with HIV. The epidemics in sub-Saharan Africa appear to be stabilizing generally, with HIV prevalence at around 7.4% for the entire region. But such a summary perspective hides important aspects. First, roughly stable HIV prevalence means more or less equal numbers of people are being newly infected with HIV and are dying of AIDS. Beneath the apparent constancy of steady prevalence levels lie devastating realities—especially in southern Africa, which accounts for one third of all AIDS deaths globally. Second, the epidemics in Africa are diverse, both in terms of their scale and the pace at which they are evolving. There is no single “African” epidemic. Some urban parts of East Africa display modest declines in HIV prevalence among pregnant women, while in West and Central Africa prevalence levels have stayed roughly steady at lower levels than in the rest of sub-Saharan Africa. National HIV data, though, hide much higher levels of infection in parts of countries, as Nigeria illustrates. Southern Africa, unfortunately, offers only slight hints of possible future declines in HIV prevalence. HIV prevalence in the Caribbean is the second-highest in the world, exceeding 2% in five countries, and AIDS has become the leading cause of death among adults aged 15–44 years in this region. Yet, a growing number of Caribbean countries are showing that the epidemic does yield to appropriate and resolute responses

Deaths among those already infected will continue to increase for some years to come even if prevention manage to cut the number of new infections to zero. However, with the HIV- Positive population still increasing, the annual number of AIDS deaths can be expected to also increase for many years to come.
 

About half of all children and young people who acquire HIV become infected before they turn 25 and typically die of the life threatening illness called AIDS before they turn 35. This Age factor makes AIDS uniquely threatening to children. By the end of 2001, the epidemic had left behind a cumulative total of 14 million AIDS orphans (those having lost one or both parents to AIDS before reaching the age of 15).
 

In 2002, an estimated 800,000 children aged 14 or younger became infected with HIV. Over 90% were babies born to HIV-Positive women who acquired the virus at birth or through their mothers breast milk. Of these, almost nine tenths were in Sub-Saharan-Africa.

In all parts of the world except Sub-Sahara-Africa, North Africa, and the Middle East, there are more men infected with HIV and dying of AIDS than women. Altogether, an estimated 2.2 million men aged 15-49 became infected during 2002, bringing the number of adult males living with HIV or AIDS by the end of the year to 19.4 million.

NORTH AMERICA, WESTERN AND
CENTRAL EUROPE

AIDS is affecting new sections of populations,
and an increasing proportion of people are becoming infected through unprotected heterosexual intercourse.

Some 64 000 [34 000–140 000] new infections occurred in North America and in Western and Central Europe in 2004, raising the number of people living with HIV in these countries to between 1.1 million and 2.2 million. Among young people 15–24 years of age, 0.1% of women [0.1–0.2%] and 0.2% of men [0.1–0.5%] were living with HIV by the end of 2004. Widespread access to life-extending antiretroviral treatment kept the number of AIDS deaths at between 15 000 and 32 000 in 2004. However, there are ample indications that prevention efforts are not keeping pace with the changing epidemics in several countries.

Sex between men and, to a lesser extent, injecting drug use remain prominent factors in the epidemics in these countries, but the patterns of HIV transmission are changing. New sections of populations are being affected, with an increasing proportion of people becoming infected through unprotected heterosexual intercourse.

In the United States of America the epidemic has altered demonstrably during the past decade. An estimated 40 000 people have been infected with HIV each year in the United States in the last ten years, but the epidemic is now disproportionately lodged among African Americans and is affecting much greater numbers of women.

In 2003, African Americans accounted for at least 25% of all AIDS cases, compared with 20% in 2001. That proportion could be higher, since the estimate was based on data collected in just 29 states. Although African Americans represent just 12% of the country’s population, over half of new HIV diagnoses in recent years have been among them (in 2002 the proportion was 54%, according to the most recent available data). Especially affected are African American women, who account for up to 72% of new HIV diagnoses in all US women. At the turn of the century already, AIDS ranked among the top three causes of death for African American men aged 25–54 and for African American women aged 35–44 years (Centers for Disease Control and Prevention, 2003a).

It should go without saying that race and ethnicity are not per se risk factors for HIV. Poverty and other forms of socioeconomic deprivation, however, are known to increase vulnerability to HIV infection. It is estimated that one in four African Americans lives in poverty, and some studies in the US have discerned a close relationship between higher AIDS incidence and lower income (Centers for Disease Control and Prevention, 2003b; Census Bureau, 2000; Diaz, et al., 1994). It is possible that high incarceration rates, particularly for African American men, could be an amplifying factor in the epidemic through injecting drug use and unprotected sex in prison institutions.

For men overall, and African American men specifically, the vast majority of HIV infections occur during injecting drug use and sex between men. High levels of risk behavior are still being found especially among younger men who have sex with men. However, heterosexual intercourse accounts for most HIV diagnoses among women, and there are strong indications that the main risk factor for many women acquiring HIV is the often-undisclosed risk behavior of their male partners. Recent research in a low-income area of New York City, for example, has shown that women were more than twice as likely to be infected by a husband or steady boyfriend than by casual sex partners. Along with injecting drug use, unsafe sex with other men on the part of male partners appears to be a significant risk factor for some women (McMahon et al., 2004). A seven-city study among men who have sex with men has found that 9% of them also had sex with women, and a more recent study among young African American men who have sex with men found that 20% of the men reported also having female sex partners (Valleroy et al., 2004; Centers for Disease Control and Prevention, 2004b).

The majority of people living with HIV in the United States of America are men who have sex with men. Evidence in recent years of increasing cases of syphilis and other sexually transmitted infections among men who have sex with men triggered concerns that risk behavior was on the rise and that a new wave of HIV infections could follow. Studies in Los Angeles and San Francisco found dramatic increases in the number of syphilis cases (from 4 to 260 in Los Angeles county in 1998–2000, and from 67 to 299 in San Francisco in the same period). However, new research suggests that those increases have not had a substantial impact on HIV incidence, which stayed steady in 1999–2002 among men who have sex with men (Centers for Disease Control and Prevention, 2004a). In addition, a new, 16-state study indicates that a majority of HIV-positive men who have sex with men have been taking precautionary measures (including condom use, abstinence or staying faithful to one partner) to prevent further HIV transmission. The study urges, though, that more intensive prevention efforts should reach the small number of HIV-positive men who still practice unsafe sex with other men (Centers for Disease Control and Prevention, 2004c).

After the introduction of antiretroviral therapy in 1995-1996, AIDS-related deaths fell steeply in the United States of America until the late 1990s and then continued to decline more gradually—from 19 005 reported AIDS deaths in 1998 to 16 371 deaths in 2002 (UNAIDS, 2004). However, the rate of death due to AIDS among African Americans was over twice as high as that among whites in 2002. African Americans now have the poorest survival rates among people diagnosed with AIDS–probably reflecting late diagnoses (often after the disease has become symptomatic) and inadequate access to quality health care services.

In Canada, most recent estimates indicate that some 56 000 people were living with HIV at the end of 2002 (Geduld et al., 2004), as many as one third of them unaware of their HIV status. Indigenous persons appear twice as likely to be HIV-infected compared with non-Indigenous persons. Most new HIV infections in Canada in 2002 were attributable to unsafe sex between men (40%) and injecting drug use (30% and declining slightly in recent years). There has been an increase in infections linked to unsafe heterosexual intercourse, a small percentage (less than 10%) of them among refugees or immigrants from high-prevalence countries in sub-Saharan Africa and the Caribbean (Health Canada, 2003).

In the 12 Western European countries with data for newly diagnosed HIV infections, HIV diagnoses in people who were infected through heterosexual contact increased by 122% between 1997 and 2002. In contrast to Canada, a large share of those diagnoses are in people originating from countries with serious epidemics, principally countries in sub-Saharan Africa (Hamers and Downs, 2004) and, for the United Kingdom, the English-speaking Caribbean. There are signs, too, that the resurgence of risky sex between men noted in previous years is leading to increased HIV transmission in some countries among men who have sex with men.

HIV diagnoses among men who have sex with men increased by 22% in 2001-2002 in Western Europe, reversing the slow decline seen in the previous years (Hamers and Downs, 2004). Data on new diagnoses should not be mistaken for HIV incidence, however, since data may reflect an increased uptake of testing services (and therefore include people who became infected several years earlier). The recent rise in new HIV diagnoses in the United Kingdom appears partly attributable to increases in HIV testing (half of the men who had sex with men who were diagnosed in 2002 had been infected for more than six years). Nevertheless, HIV infection is now the fastest-growing serious health condition in England (Department of Health United Kingdom, 2003).

In Germany, where greater treatment access had already boosted uptake of testing in the late 1990s, a recent increase in diagnoses (from 642 HIV diagnoses in 2000 to 742 in 2002) likely reflects an actual increase in new infections. Given persistently high HIV prevalence among men who have sex with men in western Europe—between 10% and 20% in several countries, and even higher in large cities—there is an urgent need to revitalize and improve safer sex campaigns for men who have sex with men (Hamers and Downs, 2004).

Although injecting drug use accounts for a diminishing share of newly diagnosed HIV infections in most Western European countries, it remains an important factor in several countries’ epidemics, among them Italy, Portugal and Spain, and in some cities in other countries. In most cases, this reflects declines in unsafe injecting practices likely associated with effective prevention efforts among injecting drug users in many western European countries. Spain offers a striking example of how a comprehensive set of harm-reduction efforts (including methadone maintenance programmes and needle-exchange projects) can reverse an epidemic among injecting drug users. New HIV infections among drug injectors reached as high as 16 000 in 1985-1986, but plummeted subsequently (De la Fuente et al., 2003).

in Spain, as in several other European countries, HIV prevalence among drug injectors can vary considerably in different parts of the country (European Monitoring Centre for Drugs and Drug Addiction, 2003). A survey of drug injectors at Spanish drug-treatment centers in Catalonia still found HIV prevalence of 38% in 2001, for example (Centre d’Estudis Epidemiológicos sobre l’HIV/Sida de Catalunya, 2001). Indeed, injecting drug use in Spain now appears concentrated largely in the north-east of the country and the Balearic Islands (De la Fuente, 2003).

In Portugal (where the rates of new HIV diagnoses are higher than elsewhere in Europe), injecting drug use still accounted for almost 50% of HIV diagnoses in 2002. HIV prevalence of 20% and higher is still being found among injecting drug users in parts of other countries (including France, Italy and Netherlands) (Hamers and Downs, 2004). A survey among drug injectors at treatment centers in Marseille, France, for example, found 22% were HIV-infected (Emmanuelli et al., 2004). Encouragingly, none of the injectors younger than 30 years was found to be HIV-positive. Nevertheless, ongoing efforts are needed to further limit HIV transmission among injecting drug users—and from them to their sexual partners. Indeed, in Portugal heterosexual infections accounted for more than 40% of newly diagnosed HIV infections in 2002. Based on available data, similar trends have been detected in some regions or provinces of Italy and Spain (Hamers and Downs, 2004).

There is an increasing trend in the share of HIV infections attributable to heterosexual intercourse in Western Europe—and with it, a rise in the number of women who are being diagnosed with the virus. In the 12 Western European countries for which data are available, the proportion of women among people newly diagnosed with HIV infection increased from 25% (1955 out of 7770) in 1997 to 38% (4269 out of 11 337) in 2002 (Hamers and Downs, 2004). In France, some two thirds of new HIV diagnoses in 2003 were in people infected during heterosexual intercourse, while in the United Kingdom the proportion was about 49% and in Germany 41%. A significant number of female sex workers are still being infected in some countries, such as the Netherlands, where a 2002-2003 study found 7% of sex workers (and up to 12% of those working in the street) in Rotterdam were HIV-positive (Van Veen et al., 2004). Approximately 5% of immigrant sex workers (male and female) in Madrid, Spain, were found to be HIV-positive during 1998–2003, according to a recent study. Most of the sex workers were from sub-Saharan Africa (Gutierrez et al., 2004).

Migrants from countries experiencing serious AIDS epidemics, notably in sub-Saharan Africa, bear a disproportionate and growing share of HIV infections throughout Western Europe. In Germany and the United Kingdom, for example, a large share of newly diagnosed, heterosexually-acquired infections in recent years have been in people originating from high prevalence countries (Hamers and Downs, 2004). In the United Kingdom, as many as three quarters of heterosexual infections probably occurred in sub-Saharan Africa, while in Sweden, more than 80% of such infections were probably acquired abroad. The majority of migrants living with HIV appear to be unaware of their serostatus; by and large, HIV diagnoses tend to occur when individuals become symptomatic or get pregnant. In most countries, migrants are not being reached with adequate, appropriate and socially relevant prevention, treatment, and care services. Remedying this situation will require concerted efforts, including forthright steps to counter the social and legal discrimination, and the administrative hurdles migrants encounter.

In the countries of central Europe (including Czech Republic and Hungary), numbers of new HIV infections have stayed stable since the late 1990s, with most of the new infections being recorded in Poland. In the Czech Republic, Hungary, Slovenia, and the Slovak Republic, sex between men is known to be the predominant mode of HIV transmission.

Unlike elsewhere in the world, a large majority of people in most countries in this region who need antiretroviral treatment do have access to it. As a result, AIDS deaths have stayed low since plummeting in the mid-to-late 1990s. In Western Europe, the number of reported deaths among AIDS patients was 3101 in 2002 (UNAIDS, 2004). Two trends, though, warrant concern. In some countries, a large share of HIV infections remain undiagnosed; in the United Kingdom, for example, an estimated one third of people with HIV do not know their serostatus and are likely to discover it only once afflicted by AIDS-related illnesses (Department of Health, 2003). And there is worrying evidence of antiretroviral drug resistance among some newly HIV-infected individuals in Western Europe (Girardi, 2003).

Overall in these countries, thousands of new infections are occurring every year and large numbers of HIV-infected persons are unaware of their HIV status. The main challenges are to provide early and effective treatment and care to all HIV-infected people, to rejuvenate prevention efforts and adapt them to the changing patterns of the epidemic, and to reduce the psychosocial, economic and physical repercussions of HIV infection.

Caribbean

The Caribbean is the second-most affected region
in the world. Among adults aged 15–44, AIDS has become
the leading cause of death.

More than 440 000 [270 000–780 000] people are living with HIV in the Caribbean, including the 53 000 [27 000–140 000] people who acquired the virus in 2004. An estimated 36 000 [24 000–61 000] people died of AIDS in the same year. Among young people 15–24 years of age, an estimated 3.1% [1.6–8.3%] of women and 1.7% [0.9–4.6%] of men were living with HIV at the end of 2004. In the Caribbean Community (CARICOM) region 370 000 [210 000–710 000] people are living with HIV, including the 48 000 [22 000–140 000] people who acquired the virus in 2004. More than 29 000 [17 000–54 000] people died of AIDS in the past year*. With average adult HIV prevalence of 2.3%, the Caribbean is the second-most affected region in the world. In five countries (the Bahamas, Belize, Guyana, Haiti and Trinidad and Tobago), national prevalence exceeds 2%. Overall, the highest HIV-infection levels among women in the Americas are in Caribbean countries and AIDS has become the leading cause of death in the Caribbean among adults aged 15–44 years (Caribbean Epidemiology Centre, PAHO, WHO, 2004). Life HIV and AIDS statistics and features, end of 2002 and 2004 Adults and children living with HIV Number of women living with HIVAdults and children newly infected with HIVAdult prevalence (%) Adult and child deaths due to AIDS2004440 000 [270 000–780 000]210 000[120 000–380 000]53 000 [27 000–140 000]2.3 [1.5–4.1]36 000 [24 000–61 000]2002420 000 [260 000–740 000]190 000[110 000–360 000]52 000 [26 000–140 000]2.3 [1.4–4.0]33 000 [22 000–57 000]expectancy at birth in 2010 is projected to be 10 years less in Haiti and in Trinidad and Tobago nine years less than it would have been without AIDS (see Figure 13) (Stanecki, 2004). Several countries and territories with economies that are dependent on tourism rank among those most heavily affected by the epidemic in this region, including the Bahamas, Barbados, Bermuda, Dominican Republic, Jamaica, and Trinidad and Tobago. Yet most countries in the region have limited capacity to track the evolution of their epidemics, and are relying on data and systems that do not necessarily match the realities they are facing. Unlike in Latin America, HIV transmission in the Caribbean is occurring largely through heterosexual intercourse (almost two thirds of all AIDS cases to date are attributed to this mode of transmission), although sex between men, which is heavily stigmatized, and in some places illegal, remains a significant—but still neglected—aspect of the epidemics. HIV transmission through injecting drug use remains rare, with the significant exception of Bermuda, where it accounts for a large share (43%) of AIDS cases, and Puerto Rico, where more than half of all infections in 2002 were associated with injecting drug use and about one quarter were heterosexually transmitted (Caribbean Epidemiology Centre, 2003; UNAIDS, 2004). As the epidemics in this region evolve, more women are being affected, and the number of new HIV infections among them now outstrips that among men. Latest estimates suggest that roughly as many women as men are now living with HIV in this region. According to a population-based survey carried out in 2002, women younger than 24 years in the Dominican Republic were almost twice as likely to be HIV-infected compared with their male peers (MAP, 2003). In Jamaica teenage girls are 2.5 times more likely than boys in the same age group (10–19 years) to be infected—due partly to the fact that some girls have sexual relationships with older men who are more likely to be HIV-infected, a trend that has also been documented in several other countries.

Haiti continues to have the largest number of people living with HIV in the Caribbean: some 280 000 [120 000–600 000] at the end of 2003 (UNAIDS, 2004). The most recent sentinel surveillance studies suggest a decline in HIV prevalence. The latest data suggest that median HIV prevalence among women (15–49 years) attending antenatal clinics has fallen from 4.5% in 1996 to 2.8% in 2003-2004 in consistently reporting sites (Ministère de la Santé publique et de la population Haiti et al., 2004). HIV prevalence among pregnant women aged 15–24 years appears to have declined by a similar magnitude as among women of all ages in the same period—from 3.6% to 2.8%. The decline in the youngest age group is usually considered to indicate a decline in incidence. The decline in the older age group of similar magnitude is not easily explained. Increased donor support of HIV surveillance in Haiti in 2003-2004 may have enhanced the quality of surveillance data, making it difficult to compare the latest information with that obtained in earlier surveillance rounds. Further investigation combining trends in prevalence data with trends in behavioral data is needed to examine the reasons for the observed decline in Haiti’s HIV prevalence .The latest round of HIV surveillance among pregnant women shows HIV prevalence varying between 1.8% and almost 7% in different parts of Haiti. Poorer, less educated women are more likely to be HIV-infected than their better-off counterparts. Recent behavioral surveillance has shown that a significant proportion of the country’s largely young population (about 60% of which is under 24 years) is sexually active and having unprotected sex. In a 2001 survey, almost half the young women (and more than half the men) said they had become sexually active before their 18th birthday and 18% of urban women in their late teens (15–19 years) had been pregnant at least once. Although general AIDS knowledge is widespread, misconceptions about the virus continue to circulate, particularly among women—an indication that there is still considerable room to expand and improve prevention efforts.

Jamaica, where an estimated 22 000 [11 000–41 000] people were living with HIV at the end of 2003, has the second-highest annual number of AIDS cases and deaths in the region, after Haiti (UNAIDS, 2004). While a handful of other islands in the region appear to be making incremental inroads against the epidemic by expanding access to antiretroviral treatment (see below), in Jamaica at least 900 AIDS cases have been reported there each year since 1999 and the disease has been claiming between 590 and 690 lives annually over the same period—more than double the numbers just four years earlier (Caribbean Epidemiology Centre, 2004). The most recent round of HIV surveillance indicates that the HIV epidemic is not abating either. HIV prevalence among pregnant women attending antenatal clinics stood at 1.4% nationally in 2002, but was as high as 2.7% and 1.9% in the parishes of St. James and of Kingston and St. Andrews, respectively. High HIV prevalence found among patients at sexually transmitted infection clinics in 2002 seemed to confirm earlier warnings of widespread unprotected sex; almost 8% of men attending these clinics were HIV-positive, as were almost 5% of women (Ministry of Health Jamaica, 2003). In the late 1990s, a study among male adolescents and young adults in Jamaica found that fully 9% had reported symptoms of sexually transmitted infections in the previous year, a clear indication of unprotected sex (Norman and Uche, 2002). There is no evidence to date that these patterns have altered significantly.

in both the Bahamas and Barbados, there are indications that stronger prevention efforts since the late 1990s could be forcing HIV-infection levels lower. In the Bahamas (see Figure 14), where an estimated 5000 people were living with HIV at the end of 2003, HIV prevalence among pregnant women fell from 4.8% in 1993 and 3.6% in 1996 to 3% in the latest round of HIV surveillance in 2002. A similar downward trend in HIV levels has been observed among patients at sexually transmitted infection clinics (Department of Public Health Bahamas, 2004). The decline in the annual number of reported AIDS cases (from 320 in 2000 to 164 in 2003) and AIDS deaths (from 272 in 2000 to 185 in 2003) probably reflects this steady drop in prevalence, along with the expansion of antiretroviral treatment access since the turn of the century (Caribbean Epidemiology Centre, PAHO, WHO, 2003).

Barbados has a smaller epidemic but there, too, a decline in HIV levels is being observed, with new HIV diagnoses among pregnant women dropping substantially between 1999 and 2003, from 0.7% to 0.3% (Kumar and Singh, 2004). Mother-to-child transmission of HIV has also been reduced since the expansion of voluntary counseling and testing services, and the provision of antiretroviral prevention regimens. The rate of mother-to-child transmission declined by 69% between September 2000 and December 2002 (St John et al., 2003). In addition, the introduction in 2001 of antiretroviral treatment for people living with HIV has reversed the trend of AIDS mortality in the island nation. The annual number of AIDS deaths decreased from 114 in 1998 to 50 in 2003, while hospital admissions for treatment of opportunistic infections fell by 42% in the same period. In Bermuda, meanwhile, the number of AIDS cases decreased by almost half (19 to 11) between 2000 and 2002 (Caribbean Epidemiology Centre, 2004; Caribbean Epidemiology Centre, PAHO, WHO, 2003). At an estimated 2.5% at the end of 2003, HIV prevalence in Guyana was the second-highest in the region. There has been a steep rise in the numbers of HIV cases reported since the mid-1990s. According to the Ministry of Health, officially reported cases probably represent less than one third of the actual number of people living with HIV. Meanwhile, fewer than one fifth of people infected with HIV—the majority of them aged between 20 and 34 years— are aware of their serostatus. Most infections are occurring through heterosexual intercourse. One recent study among miners in the country’s Amazon region has revealed an exceptionally high HIV prevalence of 6.5%. The miners, all young men, divide their lives between six to eight week work shifts and two weeks of rest at their homes near the coast. The danger of infected miners transmitting HIV to their regular partners at home or to casual partners near the mine seems substantial (Palmer et al., 2002).

The Dominican Republic, which shares Hispaniola Island with Haiti, still faces a serious epidemic with HIV prevalence among pregnant women higher than 2% nationally. In the capital Santo Domingo, HIV prevalence among 15–24 year-old pregnant women—which can offer a hint of recent infection rates—has declined from around 3% in 1995 to below 1% in 2003 (UNAIDS, 2004). This might be due to prevention efforts aimed at encouraging people to adopt safer sexual behaviors. However, the same trend is not apparent elsewhere in the country. Of particular concern is the unusually high HIV prevalence (4.9%) that has been detected among sugar cane plantation workers.

Cuba has been an exception in this region, with very low HIV prevalence, possibly due, in part, to a policy of quarantining HIV-infected people as a preventive measure during the 1980s. Cuba subsequently abandoned that policy. Meanwhile, universal free access to antiretroviral therapy has kept the number of AIDS cases and deaths very low. Though still small in scale, the country’s HIV epidemic is now growing, however. A sharp increase in newly reported HIV cases has occurred since the late 1990s, with the annual number of reported new cases growing almost five-fold between 1995 and 2000. However, Cuba’s epidemic remains small, in contrast to much of the Caribbean. Most new HIV transmission is occurring during sex between men (Caribbean Epidemiology Centre, 2003). The lack of good quality HIV-surveillance data in most Caribbean countries is hampering the ability to design and run potentially effective prevention programs, and will almost certainly undermine efforts to expand access to antiretroviral treatment. But social, not just technical, challenges will need to be confronted if the countries of this region are to bring their epidemics under control. Widespread homophobia is providing an ideal climate for the spread of HIV by driving men who have sex with men further away from the information, services, and security they need if they are to protect themselves against HIV. Meanwhile, the unequal social and economic status of women and men is acting as a powerful dynamic in epidemics that are growing amid ongoing stigma, misconceptions and denial.

Asia

National HIV infection levels in Asia are low compared with some other continents, notably Africa. But the populations of many Asian nations are so large that even low national HIV prevalence means large numbers of people are living with HIV. Latest estimates show some 8.2 million [5.4 million–11.8 million] people (2.3 million [1.5 million–3.3 million] adult women) were living with HIV at the end of 2004, including the 1.2 million [720 000–2.4 million] people who became newly infected in the past year. AIDS claimed some 540 000 [350 000–810 000] lives in 2004. Among young people 15–24 years of age, 0.3% of women [0.2–0.6%] and 0.4% of men [0.3–0.8%] were living with HIV by the end of 2004.

Asia is not just vast but diverse, and HIV epidemics in the region share that diversity, with the nature, pace and severity of epidemics differing across the region. Overall, Asian countries can be divided into several categories, according to the epidemics they are experiencing. While some countries were hit early (for example, Cambodia, Myanmar and Thailand), others are only now starting to experience rapidly expanding epidemics and need to mount swift, effective responses. They include Indonesia, Nepal, Viet Nam, and several HIV and AIDS statistics and features, end of 2002 and 2004 Adults and children living with HIV Number of women living with HIVAdults and children newly infected with HIVAdult prevalence (%) Adult and child deaths due to AIDS20048.2 million[5.4–11.8 million]2.3 million[1.5–3.3 million]1.2 million[720 000–2.4 million]0.4 [0.3–0.6]540 000[350 000–810 000]20027.2 million[4.6–10.5 million]1.9 million[1.2–2.8 million]1.1 million[540 000–2.5 million]0.4 [0.2–0.5]470 000[300 000–690 000]provinces in China. In Myanmar and in parts of India and China, HIV has become well-entrenched in some sections of society, despite modest efforts to halt the virus’ spread. Other countries are still seeing extremely low levels of HIV prevalence, even among people at high risk of exposure to HIV, and have golden opportunities to pre-empt serious outbreaks. These countries include Bangladesh, East Timor, Laos, Pakistan, and the Philippines. Some countries, by sheer virtue of size, simultaneously fit several of those descriptions: China and India are examples. These two countries, home to some 2.35 billion people, are experiencing several distinct epidemics, some already very serious.

China

Although moving at a varied pace, HIV has spread to all of China’s 31 provinces, autonomous regions and municipalities. In some, such as Henan, Anhui, and Shandong, HIV was already spreading a decade ago among rural people who sold blood plasma to supplement their incomes. Elsewhere, the virus has established a more recent but firm presence among injecting drug users and, to a lesser extent, sex workers and their clients (Zang, Ma and Xia, 2004). Much of the current spread of HIV in China is also attributable to injecting drug use and paid sex. HIV prevalence among drug injectors was measured at between 18% and 56% in six cities in the southern provinces of Guangdong and Guangxi in 2002, while in Yunnan province some 21% of injectors tested positive for HIV in 2003 (China National Center for AIDS/STD Control and Prevention, 2003). Sexual transmission of HIV from injecting drug users to their sex partners looks certain to feature more prominently in China’s fast-evolving epidemic. Some 47% of surveyed female drug injectors in Sichuan province and 21% in neighbouring Yunnan province reported selling sex for money or drugs in the previous month, according to recent studies. Condom use was reportedly quite high but it was hardly the norm. Once HIV becomes well-established in commercial sex circuits, onward spread of the virus could be quite rapid if current behavior trends persist. In 2003, almost one quarter of surveyed sex workers in Guangxi never used condoms and about one half used them only occasionally (China National Center for AIDS/STD Control and Prevention, 2003). In Sichuan, only around 40% of sex workers reported using condoms with all their clients in the previous month, according to a 2002 study. Little is known about the possible role of sex between men in China’s epidemic. A rare survey of men who have sex with men in Beijing, conducted in 2001-2002, found that approximately 3% of the men were HIV-infected (almost all of whom had been unaware of their serostatus) (Choi et al., 2003).

There are signs that efforts to boost public knowledge about HIV are bearing fruit but there remains much room for improvement. A 2003 survey found that two-in-five Chinese men and women could not name a single way to protect themselves against infection (Shengli, Shikun and Westley, 2004). In Sichuan province, more than one third of sex workers (and a similar proportion of clients) did not know that condoms offer good protection against HIV. Research suggests that few brothels in China have condom policies of the sort that helped Cambodia and Thailand bring their epidemics under control, and even fewer are keeping condoms on the premises. China can still shape the course of its epidemic. But it needs to move swiftly and with great resolve.

India

India’s epidemics are even more diverse than China’s. Latest estimates show that about 5.1 million
[2.5–8.5 million] people were living with HIV in India in 2003. Serious epidemics are underway in several states. In Tamil Nadu, HIV prevalence of 50% has been found among sex workers, while in each of Andhra Pradesh, Karnataka, Maharashtra and Nagaland, HIV prevalence has crossed the 1% mark among pregnant women. In Manipur, meanwhile, an epidemic driven by injecting drug use has been in full swing for more than a decade and has acquired a firm presence in the wider population (UNAIDS/WHO, 2003). HIV prevalence measured at antenatal clinics in the Manipur cities of Imphal and Churachand has risen from below 1% to over 5%, with many of the women testing positive appearing to be the sex partners of male drug injectors. Several factors look set to sustain Manipur’s epidemic, including the large proportion (about 20%) of female sex workers who inject drugs and the young ages of many injectors (40% of male injectors surveyed in 2002 were under 25 years of age).

There are signs that injecting drug use is playing a bigger role in India’s epidemics than previously thought. Most surveillance sites for injecting drug users are in the northern states where injecting is common behavior, but other parts of the country have yielded equally troubling evidence. In the southern city of Chennai, for example, 26% of drug injectors were already infected with HIV when a sentinel site was established there in 2000; by 2003, 64% were infected. In most cities where injecting drug users have been surveyed, at least one quarter of them—and, in Chennai, 46%—said they lived with a wife or regular sex partner (MAP, 2004). This has probably contributed to the fact that Chennai also has among the highest HIV-prevalence rates among pregnant women in the country. It is likely that many of those women were infected by partners who injected drugs.Like Manipur, the states of Maharashtra, Tamil Nadu and Andhra Pradesh have long-established HIV epidemics, but theirs are driven mainly by commercial sex. Available evidence suggests that prevention efforts in some of those states have done little to alter the epidemics’ advance.

Sentinel surveillance has revealed no significant drop in HIV prevalence among female sex workers in Mumbai, for example, despite decade-old safer-sex programs for sex workers. It appears the programs have been either too scattered or short-term to reach a large enough proportion of sex workers to make a difference. In some of these states, HIV has been rising steadily among pregnant women, most likely because clients have transmitted the virus to their regular partners. Fortunately, India does boast some significant prevention successes, such as the drop in unprotected casual sex reported in the southern state of Tamil Nadu. In 1996, 14% of truck drivers reported recent unprotected sex with a sex worker. By 2002, after concerted prevention programs were introduced, that had fallen to just 2% (AIDS Prevention and Control Project, 2003).

As elsewhere in the region, the role of sex between men in India’s epidemics remains poorly understood. What is clear is that a considerable number of men in India do have sex with other men. In a household-based survey in a low-income area of Chennai, India, 6% of men reported sex with other men. These men were over eight times more likely to be infected with HIV than other men in the survey, and 60% more likely to be infected with other sexually transmitted infections. A high proportion of men who have sex with men also reported sex with women (Go et al., 2004). For example, in a household study in India, 57% of men reporting sex with other males were married (NACO, 2002).

High risk behavior and sharp rises in HIV

HIV prevalence is rising sharply in several places where it stayed low for many years. These rises are most dramatic among people whose behaviorscarry a high risk of exposure to HIV—drug injectors, sex workers and their clients, and men who have sex with men. In Indonesia, Nepal, Viet Nam and parts of China, rapid, recent rises in HIV infection among drug injectors appear to have spurred subsequent rises in HIV infection among non-injectors who have sexual risk behaviors, “kick-starting” wider epidemics,Given the very large population numbers in these countries, continued HIV spread among those with risk behaviors and their sex partners. research has revealed ample opportunities for wider HIV transmission, the epidemic will assume diverse patterns. Risk behavior among injecting drug users in Indonesia is very common. A recent survey in three cities found 88% of the injectors had used non-sterile needles or syringes in the preceding week, yet fewer than one third said they felt at high risk of HIV infection (Pisani et al., 2003). When injecting drug users are tested for HIV, very high infection levels are found. One in two injecting drug users in Indonesia’s capital, Jakarta, now test positive for HIV, while in far-flung cities such as Pontianak (in West Kalimantan province on the island of Borneo) more than 70% of drug injectors who request HIV tests are discovering that they are HIV-positive.

Conditions also favor HIV spread through sex work. In seven Indonesian cities, an average 42% of sex workers had either or both gonorrhea or Chlamydia in 2003. Condom use ranges from irregular to rare. In 2002, fewer than one in five sex workers operating out of massage parlors and discotheques in Jakarta said they used condoms consistently (MAP, 2004). Among sex workers in brothel areas (a group that ought to be easier to reach with interventions), rates of condom use with all clients in the previous week stood at a meager 4% . The situation is even more troubling in parts of Indonesia’s easternmost province of Papua, where HIV prevalence among sex workers in Sorong, for example, had reached 17% by 2003, over five times the national average for sex workers.

There are strong signs that the virus is spreading beyond sex workers and their clients. One recent serosurvey among adults in five villages found that close to 1% were HIV-positive . Household surveys of young men and women in Jayapura and Merauke show that both young men and women in Papua report far less drug use and far more sexual activity than those in other parts of Indonesia. The data are inconclusive, but they suggest patterns of sexual networking that could favor dynamic HIV spread in the general population on Papua (Indonesia Central Bureau of Statistics and MACRO International, 2004). Unsafe injecting drug use is the wellspring of Nepal’s epidemic, too. Use of non-sterile injecting equipment is widespread and accounts for the high HIV prevalence—22% to 68% across the country in 2002—among male injectors, many of them younger than 25. Younger injectors appear more likely to report risky practices in parts of Nepal; in the east, for example, injectors under 25 were three times as likely to report using non-sterile equipment at last injection compared with older injectors. Nepal’s epidemic also highlights the potential links between HIV infection and mobility. Injecting drug users from cities with low prevalence, but who had injected drugs elsewhere, have been found to be two to four times more likely to have acquired HIV than those who had remained in their home cities. Half of the sex workers surveyed in central Nepal and who said they had worked in Mumbai (India) were HIV-infected, compared with 1.2% of those who had never been to India.

Widespread injecting drug use by sex workers makes Viet Nam’s epidemic particularly explosive. In Ho Chi Minh City, 38% of almost 1000 sex workers included in one survey injected drugs—and fully 49% of those injecting sex workers were infected with HIV (compared with 8% of those who didn’t use any drugs). In the northern port city of Haiphong, nearly 40% of all sex workers said they injected drugs, compared to one in six sex workers who did likewise in the capital, Hanoi. Drug-using sex workers are about half as likely to use condoms as those who do not use drugs, according to another study in Ho Chi Minh City. These trends probably explain a good deal of the steep rises in HIV prevalence detected in some of Viet Nam’s cities, where the virus now appears to be spreading freely among groups that are at high risk of exposure to HIV. HIV prevalence of 8% was detected in a 2003 Ho Chi Minh city survey among men who have sex with men.

Some countries, including East Timor and Pakistan, could be poised for HIV outbreaks. Until very recently the majority of HIV infections and AIDS cases reported in Pakistan were among migrant Pakistani workers who had been deported from the Gulf States. However, there has been a recent report of an HIV outbreak among injecting drug users in a small town in Pakistan’s Sindh province. Just under 10% of the drug injectors in the town of Larkana reportedly tested HIV-positive (Shah et al., 2004.) Studies among Pakistani truck drivers have found that one in three has never heard of condoms, and 19 out of 20 who bought sex from women did not use condoms. Meanwhile, nearly six out of 10 sex workers in East Timor have never heard of AIDS, four out of 10 do not recognize a condom when shown one, and zero out of 10 consistently use condoms with their clients (Pisani and Dili STI survey team, 2004).

EASTERN EUROPE AND CENTRAL ASIA

Most of the epidemics in this region are still
in their early stages—which means that timely, effective interventions
can halt and reverse them.

In Eastern Europe and Central Asia the number of people living with HIV has risen dramatically in just a few years—reaching an estimated 1.4 million [920 000–2.1 million] at the end of 2004. This is an increase of more than nine-fold in less than ten years. Some 210 000 [110 000–480 000] people were newly infected with HIV in the past year, while an estimated 60 000 [39 000–87 000] died of AIDS. Among young people 15–24 years of age, an estimated 0.8% [0.4–1.6%] of women and 1.7% [0.8–3.7%] of men were living with HIV at the end of 2004.

Diverse HIV epidemics are underway in this region. The most serious and firmly-established epidemic is in Ukraine, which is experiencing a new surge of reported infections, while the Russian Federation is home to the largest epidemic in the entire region (indeed in all of Europe). However, HIV is unevenly distributed in Russia, with about 60% of all HIV infections to date having been reported in just 10 of the country’s 89 regions. There is considerable scope for further expansion of the epidemic in this vast country—alongside great opportunities to prevent such an outcome. Several Central Asian and Caucasian republics have entered the early HIV and AIDS statistics and features, end of 2002 and 2004 Adults and children living with HIV Number of women living with HIVAdults and children newly infected with HIVAdult prevalence (%) Adult and child deaths due to AIDS20041.4 million [920 000–2.1 million]490 000[310 000–710 000]210 000 [110 000–480 000]0.8 [0.5–1.2]60 000 [39 000–87 000]20021.0 million [670 000–1.5 million]330 000[220 000–480 000]190 000 [94 000–440 000]0.6 [0.4–0.8]40 000 [27 000–58 000]stages of the epidemic, while in south-eastern Europe, HIV has acquired a tenuous presence amid behavior patterns that favor significant spread of the virus.

Amid such diversity, four features stand out. On the whole, most of the epidemics in this region are still in their early stages—which means that timely, effective interventions can halt and reverse them. Secondly, the vast majority of people living with HIV in this region are young; more than 80% of the reported infections are being found among people below the age of 30 years (by comparison, in Western Europe some 30% of people with HIV fall in that age group). Thirdly, sexual transmission of HIV is increasing in each of the most seriously-affected countries—an indication that the epidemic has gained a foothold in the wider population. Fourthly, ongoing, arduous social and economic transitions serve as the context in which extraordinarily large numbers of young people are injecting drugs. In countries with emerging epidemics demand- reduction programs that discourage drug use and harm reduction programs that reduce drug injecting and prevent HIV transmission through contaminated injecting equipment among young people can prevent larger, more extensive HIV epidemics of the kind now taking hold in Russia and Ukraine. This would entail a comprehensive set of interventions to lessen the vulnerability of young people and reduce the numbers of people initiated into drug injecting, alongside large-scale harm reduction and safer-sex programs.

The Russian Federation has the largest number of people living with HIV in the region, and accounts for some 70% of all HIV infections officially registered in Eastern Europe and Central Asia (Rhodes et al., 2002). An estimated 860 000 [420 000–1 400 000] people were living with HIV in Russia at the end of 2003, fully 80% of them aged 15–29 years and more than one third of them women (UNAIDS, 2004; Field, 2004). HIV prevalence is increasing steadily. Infection levels measured among pregnant women have risen from less than 0.01% in 1998 to 0.11% in 2003. In St. Petersburg, prevalence increased from 0.013% in 1998 to 1.3% in 2002—a hundred-fold increase.

It bears noting that the number of newly reported HIV infections in Russia has declined in the past few years. The 39 699 officially-reported new infections in 2003 were 24% fewer than the 52 349 reported in 2002, and 55% fewer than the 88 577 cases documented in 2001. It is unclear whether this possibly represents a slowing in the growth of Russia’s epidemic and, if so, what might be causing it. One possibility is that, in some of the areas currently most-affected, a majority of injecting drug users have already been tested. It is also possible that HIV prevalence has reached saturation levels among some of the currently affected drug injector populations in those areas 18(EuroHIV, 2003). In some areas fewer people were tested for HIV (the total number of drug users tested for HIV shrank from 491 526 in 2001 to 279 509 in 2003). As well, the slower transmission route of sexual intercourse may be gaining in prominence. Although HIV infections have been recorded throughout the expanse of the Russian Federation, much of the epidemic is still concentrated in 10 regions (nine of them in the more densely populated west, see Figure 18) (AIDS Foundation East-West, 2004). In the absence of effective prevention efforts, serious HIV outbreaks could follow in the rest of the country. At the heart of the country’s epidemic are the extraordinarily large numbers of young people
who inject drugs; and have active sex lives. Between 1.5 and 3 million Russians are believed to inject drugs (1% to 2% of the entire population), and an estimated 30% to 40% of injecting drug users use non-sterile needles or syringes, which massively boosts the chances of HIV transmission (Max Planck Institute for Foreign and International Law, 2000). HIV prevalence among injecting drug users is high in many parts of Russia. A recent multi-center study estimated that 65% of street injecting drug users in Irkutsk were HIV-positive (90% of them still in their teens); in Tver, 55% were infected, in Ekaterinburg the figure was 34% and in Samara 29% (Rhodes et al., 2004). Studies show that most drug injectors are young (under 25 years), male, unemployed and living in major cities (though there are signs that the practice is spreading into rural areas).

In early 2004, more than 80% of all officially reported HIV cases since the beginning of the epidemic had been among drug injectors (Russian Federal AIDS Centre, 2004). But the majority of drug injectors are sexually active—upwards of 70%, according to studies in several Russian cities. Many have regular sexual partners, some buy or sell sex (see box). Those infected with HIV are therefore liable to transmit the virus sexually unless they practice safer sex. And studies show that the majority of male drug injectors do not use condoms consistently. In Togliatti and Nizhny Novgorod, for example, 83% of male injecting drug users had not used condoms regularly in the last month, while in Mirny 23% said they never used a condom (Lowndes et al., 2002; Moshkovich et al., 2000; Filatov and Suharsky, 2002; Rhodes et al., 2004). As a result, the epidemic’s pattern is shifting and the proportion of new, reported HIV infections acquired during heterosexual intercourse has grown dramatically—from 5.3% in 2001 to almost 15% in 2002 and just over 20% in 2003. This means that more women are being infected; indeed, the overall proportion of women among people living with the virus has increased to 38% in 2003, compared Sexually-acquired HIV cases form an increasing proportion of new infections, notably in the cities of Kaliningrad, Mirny, Moscow and Vladivostok (Rhodes et al., 2004).

Official estimates put HIV prevalence in the Russian prison system at 2% to 4%—at least four times higher than in the wider population. As injecting drug use has increased, so has the proportion of prisoners who have histories of injecting drug use, as well as the number of them who are HIV-positive. Russia is now developing a program that will include prevention education for prisoners, access to condoms and to bleach (for cleaning injecting equipment) (UNAIDS, 2004).

As in other countries in the region, Russia’s HIV surveillance system captures very little information about HIV transmission between men who have sex with men. It is likely that sex between men is a more prominent factor in Russia’s epidemic than is currently evident, and that this mode of transmission could be linked to onward heterosexual transmission of the virus. Research in St Petersburg has cast some light on the issue, by showing that over one third of surveyed men who had sex with men also had sex with women in the previous three months, and most of these men had multiple male and female partners in that period. Bisexual men were more likely to have sold sexual services, and displayed the least amount of knowledge about HIV.

Men who have sex with men—like injecting drug users and sex workers—endure stigma and discrimination, both at the hands of officialdom and society at large. This should not stand in the way of setting up reliable sentinel surveillance among these men (including those who also have sex with women) and implementing appropriate prevention services that can help limit the epidemic’s spread through and beyond their ranks. These programs will need to focus on the risks of HIV infection associated both with sex between men and heterosexual intercourse (Kelly et al., 2002).

Tackling the epidemic need not be as daunting as it appears. At the moment, the epidemic still has an uneven presence across Russia. Even though HIV infections have been recorded in each of Russia’s 89 administrative territories, reported HIV prevalence is extremely low (1–150 cases per 100 000 population) in 66 of them (with almost 60% of the country’s population) (AIDS Foundation East-West, 2004). While moving toward much wider coverage of HIV programs ultimately, particular effort needs to be focused on the 10 territories responsible for over half of the total reported cases. With around 90 harm-reduction projects operating in the entire Russian Federation, there is a huge room for improvement (Rhodes et al., 2004).

Erupting with astounding speed over the past decade, Ukraine’s firmly-rooted epidemic continues to expand, as figure 21 shows. Newly registered HIV infections have been increasing annually since the turn of the century—by 7% in 2000, 13% in 2001 and 25% in 2002. Just 10 years ago, there were only 183 officially registered HIV cases in Ukraine, but by mid-2004 more than 68 000 cases of HIV infection had been officially registered (EuroHIV, 2003). These figures grossly understate the actual scale of the epidemic since they only measure infections among people who come in direct contact with the authorities and testing facilities.

More widespread sexual transmission of HIV has become an increasingly prominent factor in Ukraine’s epidemic, which has evolved to the point where a growing share of new infections are occurring during unsafe sex between people who have no direct relationships with injecting drug users. Some 30% of new HIV infections registered in 2003 occurred during heterosexual intercourse (almost three times higher than the 11% documented in 1997), and more than 40% of people with HIV infection are women, most of them in their peak reproductive years (EuroHIV, 2003). Indeed, data show that as much as 60% of HIV-infected women are under 25 years of age. In urban areas such as Odessa and Mikolyiv more than 1% of pregnant women attending antenatal clinics are HIV-positive. These trends have spurred considerable efforts to expand programs to prevent mother-to-child transmission,and these are beginning to bear fruit, with the proportion of HIV-infected babies born to infected mothers diminishing from 27% in 2001 to 12% in 2003.

At the same time, injecting drug use remains an important facet of the epidemic, particularly in the eastern and southern parts of the country. As elsewhere in the region, the vast majority (about 80%) of HIV-infected drug injectors are young (under 30 years of age). A large proportion of them—some 28%, according to a recent study in Kiev, Odessa and Donetsk—are female. The same study found widespread re-use of injecting equipment among users, more than half of whom were also sexually active. Condom use was low: just one third of the sexually active injectors had used a condom during sex in the previous month. A potent overlap between injecting drug use and commercial sex is found in places such as Donetsk, where 33% of female sex workers who inject drugs were found to be HIV-positive. Somewhat heartening was the finding that many injecting drug users who knew they were HIV-positive were either abstaining from sex (40% had not had intercourse in the previous month) or were using condoms (Booth et al., 2004).

In a country with a quickly maturing epidemic, it is disconcerting that so little is known about the role of sex between men as a contributory factor. Ukraine’s HIV-registration system requires that persons who test HIV-positive declare, if possible, how they likely acquired the virus. Since the first HIV case was detected in Ukraine in 1987, only 44 cases have been attributed to sex between men—a curiously low figure which lends credence to concerns that HIV might be spreading largely undetected among men who have sex with men (EuroHIV, 2003).

The deadly combination of HIV and tuberculosis (TB) is a serious concern in Ukraine, where 10–15% of TB cases are estimated to be multidrug resistant. Tuberculosis has become the leading cause of death among people living with HIV. This underlines the need for a significant scaling up of access to antiretroviral treatment in Ukraine. Currently, just over 500 of the estimated 45 000 people who need antiretroviral treatment in Ukraine are receiving it, despite the fact that treatment access for all is guaranteed by Ukranian law.

In the Baltic states, HIV transmission is occurring at a brisk rate, even if the overall numbers of infections remain low. At 2300 in 2002, the total number of HIV diagnoses in Latvia has risen five-fold since 1999. Just four years ago, Estonia reported 12 new HIV cases; in 2003, 840 people were newly diagnosed with the virus. In Lithuania, the 72 new HIV cases detected in 2001 increased more than five-fold the following year. Injecting drug use still accounts for the largest proportion of newly reported infections in these countries but sexual transmission is slowly gaining ground. In Belarus (where more than 5000 people had been officially diagnosed with HIV by mid-2003) and Moldova (where the figure stood at just under 1800), most infections are occurring among young drug injectors and their sexual partners (EuroHIV, 2003). In Moldova, however, HIV prevalence of almost 5% has been found in street-based sex workers, one in ten of whom also reported a history of injecting drugs—suggesting a potentially strong link in HIV transmission among drug injectors, sex workers and their clients (WHO Regional Office for Europe, 2004).

Meanwhile HIV prevalence remains very low (less than 0.3%) in most of the Central Asian and Caucasian republics, though the overall number of registered infections continues to rise—formidably in Uzbekistan, which hosts one of the youngest epidemics in the world. Almost 91% of all reported infections were diagnosed between 2001 and mid-2003, bringing to more than 2500 the total number of reported HIV cases, as Figure 22 shows. Uzbekistan’s epidemic is now developing swiftly. Already, commercial sex appears to be playing a large role. The proportion of women among people living with the virus has grown annually from just over 12% in 2001 to almost 18% in 2003. HIV infections have been recorded in all regions of the country, though the epidemic is most heavily concentrated in the capital Tashkent (48% of all registered HIV cases) and surrounding areas (20%). In Kazakhstan, where a total of just over 3600 HIV cases had been reported by mid-2003, sentinel surveillance conducted in 2003 has shown prevalence levels of 3.8% in injecting drug users and 4.6% in sex workers but there are no data concerning men who have sex with men (EuroHIV, 2003). The proportion of people living with HIV in this region who are in Kazakhstan rose from 19% in 2001 to 24% in 2003. These epidemics are growing at a fearsome pace and are concentrated currently among young people who inject drugs and/or engage in commercial sex. Kyrgyzstan’s much smaller epidemic is being propelled mainly by injecting drug use and is concentrated largely still in Osh Oblast, two regions of Chui Oblast (Jaiyl and Yssykata) and Bishkek City. In a country where it is officially estimated that at least 2% of the adult population injects drugs, huge scope exists for rapid and extensive spread of HIV. In the Caucuses, new studies indicate that significant HIV outbreaks are underway in Azerbaijan, where one in four street drug injectors in the capital, Baku, have been found to be HIV-positive. Among street-based sex workers, HIV prevalence of 11% has been detected and among their counterparts working out of cafes and saunas, prevalence of 6% has been found (WHO Regional Office for Europe, 2004). Because the epidemics are still in their early stages, they can be halted with prevention strategies that concentrate on reaching those who are currently most at risk of exposure to HIV.

In parts of south-eastern Europe (notably countries emerging from conflict and difficult transitions) drug injecting and sexual risk behavior appear to be on the increase, and rising numbers of HIV infections could soon follow. For example, the capital of Romania, Bucharest, has seen a swift rise in the number of injecting drug users during the past few years. By 2002 there were an estimated 30 000 injecting drug users in the capital (more than 1% of the city’s population); four years earlier, another exercise had estimated their number at just 1000. More than 80% of the drug injectors are under 30 years of age and over 70% of them are unemployed. An estimated 60% of the drug injectors use non-sterile needles and syringes, and 40% to 60% are estimated to be infected with hepatitis C. As of yet, sentinel surveillance has revealed very few cases of HIV infection among injecting drug users.

Overall in Eastern Europe and Central Asia, current case reports reflect the situation only among those people and groups (chiefly injecting drug users) who come into contact with HIV testing programs. Consequently, little is known about HIV spread among people who do not engage with the authorities and/or testing services. Among the rare studies of men who have sex with men is a recent, small investigation conducted in Ekatarinaburg, which found HIV prevalence of 5% among men who have sex with men, one third of whom also had sexual relations with women and half of whom never used a condom (WHO Regional Office for Europe, 2004). Networks of men who have sex with men have been documented in several countries and scattered surveys of sexual behavior (for example, in the Russian Federation and Ukraine) have pointed to high levels of unprotected sex. Throughout the region, men who have sex with men are routinely stigmatized and discriminated against, and in many countries the likely role of sex between men in the epidemic is not being recognized.

Latin America

Only Guatemala and Honduras have national HIV prevalence
of over 1%, but lower prevalence in other countries disguises serious, localized epidemics.

More than 1.7 million [1.3 million–2.2 million] million people are living with HIV in Latin America. In 2004, around 95 000 [73 000–120 000] people died of AIDS, and 240 000 [170 000–430 000] were newly infected. Among young people 15–24 years of age, an estimated 0.5% [0.4–0.9%] of women and 0.8% [0.6–1.3%] of men were living with HIV at the end of 2004.

Two countries in this region—Guatemala and Honduras—have national adult HIV prevalence of over 1%. But lower prevalence in other countries disguises the fact that serious, localized epidemics are also underway in several other countries—not least Brazil, which accounts for more than one third of the people living with HIV in Latin America.

Brazil’s epidemic has dispersed into all regions of this vast country, and displays some variation. At first affecting mainly men who have sex with men and then injecting drug users, the epidemic has grown more heterogeneous. Heterosexual transmission is now responsible for a growing share of HIV infections, with women increasingly affected (Marins et al., 2003). Lower socioeconomic status has been found to correlate strongly with higher prevalence among sex workers in Santos and Sao Paulo, according to one new study. Overall, 7% of the sex workers were HIV-positive, but among those living in urban slums, HIV levels were 18% and among illiterate women in their ranks they reached 23% (Gravato et al., 2004). Although national HIV prevalence among pregnant women has stayed stable at below 1% for the past five years, considerably higher levels have been found in some areas: 3% to 6% in a study conducted among pregnant women in Rio Grande do Sul state who did not regularly attend antenatal clinics. Most of these women had very-low incomes and were poorly educated (UNAIDS/WHO, 2003). The Brazilian government has introduced an initiative to recruit and offer testing to all pregnant women, provide services to prevent mother-to-child transmission and, if applicable, treat the women and their infants.

The role of injecting drug use in Brazil’s epidemic should not be underestimated. In some areas, injecting drug users constitute at least half of AIDS cases. Harm reduction programmes in some cities have been associated with steep drops in HIV prevalence among injecting drug users in recent years—notably in Salvador where prevalence fell from 50% in 1996 to 7% in 2001 (Ministerio da Saude do Brasil, 2001). In the south of the country, though, injecting drug users remain at high risk of HIV infection. Prevalence among users in Porto Alegre was 64% in 2003, while in Itajai it was 31%, indicating the need for more effective prevention programmes (Caiaffa et al., 2003).

HIV in Argentina remains concentrated largely in the urban areas of Buenos Aires, Cordoba and Santa Fe provinces, with an estimated 65% of HIV infections occurring in the capital Buenos Aires and its surrounding areas. However, the epidemic is changing. Throughout the 1980s and much of the 1990s, HIV transmission occurred mainly through injecting drug use, predominantly involving men. But sexual transmission of HIV—mainly from infected drug users to their female partners, as well as between men who have sex with men—has become more prominent, accounting for an estimated 80% of all reported AIDS cases. HIV prevalence in pregnant women was 0.4%, in 2002, and the male-to-female ratio among people living with HIV narrowed from 15:1 in 1988 to 3:1 in 2002. Most new infections appear to be occurring among the poorest and least-educated urban inhabitants (Ministerio de Salud Argentina, 2003; de los Pando et al., 2003). The paucity of prevention efforts in Argentina directed at men who have sex with men is a concern, given HIV prevalence of 14% detected among them in Buenos Aires and the fact that just one-in-seven of the men who tested positive had been aware of their serostatus (Avilla et al., 2004). In Uruguay, where about three quarters of all registered HIV cases have been in the capital, Montevideo, or its surrounding areas, there has been an alarming increase in the number of people living with HIV who are either injecting drug users or their sex partners. At least one quarter of HIV cases are injecting drug users and almost one half of them are younger than 25 (Osimani, 2003). A 2002 survey in Montevideo found almost 10% of injecting drug users were HIV-infected. Very high prevalence (21%) was measured a year earlier among men who have sex with men in the capital (US Bureau of the Census – HIV/AIDS Surveillance Database, 2003).

Until recently, the epidemics in the Andean area have been lodged largely among sex workers, their clients and men who have sex with men. However, this is beginning to change as the virus spreads increasingly to the wives and girlfriends of these men. One recent study in Lima, Peru, for example, found that almost 90% of HIV-positive pregnant women had had just one or two sex partners in their lifetimes (Alarcon et al., 2003). The women’s HIV risk depended almost exclusively on the sexual behavior of their male partners, and those most at risk were young women (Johnson et al., 2003). In a general population study in 24 Peruvian cities, 44% of men aged 18 to 29 years said they paid for sex (45% of them did not consistently use condoms with sex workers) and 12% said they had sex with other men (68% of them did not use condoms consistently in those encounters). Other urban Peruvian studies have found that 87% of men who had sex with men also slept with women, confirmed very-low rates of condom use, irrespective of the partner’s sex, and revealed high levels of sexually transmitted infections such as syphilis and herpes (Guanira et al., 2004). Given the consistently high HIV prevalence detected in recent years in groups of men who have sex with men in Peru—12% in Iquitos in 2002 and 22% in Lima in the same year—there is significant scope for wider HIV spread (MAP, 2003). Other research suggests that similar patterns of HIV transmission could be significant factors in epidemics elsewhere in the region. With an estimated 110 000 [47 000–170 000] people living with HIV at the end of 2003, Venezuela has one of the largest epidemics in the region. There, HIV is spreading mainly through unsafe sex, much of it between men, a significant proportion of whom also have sex with women (Minsterio de Salud y Desarrollo Social de Venezuela, 2003).

Little information is available on Ecuador’s epidemic, but behavioral studies in rural and urban areas point to several factors that could aid the spread of HIV—including early sexual debut (one survey showed 43% of high-school students were sexually active) and low rates of condom use (almost half the students never used one). Indeed, two recent studies (in 2002 and 2003) revealed HIV prevalence of 12% to 14% and 21% among men who have sex with men in Quito and Guayaquil, respectively. Much lower prevalence (under 2%) was found among female sex workers (National AIDS Programme Ecuador, 2002). Bolivia’s epidemic is concentrated largely among sex workers and their clients, and men who have sex with men. Most HIV infections reported to date have been in the cities of Santa Cruz and La Paz (Khalsa, Francis and Mazin, 2003). HIV prevalence among registered sex workers is very low—around 0.5%—but the health authorities believe the rate could be much higher among non-registered sex workers. If so, this is a concern in a country where 7% to 8% of adult men are believed to frequent sex workers. Among men who have sex with men, HIV prevalence is an estimated 3% to 5%. Little information on Colombia’s epidemic is available. Most recent data show low HIV prevalence among female sex workers in Bogotá (0.7% in 2001-2002), but very high prevalence among men who have sex with men (18% in 2000) (Khalsa, Francis and Mazin, 2003). In Central America, where the epidemic to a large extent is still concentrated in large urban areas, the numbers of HIV infections have been rising in several countries (including El Salvador, Nicaragua and Panama) since the late 1990s, but HIV prevalence remains highest in Guatemala and Honduras. Among people living with HIV, men outnumber women by roughly 3:1 in most countries.

Worst-affected is Honduras, where adult prevalence of almost 2% meant that an estimated 63 000 [35 000–110 000] people were living with HIV at the end of 2003 (UNAIDS, 2004). AIDS-related diseases are now estimated to be the second-leading cause of death in Honduras. The country’s epidemic has matured considerably, spreading among the wider population in some parts (such as Valle de Sula), while concentrated in other parts among sex workers and men who have sex with men and other vulnerable groups. HIV prevalence of 13% was measured in a study among men who have sex with men in 2001. Earlier studies found HIV levels of 7% among a sample of prisoners and 8.4% among the Garifuna population (Secretaria de Salud de Honduras, 2004).

In this subregion, HIV is spread predominantly through sex, and the highest levels of HIV infection are found in men who have sex with men and in female sex workers. Among the latter, HIV rates have varied considerably from about 1% in Nicaragua and Panama, to above 10% in Honduras and El Salvador. In most Central American countries, street-based sex workers are at least twice as likely to be HIV-infected as their counterparts working out of brothels, bars and hotels (various Ministries of Health, 2003; MAP Report, 2003). In Guatemala, HIV prevalence of 3.6% and 15% has been measured among brothel-based and street-based sex workers, respectively, while similar HIV levels (4% and 14%, respectively) have been found in Honduras (Secretaria de Salud de Honduras, 2003; Ministerio de Salud Pública y Asistencia Social de Guatemala, 2003). In El Salvador, HIV prevalence of 16% was found among street-based sex workers (in San Salvador and Puerto de Acajutla) (Ministerio de Salud Pública y Asistencia Social de El Salvador, 2003). It bears noting that, on the few occasions where prisoners have been tested for HIV, high prevalence has been found (7% in Honduras in the late 1990s, for example).

Sex between men is a major factor in the epidemics throughout the region, dramatically so in Costa Rica. There, more than half of AIDS cases in 1998–2002 were among men who have sex with men, a significant percentage of whom also have sex with women (Ministerio de Salud de Costa Rica, 2003). Roughly one third of HIV infections in Panama and Nicaragua are attributable to unprotected sex between men; in Panama, a 2002 study among men who have sex with men found almost 11% were infected with HIV (Ministerio de Salud de Nicaragua, 2004; Ministerio de Salud de Panama, 2004). Meanwhile, various studies have shown HIV levels in groups of men who have sex to be uniformly high in other countries, ranging from 9% to 13% in Guatemala, Honduras, Nicaragua and Panama to almost 18% in El Salvador. In each of these Central American countries, large proportions of men who have sex with men also report having female sexual partners (various Ministries of Health, 2003). Bisexuality therefore constitutes a significant bridge for HIV transmission into the wider population. Similarly, the wives or regular partners of sex worker clients face an elevated risk of HIV infection, even when they themselves have only one sex partner.

Mexico, national prevalence in the adult population has remained well under 1%, but shows marked regional variance. In the Baja California, District Federal, Quintana Roo and Yucatan states, prevalence hovers at 0.5%, while in Hidalgo, San Luis Potosi and Zacatecas states it is much lower, at an average 0.1% (Bravo-Garcia and Magis, 2004). In the past few years, much higher rates of HIV have been found among injecting drug users (up to 6%) and men who have sex with men (up to 15%). According to the country’s AIDS Registry, overall heterosexual transmission of HIV has increased in recent years. It is difficult to determine the extent to which high-risk behaviours such as injecting drug use or sex between men (which have been widely documented in Mexico) are contributing to the transmission of HIV in the country (Minichielloa et al., 2002).

Shadowing the considerable variation in Latin America’s epidemics, however, is a common and troubling pattern. In several countries, there is still a mismatch between prevention spending priorities and the main epidemiological features of countries’ epidemics. Most countries direct the bulk of their prevention expenditure to sex worker programming. Prevention spending does not yet reflect the fact that sex between men is a driving force in the epidemic throughout the region—with Peru the notable exception. The disparity is most pronounced in Central America. Meanwhile, among those countries where injecting drug use features prominently in their epidemics, only Argentina and Brazil appear to have prioritized their prevention spending accordingly . Much better use can be made of epidemiological and other pertinent data for designing tailored HIV prevention programming.

On the treatment front, Brazil remains a beacon among developing countries. It continues to offer all people living with HIV access to antiretroviral drugs via its national health system when they need it. As a result, the survival time of AIDS patients has increasingly dramatically. A recent study calculated that median survival was just under five years (58 months) for people diagnosed with AIDS in 1996 (Marins et al., 2003) while it was only 18 months for those diagnosed in 1995. AIDS cases and AIDS mortality have declined in several other countries, including Argentina, Costa Rica and Panama, after expansion of antiretroviral treatment access.

Oceania

An estimated 35 000 people [25 000–48 000] in Oceania are living with HIV. Although less than 700 [<1700] people are believed to have died of AIDS in 2004, about 5000 [2100–13 000] are thought to have become newly infected with HIV. Among young people 15–24 years of age, an estimated 0.2% of women [0.1–0.4%] and 0.2% of men [0.1–0.3%] were living with HIV by the end of 2004.

The annual number of new HIV diagnoses in Australia has gradually increased from 650 in 1998 to about 800 in 2002. A growing share of those diagnoses was in people who had become infected in the previous year—which suggests that the increase in new diagnoses could be linked to a revival of unsafe sex. The annual number of HIV diagnoses in women has stayed relatively stable, but more of those diagnosed infections occurred through heterosexual intercourse—either in a high-prevalence country or with a partner from a high-prevalence country. As is the case in New Zealand, transmission of HIV in Australia continues to be mainly through sexual intercourse between men, which accounted for more than 85% of new HIV diagnoses in the five years up to 2002. Injecting drug use was responsible for about 4% and heterosexual intercourse for 8.5% of newly acquired infections in that period. In a 2002 cross-sectional survey among men who have sex with men in Sydney, an increasing proportion of respondents reported unprotected anal sex with casual partners (25% compared with 18% in 1998-1999). Surveys in other cities have made similar findings. Recent gonorrhea surveillance data, too, have pointed to a possible increase in sexual risk behavior among men who have sex with men, underlining the need to reinvigorate prevention efforts aimed at men—especially young men—who have sex with other men (National Centre in HIV Epidemiology and Clinical Research, 2003).

The per capita rates of HIV diagnoses among Indigenous people in Australia since 1993 have been similar to those in non-Indigenous people. But higher proportions of diagnoses have been among women (36% in Indigenous compared to 11% in non-Indigenous people) and have been associated with injecting drug use (20% in Indigenous people compared to 4% in non-Indigenous people). At least half the estimated 14 000 people living with HIV in Australia are receiving antiretroviral therapy, reflecting both the age of the Australian epidemic and extensive treatment access.

Papua New Guinea, which shares an island with one of Indonesia’s worst-affected provinces, Papua, has the highest prevalence of HIV infection in the Pacific. An estimated 0.6% [0.3%–1.0%] of adults—roughly 16 000 [7800–28 000] people of the adult population of about 2.6 million—were living with HIV at the end of 2003 (UNAIDS, 2004). The annual number of new HIV infections detected in Papua New Guinea has been increasing progressively since the mid-1990s, exceeding 1000 in 2003, as Figure 24 illustrates (National AIDS Council Secretariat and Department of Health, 2003). In the same year, 1.4% of pregnant women at antenatal clinics in the capital Port Moresby tested HIV-positive, while in Lae, in the central highlands, 2.5% of pregnant women were HIV-infected (MAP, 2004). More than twice as many young women (aged 15–24 years) as men have been diagnosed with HIV. And in 2003 for the first time, overall, more HIV infections were detected in women than in men, as Figure 24 shows (National AIDS Council Secretariat and Department of Health, 2003).

Papua New Guinea’s HIV and AIDS surveillance capacity is limited and needs to be enhanced urgently. Available data suggest the epidemic is centered on commercial and casual sex, most of it heterosexual. High HIV prevalence has been found among sex workers (above 10% in the capital, Port Moresby, for example) (National AIDS Council Secretariat and Department of Health, 2003).

Recent household surveys of young men and women in Jayapura and Merauke, on the Indonesian side of the border, hint at some of the possible dynamics of HIV spread on the island overall. Unmarried women aged 15–24 in Papua were almost 10 times and young men five times more likely to be sexually active compared with their counterparts elsewhere in Indonesia (Indonesia Central Beureau of Statistics and MACRO International, 2004). In addition, 29% of sexually active young Papuan women reported having sex with men at least 10 years older than they were themselves. Because older men are more likely to be infected with HIV, such age mixing serves as a passageway for the virus from older to younger generations.

In an epidemic that centers mainly on commercial sex, the routes along which HIV can spread through the population are comparatively limited; most at risk are sex workers, their clients and the clients’ regular female partners. Where many men frequent sex workers, this can generate a serious epidemic—as Thailand discovered. But the combination of widespread commercial sex and multiple non-commercial partners seen in Papua New Guinea is ominous, since it enables the epidemic to assume much greater proportions. A high incidence of rape, sexual aggression and other forms of violence against women appear to be aiding the epidemic’s growth. According to one study, up to 70% of women have experienced domestic violence, while other studies have put the figure even higher (Brouwer, Harris, Tanaka, 1998). Many questions remain unanswered about Papua New Guinea’s epidemic but existing data highlight the need for urgent action to improve HIV prevention and AIDS care services if Papua New Guinea is to avert a rampant epidemic which will have ramifications for years to come.

HIV-infection levels appear to be very low in other parts of Oceania, but the data are extremely limited. On remote islands, seafarers and their partners appear to be most at risk; on Kiribati, for example, 9% of seafarers included in a recent study had Chlamydia and 3% syphilis, although HIV prevalence was still low (at 0.3%) (Sullivan et al., 2004). High rates of other sexually transmitted