HIV Prevalence Estimates from the Demographic and Health Surveys Updated June 2010
This reports summarizes the HIV prevalence estimates provided in MEASURE Demographic and Health Surveys. The MEASURE DHS project is implemented by ICF Macro, an ICF International Company, and funded by the United States Agency for International Development (USAID). The opinions expressed herein are those of the authors and do not necessarily refl ect the views of the USAID. This report was prepared by Erica Nybro and Bernard Barrere, ICF Macro. Recommended citation: ICF Macro. 2010. HIV Prevalence Estimates from the Demographic and Health Surveys. Calverton, Maryland: ICF Macro. For more information about the MEASURE DHS project, contact: ICF Macro 11785 Beltsville Drive, Calverton, Maryland, 20705 USA Telephone: 301-572-0200 Fax: 301-572-0999 www.measuredhs.com. Photos: Photos used in this report are representative of the countries highlighted, but are not necessarily of individuals living with HIV/AIDS. For a full list of photo credits and captions, see inside back cover.
Page 1 Introduction Demographic and Health Surveys The Demographic and Health Surveys project (MEASURE DHS) has been conducting surveys in Africa, Asia, Latin America, and Eastern Europe since 1984. DHS has earned a worldwide reputation for collecting and disseminating accurate, nationally representative data on fertility, family planning, maternal and child health, child survival, malaria, nutrition, and HIV/AIDS. DHS household surveys typically interview a nationally representative sample of over 10,000 women and men age 15-49. In recent years, blood tests have been added to the verbal interview to test for various health conditions, including anemia, and more recently, malaria and HIV. MEASURE DHS is implemented by ICF Macro based in Calverton, Maryland, and is funded by the United States Agency for International Development (USAID) and the President s Emergency Plan for AIDS Relief (PEPFAR). Individual surveys also receive funding from national governments, as well as multilateral and bilateral donors. HIV Estimates from the Demographic and Health Surveys The MEASURE DHS project has changed the way the world measures HIV prevalence. Prior to 2001, HIV prevalence was estimated largely from sentinel surveillance systems that monitored HIV rates in pregnant women attending antenatal care. This was, at the time, the best available proxy measure for national prevalence. In 2001, DHS included testing for HIV in the Mali Demographic and Health Survey, providing for the fi rst time ever HIV prevalence estimates from a nationally representative, population-based sample. Unlike sentinel surveillance, population-level testing includes men, nonpregnant women, and those without access to health facilities, thereby providing a representative sample of a national population. UNAIDS and national governments have adjusted their offi cial HIV prevalence estimates in response to the nationally representative HIV prevalence data provided through DHS surveys. Including HIV testing in the DHS also improves understanding of the HIV pandemic, because HIV test results are connected to the full DHS survey data. This allows for a more comprehensive examination of the sociodemographic and behavioral factors associated with HIV infection, such as age, education, residence, wealth, marital status, and higher-risk sexual behaviors than is possible with sentinel surveillance data. Between 2001 and 2009, DHS has included population-based HIV testing as part of 36 surveys in 31 countries, resulting in revised HIV estimates for these countries and a new understanding of the global HIV epidemic.
Page 2 Data included in this report This report includes data from MEASURE DHS surveys. It is an update of the 2008 document of the same name. The data are cross-sectional and provide a snapshot of the current situation in the 31 countries that included HIV testing in their surveys. The Dominican Republic, Kenya, Mali, Tanzania, and Zambia have had two surveys with HIV testing done since 2001. For these fi ve countries, the most recent survey fi ndings are presented throughout the report. Trends in HIV prevalence results are discussed on pages 20-21. For more information For more information on MEASURE DHS surveys, methodology, and results, visit: www.measuredhs.com To build your own tables using MEASURE DHS data, visit: STATcompiler (www.statcompiler.com) and the HIV/AIDS Survey Indicators Database (www.measuredhs.com/hivdata) For interactive mapping tools, visit STATmapper (www.statmapper.com) and HIVmapper (www.hivmapper.com)
Country Summary of Surveys with HIV Prevalence Type of survey Year Total number testedall ages Number of women 15-49 tested Number of men 15-49 tested Type of specimen Asia Cambodia DHS 2005 14,703 1 8,047 6,656 DBS India DHS 2005-06 102,946 2 53,332 46,506 DBS Vietnam (Hai Phong Prov.) AIS 2005 1,675 1 921 754 DBS Caribbean Dominican Republic* DHS 2002 22,725 3 10,732 10,707 Oral fluids Dominican Republic DHS 2007 51,109 3 24,564 23,513 DBS Haiti DHS 2005 10,062 3 5,230 4,321 DBS West/Central Africa Benin DHS 2006 9,599 4 5,025 3,949 DBS Burkina Faso DHS 2003 7,515 3 4,086 3,065 DBS Cameroon DHS 2004 10,352 1 5,227 4,672 DBS Cape Verde** DHS 2005 5,596 3 2,863 2,601 DBS Central African Republic** MICS 2006 10,592 3 5,413 4,657 DBS Congo (Brazzaville) AIS 2009 12,110 1 6,438 5,671 DBS Côte d Ivoire AIS 2005 8,436 1 4,413 4,023 DBS DR Congo DHS 2007 8,504 3 4,492 4,012 DBS Ghana DHS 2003 9,566 3 5,097 4,047 DBS Guinea DHS 2005 6,836 3 3,772 2,616 DBS Liberia DHS 2007 11,733 1 6,382 5,351 DBS Mali* DHS 2001 6,846 3 3,854 2,978 DBS Mali DHS 2006 8,629 3 4,528 3,614 DBS Niger DHS 2006 7,673 3 4,406 2,856 DBS São Tomé and Principe DHS 2008-09 4,710 3 2,378 2,110 DBS Senegal DHS 2005 7,823 3 4,278 3,226 DBS Sierra Leone DHS 2008 6,475 3 3,448 2,726 DBS East Africa Ethiopia DHS 2005 11,042 3 5,736 4,804 DBS Kenya DHS 2003 6,194 2 3,151 2,851 DBS Kenya DHS 2009 6,906 2 3,641 3,066 DBS Rwanda DHS 2005 10,419 3 5,656 4,361 DBS Tanzania AIS 2003-04 10,747 1 5,753 4,994 DBS Tanzania AIS/MIS 2007-08 15,044 1 8,179 6,865 DBS Uganda AIS 2004-05 18,525 5 9,391 7,515 Venous & DBS Children 0-4: 8,374 4,226 4,148 DBS Southern Africa Lesotho DHS 2004-05 5,286 3 3,031 2,012 DBS Malawi DHS 2004 5,266 2 2,686 2,465 DBS Swaziland DHS 2006-07 9,177 6 4,424 3,763 DBS Children 2-14: 3,589 1,841 1,747 DBS Zambia* DHS 2001-02 3,950 3 2,073 1,734 DBS from venous Zambia DHS 2007 10,876 3 5,502 4,424 DBS Zimbabwe DHS 2005-06 13,049 2 6,947 5,848 DBS DHS: Demographic and Health Survey; AIS: AIDS Indicator Survey; MICS: Multiple Indicator Cluster Survey (UNICEF); MIS: Malaria Indicator Survey; DBS: dried blood spots *HIV prevalence not linked to full DHS survey; **Non-MEASURE surveys; 1 women 15-49 and men 15-49; 2 women 15-49 and men 15-54; 3 women 15-49 and men 15-59; 4 women 15-49 and men 15-64; 5 women and men 15-59, 6 women 15+, men 15+
Page 4 Summary of HIV Testing Since 2001, more than 480,000 women and men worldwide have been tested for HIV through the MEASURE DHS project s Demographic and Health Surveys and AIDS Indicator Surveys. HIV testing is usually conducted on women and men age 15-49, although some countries have tested children, and others have tested older adults. The sample size for testing ranges from 1,675 in Hai Phong Province, Vietnam, to over 102,000 in India. Testing Protocol The DHS HIV testing protocol provides for informed, anonymous, and voluntary testing of women and men interviewed. The testing protocol undergoes a host country ethical review as well as an ethical review at ICF Macro. In countries with CDC involvement, the testing protocol is also reviewed by CDC. The testing is simple; in most cases, the interviewer collects dried blood spots (DBS) on fi lter paper from a fi nger prick and the fi lter paper is transported to a laboratory for testing. The laboratory protocol includes an initial ELISA test, and then retesting of all positive tests and 5% to 10% of the negative tests with a second ELISA. For those tests with discordant results on the two ELISA tests, another test, usually a Western blot, is used to determine the result.
Page 5 HIV Prevalence: Adults Age 15-49 Asia Caribbean West/ Central Africa Cambodia 2005 India 2005-06 Vietnam (Hai Phong Prov.) 2005 Dominican Republic 2007 Haiti 2005 Benin 2006 Burkina Faso 2003 Cameroon 2004 Cape Verde 2005 Central African Republic 2006 Côte d Ivoire 2005 Congo 2009 DR Congo 2007 Ghana 2003 Guinea 2005 Liberia 2007 Mali 2006 Niger 2006 0.8 1.2 0.4 1.3 1.5 1.5 1.2 0.7 2.2 1.8 2.2 São Tomé and Principe 2008-09 1.5 Senegal 2005 0.7 Sierra Leone 2008 0.6 0.28 0.5 1.5 3.2 5.5 4.7 6.2 National HIV Prevalence National HIV prevalence rates are highest in the countries surveyed in southern Africa, with more than 14% of women and men age 15-49 infected with HIV in four countries (Lesotho, Swaziland, Zambia, and Zimbabwe). East Africa has more moderate rates of HIV infection, ranging from 1.4% in Ethiopia to 6.4% in Uganda. Most countries in West/Central Africa have HIV prevalence rates under 2%, although the rates in Cameroon, Central African Republic, Congo, and Côte d Ivoire are higher. HIV prevalence in Asia is quite low fewer than 1% of adults tested in Cambodia, India, and Hai Phong Province, Vietnam are HIV-positive. HIV prevalence is also low in the Caribbean, where about 1% of adults in the Dominican Republic are infected, and just over 2% are infected in Haiti. Ethiopia 2005 1.4 East Africa Kenya 2009 Rwanda 2005 3.0 6.3 Tanzania 2007-08 5.7 Uganda 2004 6.4 Lesotho 2004 23.5 Southern Africa Malawi 2004 Swaziland 2006-07 Zambia 2007 11.8 14.3 25.9 Zimbabwe 2005-06 Percent of women and men 15-49 18.1
Page 6 National HIV Prevalence Percent of women and men age 15-49 who are HIV-Positive
Page 7 HIV Prevalence in Africa at the Subnational Level In most cases, DHS data provide HIV prevalence estimates at the regional (subnational) level. Some countries have fairly uniform prevalence; Lesotho, Swaziland, Zambia, and Zimbabwe have relatively high HIV prevalence across all regions, while much of West Africa is consistently low. There is greater regional variation in Cameroon, Central African Republic, Côte d Ivoire, Ethiopia, Ghana, Kenya, Rwanda, Tanzania, and Uganda.
Page 8 HIV Prevalence by Sex Cambodia 2005 0.6 0.6 Asia India 2005-06 0.22 0.36 Vietnam (Hai Phong Prov.) 2005 0.2 0.9 Dominican Republic 2007 Caribbean Haiti 2005 0.8 0.8 2.3 2.0 Benin 2006 1.5 0.8 Burkina Faso 2003 1.8 1.9 Cameroon 2004 6.8 4.1 Cape Verde 2005 0.3 0.5 Central African Republic 2006 4.3 7.8 Congo 2009 2.1 4.1 West/ Côte d Ivoire 2005 6.4 2.9 Central DR Congo 2007 1.6 0.9 Africa Ghana 2003 2.7 1.5 Guinea 2005 1.9 0.9 Liberia 2007 1.8 1.2 Mali 2006 1.4 0.9 Niger 2006 0.7 0.7 São Tomé and Principe 2008-09 1.3 1.7 Senegal 2005 0.9 0.4 Sierra Leone 2008 1.7 1.2 HIV Prevalence by Sex In most countries in sub-saharan Africa where HIV is mainly transmitted via heterosexual contact, HIV prevalence is higher among women than men. Prevalence among women is highest in Swaziland, where over 30% of women tested are HIV-positive, compared to 20% of men. In contrast, in Asia, where HIV is mainly concentrated in high-risk groups, HIV prevalence is slightly higher among men than women. There is little difference between men and women in the Caribbean. In general, the difference between women and men is smaller in countries where overall HIV infection rates are low. East Africa Ethiopia 2005 Kenya 2009 Rwanda 2005 Tanzania 2007-08 Uganda 2004 1.9 0.9 8.0 4.3 3.6 2.3 4.6 6.6 5.0 7.5 Women Men Southern Africa Lesotho 2004 Malawi 2004 Swaziland 2006-07 Zambia 2007 Zimbabwe 2005-06 13.3 10.2 16.1 12.3 14.5 19.2 19.7 21.1 26.4 31.1 Percent of women and men 15-49
Page 9 Caribbean HIV Prevalence by Residence Cambodia 2005 0.4 1.4 Asia India 2005-06 0.25 0.35 Vietnam (Hai Phong Prov.) 2005 0.3 1.0 Dominican Republic 2007 Haiti 2005 Benin 2006 Burkina Faso 2003 Cameroon 2004 Central African Republic 2006 1.0 0.7 2.0 2.3 0.9 1.7 1.3 Cape Verde 2005 0.3 0.5 3.6 4.0 4.7 Congo 2009 2.8 West/ 3.3 4.1 Central Côte d Ivoire 2005 5.4 Africa DR Congo 2007 0.8 1.9 Ghana 2003 2.0 2.3 Guinea 2005 1.0 2.4 Liberia 2007 0.8 2.5 Mali 2006 0.9 1.6 Niger 2006 0.5 1.4 São Tomé and Principe 2008-09 0.8 2.2 Senegal 2005 0.7 0.7 Sierra Leone 2008 1.0 2.5 6.7 8.3 HIV Prevalence by Residence In most countries, prevalence is higher in urban areas than in rural areas. The Dominican Republic, São Tomé and Principe, and Senegal are the only exceptions. The difference between urban and rural areas is most dramatic in Ethiopia, where the proportion of women and men who are HIVpositive in urban areas is almost eight times higher than among those living in rural areas. East Africa Ethiopia 2005 Kenya 2009 Rwanda 2005 Tanzania 2007-08 Uganda 2004 0.7 2.2 5.5 6.0 7.2 7.3 4.7 8.7 5.7 10.1 Rural Urban Southern Africa Lesotho 2004 Malawi 2004 Swaziland 2006-07 Zambia 2007 Zimbabwe 2005-06 10.8 10.3 17.1 19.7 17.6 18.9 21.9 23.8 29.1 31.4 Percent of women and men 15-49
Page 10 HIV Prevalence by Age In general, HIV prevalence increases until about age 30-34 for women, and then declines. Among men, prevalence is lower than women in the early years and then hits its peak in the late 30s or early 40s. In some countries HIV prevalence is quite high among older adults; more than 10% of adults age 45-49 in Swaziland, Zimbabwe, and Malawi are HIVpositive. 50 40 30 HIV Prevalence by Age: Women Percent 20 10 0 Swaziland 2006 Malawi 2004 Zambia 2007 Tanzania 2007-08 Cameroon 2004 Haiti 2005-06 15-19 20-24 25-29 30-34 35-39 40-44 45-49 In Zambia, for example, HIV prevalence among women peaks at age 30-34, while men s prevalence peaks at age 40-44. 50 HIV Prevalence by Age: Men 40 Percent 30 20 10 0 Swaziland 2006 Zambia 2007 Malawi 2004 Tanzania 2007-08 Cameroon 2004 Haiti 2005-06 15-19 20-24 25-29 30-34 35-39 40-44 45-49
Page 11 Testing of Children Some countries also choose to test younger children. Uganda tested children under age 5, while in Swaziland, children age 2-14 were tested. Prevalence among children under 5 in Uganda is less than 1%. Younger children have a higher rate of infection than older children. In Swaziland, 5% of the youngest children are HIV-positive. Prevalence decreases with age among children, as those who are born with HIV have usually died before their early teen years, and have not yet been exposed to HIV through sexual activity. Uganda and Swaziland: HIV Prevalence among Children Percent Percent Girls 1.0 1.1 1.0 Boys Uganda 2004-05 <18 months Total 0.5 0.5 0.5 18-59 months Swaziland 2006-07 5.5 4.8 5.1 4.8 4.2 3.6 3.3 2.6 1.9 2-4 5-9 10-14 Age in years Swaziland: HIV Prevalence among Older Adults Percent Women 28.3 26.0 24.3 9.6 Men 17.4 12.7 Total 13.3 9.5 7.0 Testing of Older Adults While many countries test women up to age 49 and men up to age 55 or 59, the 2006-07 Swaziland DHS included testing of all women and men in the household. In Swaziland, one in four adults age 50-54 and one in ten adults 60 and over is infected with HIV. 50-54 55-59 60+ Age
Page 12 HIV Prevalence among Youth 15-24 Asia Caribbean Cambodia 2005 0.3 0.1 India 2006 0.11 0.09 Vietnam (Hai Phong Prov.) 2005 too few cases Dominican Republic 2007 Haiti 2005 0.4 0.2 1.5 0.6 Benin 2006 1.0 0.3 Burkina Faso 2003 1.3 0.7 Cameroon 2004 1.4 Cape Verde 2005 <0.1 0.2 Central African Republic 2006 1.0 Congo 2009 0.7 2.4 West/ Côte d Ivoire 2005 2.4 0.3 Central 0.5 Africa DR Congo 2007 1.0 Ghana 2003 1.2 0.1 Guinea 2005 1.2 0.6 Liberia 2007 1.6 0.5 Mali 2006 0.9 0.5 Niger 2006 0.5 0.1 São Tomé and Principe 2008-09 1.0 0.6 Senegal 2005 0.1 Sierra Leone 2008 0.5 1.4 4.8 5.7 HIV Prevalence among Youth HIV prevalence among youth is an important indicator of the future of the epidemic. In most countries, HIV prevalence among youth is quite low. In high-prevalence countries such as Lesotho, Swaziland, and Zimbabwe, however, more than 10% of young women are already infected with HIV. The difference between young women and men is especially striking in Cameroon and the Central African Republic. It is important to note that in almost all countries, young women are far more likely to be infected than young men. This is related to the fact that women marry and have sex earlier than men in most countries. East Africa Ethiopia 2005 Kenya 2009 Rwanda 2005 Tanzania 2007-08 Uganda 2004-05 1.1 0.2 1.1 1.5 0.4 1.1 1.1 4.5 3.6 4.3 Women Men Southern Africa Lesotho 2004 6.0 Malawi 2004 9.1 2.1 Swaziland 2006 5.9 Zambia 2007 8.5 4.3 Zimbabwe 2005-06 11.0 4.2 Percent of women and men 15-24 15.4 22.7
Page 13 HIV Prevalence among Youth by Age Women Men Cameroon 2004 Haiti 2005-06 5.7 3.0 2.7 1.6 0.5 0.8 15-17 18-19 20-22 2.6 1.2 0.7 0.0 0.4 0.5 15-17 18-19 20-22 Mali 2006 0.8 0.3 0.4 0.2 1.2 0.8 11.8 1.7 2.2 23-24 2.3 23-24 1.4 0.8 Older Youth at Increased Risk HIV prevalence is much lower among the youth age 15-17 than those age 23-24 in most countries and among both women and men. This indicates that HIV prevalence increases quickly with age. This increase in prevalence mirrors the increased exposure to risk. As young people become sexually active and have more sexual partners they are more likely to be exposed to HIV. Young women are especially vulnerable to HIV due to their early age of marriage and sexual debut. Young women age 23-24 in Cameroon, for example, are 7 times as likely to be HIV-positive as young women age 15-17. 15-17 18-19 20-22 23-23 Uganda 2004-05 5.5 3.9 1.9 2.3 0.3 0.2 15-17 18-19 20-22 7.7 2.5 23-24 Zimbabwe 2005-06 22.3 9.9 12.6 10.6 3.4 2.9 3.3 3.2 15-17 18-19 20-22 23-24
Page 14 HIV Prevalence by Education There is no consistent pattern in HIV prevalence by education. In Ethiopia and Zambia, as well as Burkina Faso, Cameroon, the Central African Republic, Congo, Rwanda, and Uganda, HIV prevalence is higher among educated men and women. In other countries this association with education occurs only for men (Malawi) or only for women (the Democratic Republic of the Congo and Guinea). In contrast, HIV prevalence is lower among educated women and men in the Dominican Republic, Haiti, India, Senegal, Tanzania, and Zimbabwe, although in these cases, the pattern is more evident for men than for women. In some countries, there is no pattern at all (Cambodia, Côte d Ivoire, Ghana, Kenya, Lesotho, Mali, Niger, Sierra Leone, and Swaziland). Percent among women and men 15-49 Ethiopia and Zambia: HIV Prevalence Increasing with Education Women Men Ethiopia 2005 Zambia 2007 5.5 1.0 0.8 2.5 0.5 2.0 No education Primary Secondary + 10.8 7.7 No education 15.8 11.0 Primary 17.4 13.2 Secondary 21.3 17.6 More than secondary In both Ethiopia and Zambia, HIV prevalence rises with each subsequent level of education for both women and men. Percent among women and men 15-49 Dominican Republic and Zimbabwe: HIV Prevalence Decreasing with Education Women Men 3.7 1.6 1.3 1.6 1.0 0.8 No education/ preschool Dominican Republic 2007 Zimbabwe 2005-06 Primary 1-4 Primary 5-8 0.3 0.5 Secondary/ university 23.4 20.0 No education 22.4 15.0 Primary incomplete 20.7 14.3 Primary complete 15.8 12.8 Secondary + In Zimbabwe, HIV prevalence among men with no education is twice as high as prevalence among those with secondary or more education.
Page 15 HIV Prevalence by Wealth For most countries, prevalence is higher among women and men in the wealthiest households than among those in the poorest households. The relationship between wealth and HIV prevalence is generally stronger among women than among men. There are some exceptions to this pattern, however. In the Dominican Republic, Haiti, Benin, Congo, Ghana, São Tomé and Principe, and Zimbabwe there is no clear relationship between wealth and HIV infection. HIV Prevalence by Household Wealth Quintile Cameroon 2004 Poorest Lowest Second Middle Richest Fourth Highest Ethiopia 2005 Lowest Poorest Second Middle Fourth Richest Highest 0.3 0.7 1.0 0.3 0.4 0.9 3.1 1.4 4.1 2.2 4.7 5.3 5.3 0.2 0.4 2.2 6.1 8.1 9.4 8.0 Women Men HIV prevalence tends to increase as household wealth increases. In some countries, prevalence increases incrementally with wealth, while in others, like Ethiopia, only the wealthiest have a markedly higher rate of HIV infection. Malawi 2004 Poorest Lowest Second Middle Fourth Richest Highest Rwanda 2005 Poorest Lowest Second Middle Fourth Richest Highest 4.4 4.6 2.6 1.3 2.2 1.7 3.6 2.0 3.4 2.1 4.1 6.5 10.9 10.3 12.7 12.1 14.6 11.7 14.9 18.0 Household wealth is based on ownership of assets, materials used for housing construction, water access, and sanitation facilities. Sierra Leone 2008 Lowest 0.8 Poorest 0.2 Second 1.9 0.5 Middle 1.0 1.0 Fourth 2.3 Richest 1.9 Highest 2.4 2.1 Percent of women and men 15-49
Page 16 HIV Prevalence by Marital Status The relationship between marital status and HIV prevalence is consistent across almost all DHS countries: women and men who are divorced/separated or widowed are far more likely to be HIVpositive than those who are currently married. And women and men who have never been married are least likely to be infected. Despite low national prevalence, the relationship between HIV infection and marital status in India is still strong. Widowed women are more than 6 times as likely to be HIV-positive as currently married women. Men who are separated or divorced are 4 times as likely to be HIV-infected as married men. India 2005-06: HIV Prevalence by Marital Status Women Men 1.9 1.5 1.1 HIV Prevalence by Marital Status Benin 2006 Cambodia 2005 Congo 2009 Kenya 2009 Women 0.5 0.2 1.3 1.2 Never Currently in union in union 0.1 0.1 Never in union Men 6.6 4.2 3.3 2.6 2.8 1.1 Never in union Currently in union 7.2 2.3 Divorced/ separated 2.0 0.7 0.9 1.1 Currently in union Divorced/ separated Divorced/ separated 10.6 too few cases Widowed 2.9 Widowed 14.4 43.1 4.1 too few cases Widowed 0.5 <0.1 0.2 0.2 Never Currently married married Divorced/ separated 0.1 Widowed 3.7 1.5 Never in union 6.7 6.0 Currently in union 16.8 9.7 Divorced/ separated too few cases Widowed Zimbabwe 2005-06 66.7 57.7 35.8 35.5 8.4 4.3 19.8 22.7 Never in union Currently in union Divorced/ separated Widowed
Page 17 HIV Prevalence among Couples The best way to avoid contracting HIV is to avoid having sexual intercourse with a partner who is HIV-positive. However, in many married and cohabiting couples, one partner is positive. These couples are called discordant because one partner is HIV-positive and the other is negative. Identifying discordant couples is an essential step in reducing the spread of HIV. These couples must take extra steps to avoid transmission. Discordant Couples Asia Cambodia 2005 India 2006 0.5 0.39 Woman positive, man negative Man positive, woman negative Caribbean Dominican Republic 2007 Haiti 2005 1.2 0.8 2.0 3.2 West/ Central Africa Benin 2006 Burkina Faso 2003 0.7 0.7 1.4 Cameroon 2004 Côte d Ivoire 2005 Ghana 2003 Guinea 2005 0.9 0.9 Mali 2006 0.80.31.1 1.8 2.7 2.6 Congo 2009 2.9 1.8 3.7 2.3 1.5 1.6 0.7 1.0 1.7 Niger 2006 0.40.61.0 São Tomé and Principe 2008-09 0.4 2.1 2.5 Sierra Leone 2008 1.2 0.7 1.9 3.1 4.7 5.3 6.0 Lesotho: HIV Prevalence in Couples Both HIV negative, 66% Both HIV positive, 20% Man positive, woman negative, 9% Woman positive, man negative, 5% One in seven couples in Lesotho is discordant; both partners are HIV-positive in one in fi ve couples. Only two-thirds of couples are HIV-free. East Africa Ethiopia 2005 Kenya 2009 Rwanda 2005 Uganda 2004 1.0 0.8 0.8 1.4 1.8 3.2 2.5 2.2 Tanzania 2007-08 2.8 3.5 1.8 2.8 4.6 5.7 6.3 Lesotho 2004 4.5 8.9 14.3 Malawi 2004 4.0 5.7 9.7 Southern Africa Swaziland 2006 8.7 7.7 Zambia 2007 4.6 6.6 11.2 Zimbabwe 2005-06 5.2 8.1 13.3 16.4 Percent of couples
Page 18 Cameroon 2004 1 partner Ethiopia 2005 1 partner Guinea 2005 1 partner HIV Prevalence by Number of Sexual Partners in Past Year Women Men 2 or more partners 2 or more partners 2 or more partners 1.9 1.2 2.2 0.8 3.3 1.6 6.9 3.7 4.0 10.3 7.1 13.2 HIV Prevalence by Sexual Behavior Sex with multiple partners is associated with higher levels of HIV prevalence among women and men in Cameroon, Ethiopia, Guinea, India, Lesotho, Malawi, Rwanda, Sierra Leone, Tanzania, Uganda, and Zambia. This relationship also holds true for women, but not men, in Burkina Faso, Côte d Ivoire, Ghana, Haiti, and Zimbabwe. Uganda 2004-05 1 partner 2 or more partners 7.3 5.5 8.1 14.0 Malawi 2005 1 partner 2 or more partners 13.3 11.0 16.2 43.7 Rwanda 2005 1 partner 2 or more partners 3.5 3.3 7.9 4.1 Zambia 2007 1 partner 2 or more partners 16.1 12.8 20.0 Percent of women and men 15-49 31.7
Page 19 HIV Prevalence by Number of Lifetime Partners Women and men with more lifetime sexual partners are more likely to have been exposed to HIV. DHS data confi rm the association between number of lifetime partners and HIV prevalence in almost all of the surveys for which this information was collected: Cambodia, Cameroon, Congo, Côte d Ivoire, the Dominican Republic, Ethiopia, Guinea, Kenya, Mali, Niger, Rwanda, Swaziland, Tanzania, Zambia, and Zimbabwe. HIV Prevalence by Lifetime Number of Sexual Partners Women Men Cameroon 2004 Dominican Republic 2007 12.4 14.2 5.0 Percent 8.2 9.3 5.3 2.7 2.3 3.3 0.5 1 2 3 or 4 5-9 Number of partners 5.5 10+ Percent 1.7 1.3 0.9 0.6 0.7 0.9 0.2 0.2 1 2 3 or 4 5-9 Number of partners 1.1 10+ In Cameroon, risk of HIV infection increases steadily as women and men have more sexual partners. As in most countries, the association between number of partners and HIV prevalence is especially strong for women in the Dominican Republic. Guinea 2005 16.0 Zimbabwe 2005-06 42.2 43.9 37.1 31.1 Percent 5.1 1.0 0.9 2.2 0.5 1.5 0.1 1 2 3 or 4 5-9 Number of partners too 2.6 few cases 10+ In Guinea, 16% of women with 5 to 9 lifetime sexual partners are HIV-positive, compared to only 1% of those with only one lifetime sexual partner. Percent 18.1 6.6 14.8 20.3 1 2 3 or 4 5-9 Number of partners 22.1 too few cases 10+ Prevalence is high even among those with only one partner in Zimbabwe, and yet HIV-infection rates more than double as women and men have more lifetime partners.
Page 20 Trends in HIV Prevalence Since 2001, MEASURE DHS surveys have included HIV testing in 31 countries. In fi ve of these countries Dominican Republic, Kenya, Mali, Tanzania, and Zambia testing has been included in two surveys, which provides the opportunity to examine trends. However, trend data must be viewed with caution, as only some changes are statistically signifi cant. In the charts below, changes in HIV prevalence that are statistically signifi cant are marked with an asterisk. For further discussion of statistical signifi cance and confi dence intervals, see the Tanzania example on page 21. HIV Prevalence Trends Dominican Republic 2002 2007 Mali 2001 2006 1.0 0.9 1.1 0.8* 0.8 0.8 1.7 2.0 1.2 1.4 1.3 0.9 Total Women Men Total Women Men Tanzania (Mainland) 2003-04 2007-08 Kenya 2003 2009 Zambia 2001-02 2008 7.0 7.7 5.8* 6.8 6.3 4.7* 6.7 6.3 8.7 8.0 4.6 4.3 15.6 14.3 17.8 16.1 12.912.3 Total Women Men Total Women Men Total Women Men While it may appear that HIV prevalence has decreased in all fi ve countries, these decreases are only statistically signifi cant in Tanzania and the Dominican Republic. It cannot be concluded that HIV prevalence has truly decreased in Mali, Kenya, or Zambia.
Page 21 Understanding Trends in HIV Prevalence: Tanzania Case Study The 2003-04 Tanzania HIV/AIDS Indicator Survey (THIS) was the fi rst time HIV testing was carried out in a national, population-based survey in Tanzania. The 2003-04 THIS tested 5,753 women and 4,994 men age 15-49. The THIS reported that HIV prevalence in Mainland Tanzania* was 7.0%: 7.7% among women and 6.3% among men. It is important to note, however, that all surveys have a certain degree of error, especially for estimating events that are not very common. The reported national prevalence in the 2003-04 THIS of 7.0% is only an estimate. Based on statistical calculations, analysts are confi dent that the true HIV prevalence lies within a range, called the confi dence interval. This range in the 2003-04 THIS is 6.3%-7.8%. In other words, statisticians are confi dent that the actual HIV prevalence in Tanzania in 2003-04 was between 6.3% and 7.8%; 7.0% marks the middle of that range. In 2007-08, the National Bureau of Statistics implemented the Tanzania HIV and Malaria Indicator Survey (THMIS), the second national survey to include population-level HIV testing. The THMIS tested 8,179 women and 6,865 men age 15-49 for HIV, using the same methods as in the 2003-04 THIS. The 2007-08 survey reported that 5.8% of Tanzanians living on the Mainland are HIV-positive. While this is lower than the result of 7.0% reported in 2003-04, it is important to take a closer look at the confi dence intervals (shaded areas in graph at right). In 2003-04 the confi dence interval was 6.3% to 7.8%. For 2007-08, the confi dence interval is 5.3% to 6.4%. These two ranges overlap, but just barely, suggesting that the decrease in the two estimates represents a true change in the population. Because there is some overlap between the confi dence intervals, a second test is required to assess statistical signifi cance. In the case of Tanzania, the second test concluded that the decrease in overall HIV prevalence between the two surveys is indeed signifi cant. The case for men HIV Prevalence Trends for Mainland Tanzania: Confidence Intervals 2003-04 THIS 2007-08 THMIS Total 15-49 8.6 7.6 7.7 5.8 6.7 6.0 Women 15-49 Men 15-49 is similar: the confi dence intervals for men just barely overlap, and the second statistical test confi rmed that this decrease is statistically signifi cant. One cannot say, however, that prevalence among women has decreased. While prevalence among women appears to have changed, the confi dence intervals for the two surveys prevalence rates overlap quite a bit. Therefore, it is not possible to say with confi dence that the lower HIV prevalence found among women in the 2007-08 survey represents a real decline in the population. *In 2003-04, the THIS included testing only in Mainland Tanzania. In 2007-08, the THMIS included testing in Zanzibar as well as Mainland Tanzania. For the purpose of comparison, all fi gures on this page have been limited to Mainland Tanzania. 7.0 7.8 6.3 6.4 5.3 6.8 6.3 7.1 5.4 5.4 3.9 4.7
Page 22 The Way Forward Survey data provide valuable insights into the patterns of HIV infection, allowing countries to channel scarce resources and design more effective prevention and care programs. MEASURE DHS has helped countries conduct HIV testing in population-based surveys for almost 10 years. Interest in population-based testing continues to grow in both high and low prevalence countries. MEASURE DHS surveys with HIV testing are currently planned or underway in Angola, Burkina Faso, Burundi, Cameroon, Côte d Ivoire, Gabon, Lesotho, Malawi, Mozambique, Rwanda, Uganda, and Zimbabwe. Several of these countries are doing HIV testing for the second time. not clear cut. Has HIV prevalence decreased because the epidemic has slowed, or because many HIV-positive individuals have died? Has prevalence increased because there are more new cases of the disease or because so many more are surviving on anti-retroviral treatment? DHS data must be considered along with other sources of data to assess the true HIV situation in a given context and develop appropriate, evidence-based interventions. Such evidence-based programming will increase effectiveness and effi ciency, leading to improved prevention of HIV transmission and treatment of those living with HIV/AIDS. While trend data are important, they must be interpreted with caution. Even in a case where a change in HIV prevalence is statistically signifi cant, interpretation of these results is
Photo credits Cover photo: 2009 Gaurav Gaur/Social Activist, Courtesy of Photoshare. Love Life, Stop AIDS is written out with earthen lamps to pay homage to all those who have died due to HIV/AIDS on International AIDS Candlelight Day, at Government Model Senior Secondary School, Sector 22-A in Chandigarh, India. Page 1: 2008 Alexandria Smith, Courtesy of Photoshare. A woman gives blood during a voluntary counseling and testing (VCT) outreach provided by Hope Clinic Lukuli in Ggaba, Uganda.Ggaba is a fi shing village in Kampala with a high HIV rate due to the transient nature of fi sherman. Page 2: 2009 David Snyder, Courtesy of Photoshare. A pharmacist at a clinic in Nairobi, Kenya s sprawling Kibera slum displays some of the ARV medications in-stock for HIV patients who visit the clinic. The clinic is supported by the Centers for Disease Control, which conducts disease surveillance of the surrounding community. Page 4: Alfredo Fort, ICF Macro. Page 9: 2008 Bob Msangi, Courtesy of Photoshare. A man in Tanzania listens to an HIV/AIDS radio program as part of the STRADCOM (Strategic Radio Communication for Development) project. Page 11: Jasbir Sangha, ICF Macro. Page 13: 2007 Pradeep Tewari, Courtesy of Photoshare. An adolescent peer volunteer demonstrates correct condom use to the students of Government Sr.Sec.School, Khuda Lahora, Chandigarh, India, under the Adolescent Development and Empowerment Project run by the Ministry of Sports and Youth Welfare, Government of India. Page 15: Kia Reinis, ICF Macro. Household water collection in Cambodia. Page 16: 2009 Robyn Iqbal, Courtesy of Photoshare. A young HIV positive woman and her second husband await treatment in the HIV ward of a large municipal hospital in Tamil Nadu, India. Born into a poor family, she was married off at 15 to her fi rst husband (not pictured) who physically abused her, burned her with cigarettes, hit her with pans and threw away her meager income on alcohol. She could not choose her own clothes, go out with friends, or even turn to her family for help because he paid them off. Her husband had many women and infected her with HIV at age 16. She had two children (both HIV negative) before fi nally divorcing at 19, and her father took her children from her for fear she would infect them. Page 18: 2008 Mali Health Organizing, Courtesy of Photoshare. A community health worker sells raffle tickets at Mali Health Organizing Project s annual World AIDS Day Celebration. Page 22: 2009 Community Peer Educators, Courtesy of Photoshare. Health workers in Uganda conduct HIV testing as the community mobilizers look on. Inside back cover: 2009 Netsanet Assaye, Courtesy of Photoshare. A voluntary counseling and testing center in Majengo, Nairobi, Kenya. Commercial sex workers visit often to get free condoms and tested for their HIV status. Back cover: 2009 Cindi Cohen, Courtesy of Photoshare. A community health prevention volunteer demonstrates the proper use of condoms to a group of women from the rural community of Tevele, Massinga district, Mozambique. The community of Tevele has formed a Community Health Initiative to address the most signifi cant health issues in their community, including HIV/AIDS.