HIV/AIDS in Tanzania Lisa Garbus, MPP AIDS Policy Research Center, University of California San Francisco Published July 2004. (c) 2004 Regents of the University of California. All Rights Reserved. Table of Contents PREFACE 4 ACKNOWLEDGMENTS 4 EXECUTIVE SUMMARY 6 EPIDEMIOLOGY 8 HIV SENTINEL SURVEILLANCE 8 METHODOLOGY 8 TRANSMISSION PATTERNS 10 HIV INCIDENCE 10 BLOOD DONORS 11 AIDS CASES 12 AIDS MORTALITY 12 U.N. ESTIMATES 13 POLITICAL ECONOMY AND SOCIOBEHAVIORAL CONTEXT 16 OVERVIEW 17 COLONIAL ERA 18 POSTCOLONIAL CONTEXT 18 ECONOMY 19 POVERTY 20 DEBT 21 GOVERNANCE 22 POPULATION DYNAMICS 23 HUMAN DEVELOPMENT 23 MOBILE AND DISPLACED POPULATIONS 24 EDUCATION 28 HEALTH SYSTEM 28 SEXUAL & REPRODUCTIVE HEALTH 30 SEXUALLY TRANSMITTED INFECTIONS 33 TUBERCULOSIS 35 GENDER 35 SEXUAL VIOLENCE 38 STIGMA AND DISCRIMINATION 40 AWARENESS AND KNOWLEDGE OF HIV/AIDS 41 SEXUAL BEHAVIOR 42 TRANSACTIONAL SEX 46 SEX WORK 47 MALE CIRCUMCISION 48 ALCOHOL AND DRUG USE 49 IMPACT 51 DEMOGRAPHIC 51 MACROECONOMIC 53 HEALTH SECTOR 54 HOUSEHOLDS 54 ORPHANS AND OTHER VULNERABLE CHILDREN 55 EDUCATION 57 AGRICULTURE 57 INDUSTRY 58 RESPONSE 59 GOVERNMENT 59 MULTISECTORAL RESPONSE 59 HUMAN RIGHTS 61 NONHEALTH MINISTRIES 63 ZANZIBAR 64 BUDGETS 64 CIVIL SOCIETY 70 EXTERNAL DONORS 71 CONDOMS 76 VCT 76 PTMCT 79 CARE AND SUPPORT 80 OPPORTUNISTIC INFECTIONS 81 ANTIRETROVIRAL THERAPY 81 HIV PREVENTION TRIALS NETWORK (HPTN) 82 FEMALE-CONTROLLED PREVENTION TECHNOLOGIES 83 MILITARY 83 ASSESSMENT OF GOVERNMENT RESPONSE 83 COMMERCIAL SECTOR RESPONSE 84 REFERENCES 86 Preface The Country AIDS Policy Analysis Project is managed by the AIDS Policy Research Center at the University of California San Francisco. The project is funded by the U.S. Agency for International Development, Cooperative Agreement PHN-A-00-01-00001-00. Stephen F. Morin, PhD, is the project's principal investigator. The project received additional support from the International Training and Education Center on HIV (I-TECH), a collaboration of the University of Washington and UCSF funded through a cooperative agreement with the HIV/AIDS Bureau of the U.S. Health Resources and Services Administration. The views expressed in the outputs of the Country AIDS Policy Analysis Project do not necessarily reflect those of USAID or I-TECH. The Country AIDS Policy Analysis Project is designed to inform planning and prioritizing of effective and equitable HIV/AIDS prevention and treatment interventions through multidisciplinary research on HIV/AIDS. The project evolved from the acute need for analysis of the epidemiology of HIV/AIDS in tandem with analysis of countries' political economy and sociobehavioral context¾at household, sectoral, and macro levels. This multidisciplinary analysis aims to: * help inform national HIV/AIDS policies * strengthen ability to plan, prioritize, and implement effective interventions * highlight the range of sectoral interventions that may affect or be affected by HIV/AIDS * facilitate multisectoral/interministerial coordination * facilitate intercountry information sharing * increase national and subregional capacity for effective partnerships The project develops and disseminates online analyses of HIV/AIDS in 12 USAID priority countries: Ethiopia, Kenya, Malawi, Senegal, South Africa, Uganda, Tanzania, Zambia, Zimbabwe, Brazil, Cambodia, and India Each analysis is linked with national strategic plans for HIV/AIDS prevention, care, and support. Analyses also include a comparative table of 70 key HIV/AIDS and socioeconomic indicators. The primary audience for the country analyses is in-country HIV/AIDS planners, including those from government ministries and agencies, multi- and bilateral donors, international and local NGOs, health care institutions, prevention programs, academia, affected communities, and the private sector. International investigators and policymakers also report using the analyses in their work. All country analyses undergo peer review at the AIDS Research Institute of the University of California San Francisco. In addition, two in-country experts from each profiled country serve as peer reviewers. A scientific advisory board also reviews all analyses. Acknowledgments The following individuals served as peer reviewers and provided valuable inputs to this paper: Dr. Ariane van der Straten, University of California San Francisco; Dr. Jessie Mbwambo, Muhimbili University College of Health Sciences, Dar es Salaam, Tanzania; and Dr. Frank F. Mosha, National Institute for Medical Research, Mwanza, Tanzania. They are not responsible for any errors of fact or judgment. Executive Summary The first AIDS cases in Tanzania were reported in the Kagera region in 1983. By 1987, all regions of the country had reported AIDS cases. HIV sentinel surveillance (HSS) from 2002 found an overall prevalence of 9.6 percent. Comparison with earlier HSS findings and discernment of trends are rendered difficult as prior to the introduction of a standardized HSS system in 2001, sites used a variety of methods to test for HIV and report findings. UNAIDS estimated that 1.6 million Tanzanians were living with HIV/AIDS at the end of 2003; of them, 1.5 million were adults, and adult HIV prevalence was 8.8 percent. Women comprised 56 percent of Tanzanian adults with HIV/AIDS. Although the recent performance of Tanzania's macroeconomy has been impressive, transforming macroperformance into microlevel benefits remains a major challenge. Tanzania remains one of the world's poorest countries, with 2002 GNI of US$290. Despite debt relief under HIPC, the country's debt sustainability prospects are tenuous. Access to health care is constrained by distance to facilities, poor road infrastructure, and lack of vehicles for transportation. Most public dispensaries do not have sufficient funds to provide proper facilities and services, and the poor cannot afford private facilities. The country's high maternal mortality ratio is an indication not only of poor reproductive health, but also of women's low status and poor access to basic health services. Tanzania had the world's 14th-highest burden of TB. Up to 50 percent of TB patients are coinfected with HIV, and TB is the leading cause of death among AIDS patients. Stigma and discrimination remain major barriers to HIV/AIDS prevention and care. Misperceptions about HIV transmission persist. Despite generally widespread knowledge of condoms, urban-rural differentials with regard to ability to obtain condoms are high among both men and women. AIDS is projected to reduce life expectancy by up to 17 percent and population size by up to 15 percent. AIDS has already increased mortality by 11 percent. HIV/AIDS has resulted in welfare losses equivalent to 47.2 percent of GDP. The epidemic continues to impose a heavy burden on the health care system. Households affected by AIDS are experiencing significant reductions in income and increases in health expenditures. The percent of orphans attributed to AIDS rose from 4 percent in 1990 to 42 percent in 2001; it is projected to increase to 54 percent by 2010. In 1985, the government established a national AIDS task force. In 1987, it created the National AIDS Control Program, which subsequently established AIDS coordinators in each of the country's districts. The initial response failed to reverse the trend of the epidemic at national level. Active political commitment began to accelerate in late 1999, when President Mkapa declared HIV/AIDS a national disaster and called for the entire nation¾particularly political, civil, and religious leaders¾to ramp up efforts to combat HIV/AIDS. In 2001, the National Policy on HIV/AIDS was approved. Its overall goal is to provide a framework for leadership and coordination of the national multisectoral response to the epidemic. In 2002, the Tanzania Commission for AIDS was established. The National Multisectoral Strategic Framework on HIV/AIDS 2003-2007 delineates the basic approaches and principles that guide the national response. There are numerous NGOs and CBOs¾including associations of PLWHA¾providing critical prevention and care services. Current capacity to treat OIs is highly constrained, with wide variability in availability of OI drugs. Few Tanzanians who are clinically eligible for ART can afford treatment. Other factors that contribute to low uptake include lack of social support networks and confidentiality issues. The government of Tanzania is strongly committed to addressing HIV/AIDS. It has made significant progress in establishing an institutional framework required to scale up the response. All districts have prepared HIV/AIDS plans. The country has successfully secured major external financing, including World Bank loans and GFATM grants, to scale up HIV/AIDS interventions. Among the country's challenges in addressing HIV/AIDS: * Data from Tanzania's HSS have not been reliable. Therefore, it is difficult to determine the epidemic's dynamics and thus plan adequately for them. * Currently, the HIV/AIDS response does not appear to be targeting prevention interventions at populations at high risk of acquiring HIV. Specific at-risk populations¾such as highly mobile populations, sex workers, truck drivers, fishermen, MSM, and prisoners¾have little or no access to prevention programs in general, much less to programs tailored to their needs. * Although ministerial and district AIDS committees have been established, most of them are not yet effective. * The current lack of structured and comprehensive school-based HIV/AIDS programs is a major constraint. * Much of Tanzania's institutional infrastructure has limited capacity to respond to HIV/AIDS. Constraints include providers' knowledge gaps, limited human and financial resources, commodity shortages, and poor management capacity. Personnel and other human resource limitations are likely to become increasingly acute as the response to the epidemic is scaled up and flows of HIV/AIDS funds into the country increase. * There are no laws to protect the rights of PLWHA. * Despite macroeconomic progress, Tanzania remains one of the world's poorest countries. Its HIV/AIDS activities continue to be heavily reliant on external donors, a scenario that raises concerns regarding sustainability. Epidemiology HIV Sentinel Surveillance The first AIDS cases in Tanzania were reported in the Kagera region in 1983.[1] By 1987, all regions of the country had reported AIDS cases. In that year, the National AIDS Control Program (NACP) was established.[2] In the late 1980s, HIV and syphilis serosurveillance in antenatal clinics (ANCs) was initiated in one region.[1] In 1990, HIV and syphilis sentinel surveillance among ANC attendees was established in 24 sites spanning 11 of mainland Tanzania's regions.[3] Some HIV sentinel surveillance (HSS) has been conducted annually since 1992. There are, however, few clinics for which a consistent time series can be constructed; moreover, in the second half of the 1990s, fewer clinics reported than had done so in earlier years. Kagera, Mwanza, Arusha, Kilimanjaro and Mbeya regions and Dar es Salaam have more extensive information than other regions.[4] No ANC surveillance survey was conducted in 2001, presumably because NACP was in the midst of designing a new HSS system, as discussed below. Collaborating with WHO and CDC[5], NACP examined the strengths and weaknesses of the existing HSS system in 1999. The assessment resulted in revised and improved methods for HIV and syphilis sentinel surveillance. Using these and other resources, NACP revised the protocol for ANC surveillance. New methods resulting from this revision included a three-month data collection period, the introduction of dried blood spot filter paper cards technology, standardization of HIV tests, and quality assurance measures.[1] Methodology NACP follows the WHO recommendations for sampling and HIV testing, and currently collaborates with the U.S. Centers for Disease Control and Prevention (CDC) on HIV testing protocols.[1, 3] The surveillance population comprises pregnant women attending ANCs for the first time for a particular pregnancy at a sentinel site. A blood sample drawn for routine estimation of hemoglobin and syphilis screening is subjected to unlinked anonymous HIV testing using the ELISA test.[3] 2000 Findings A total of 6,505 ANC attendees were recruited from 28 sites. Prevalence ranged from 4.2 percent (95% CI: 1.0-12.7) in Igekemaja (Mwanza) to 32.1 percent (95% CI: 24.9-40.1) in Ipogoro (Iringa). Of the 28 sites, 22 (78.6 percent) reported HIV prevalence over 10 percent. HIV prevalence was highest in the younger age groups (14-24 and 25-34).[3] In comparing 2000 findings with those from earlier HSS, one must bear in mind that NACP's plan to strengthen HSS through a new system of site selection and quality assurance took effect after the 2000 HSS.[3] Sites did not use standardized procedures for blood collection and other surveillance methods, impeding the reliability and quality of data collected.[1] Moreover, many sites created in 1990 had been replaced by new ones by 2000.[3] During the 1990s, prevalence across sites fluctuated greatly. For example, in the Mbeya sites, prevalence rose from 15.4 percent in 1992 to 20.3 percent in 1994, fell back to 15.4 in 1998, and rose to 18.6 in 2000. At the Iringa Reg Hospital site, prevalence rose from 24.9 percent in 1998 to 40.1 percent in 1999, falling to 4.6 percent in 2000.[3] One might infer that the weaknesses in the HSS system accounted for some portion of these fluctuations. Extrapolating the 2000 findings to the Tanzania mainland adult population, 1,810,353 persons ages 15 and above (690,779 males and 1,119,574 females) were living with HIV. Of them, 1,506,703 (561,258 males and 945,445 females) were ages 15 to 49.[3] 2002 Findings As mentioned above, to strengthen HSS, NACP established a new system of site selection from 2001 onward, wherein four sentinel sites in each of six selected regions are chosen. These regions are sampled to represent various geographical areas of the country. The four sites in each region include one urban site, one semiurban or roadside site, and two rural sites (located in a rural health center, dispensary, or independent clinic).[1, 3] Between January and April 2002, a new round of HIV and syphilis serosurveillance was conducted at 24 ANC sites located in six regions: Dar es Salaam, Dodoma, Kagera, Kilimanjaro, Mbeya and Mtwara. Serosurveillance was conducted at urban, semiurban, and rural clinics.[1] A total of 7,275 ANC attendees were sampled. The number of enrollees ranged from 862 in Mtwara to 1,697 in Dar es Salaam. A total of 695 women tested HIV-positive, for an overall HIV prevalence of 9.6 percent (95% CI: 8.9-10.2). HIV prevalence ranged from 5.6 percent in Kagera (95% CI: 4.4-6.7) to 16.0% in Mbeya (95% CI: 13.9-17.8). Across the Dar es Salaam sites, HIV prevalence was 12.8 percent (95% CI: 11.3-14.4).[1] Of the 24 ANC sites participating in the study, eight (33.3 percent) reported HIV prevalence greater than 10 percent: one clinic in Mtwara (urban), one in Dodoma (semiurban-roadside), two in Mbeya (urban and border), and all four urban clinics in Dar es Salaam. Attendees from clinics located in urban areas had higher HIV prevalence than those recruited from rural clinics (p<0.001).[1] In all regions, HIV prevalence was highest among women ages 25 to 34. Single women had higher prevalence than married women (p<0.001). Prevalence among single women who had been pregnant before (18.3 percent) was nearly twice that of married women who had previously been pregnant (9.8 percent) (p<0.0001). (NB: Most HIV-positive attendees in this study had previous pregnancies). Women who reported some or more than primary education had higher prevalence in some regions than women with no education (p<0.0001). Over 80 percent of women coinfected with HIV and syphilis were married, had formal education, and lived in urban areas.[1] Comparison of 2002 findings with previous HSS is difficult, as the 2002 HSS occurred under the new NACP system previously described. Overall HIV prevalence among ANC attendees in 2002 was 9.6 percent, ranging from 5.6 to 16.0 percent across six regions¾dramatically lower than findings from the previous HSS in 2000, in which regional prevalences ranged from 4.0 to 32.1 percent. Note, however, that as discussed above, comparisons are hindered by past use of a variety of nonstandardized methodologies and substandard quality control measures. [1] Zanzibar The first three AIDS cases in Zanzibar were identified in 1986 at Mnazimmoja Hospital. HSS among ANC attendees has been conducted since 1987, with site sample sizes varying by year. In 1998, a standardized HSS protocol was developed.[6] Thus, findings must be viewed in the context of nonstandardized data collection and lack of quality assurance measures. HSS indicated that HIV prevalence among ANC sites rose from 0.3 percent in 1987 to 0.7 percent in 1999, peaking at 3.8 percent in 1995.[6] According to a June 2003 study completed by Tanzania's MOH with participation and support from WHO, UNDP, UNICEF, and Muhimbili University College of Health Sciences, HIV prevalence in Zanzibar was estimated at 0.6 percent for the general population.[7] Transmission Patterns In January 2004, Tanzania's Global Fund Country Coordinating Mechanism reported that heterosexual transmission accounted for 82 percent of HIV infections; transmission from mother to child represented 6 percent of infections.[8] Among AIDS cases reported in 2000, heterosexual transmission was the primary mode of infection (77.2 percent). MTCT accounted for 3.4 percent of AIDS cases; transmission via infected blood/blood products accounted for 0.4 percent. For 19 percent of cases, the transmission mode was not stated.[3] There are no published data on HIV transmission via MSM or IDU. HIV Incidence Given the difficulty in discerning trends in HIV prevalence due to lack of standardization across HSS cycles, making inferences about HIV incidence using prevalence data from the youngest age groups must be approached with caution. A study conducted by researchers from the German Agency for Technical Cooperation (GTZ), Tanzanian Ministry of Health, London School of Hygiene and Tropical Medicine, University of Munich, and Technical University of Berlin analyzed age-specific trends in HIV seroprevalence among ANC attendees in four areas of Mbeya region between 1988 and 2000 and in one area of Rukwa region between 1991 and 1999. (NB: As mentioned above, Kagera and Mbeya are among the regions having the most extensive HSS data.[4]) Indicators measuring behavioral change and attendance rate of patients with STIs were monitored using findings from routine surveillance, complemented by two Knowledge, Attitude, and Practice (KAP) surveys in 1995 and 1999. The researchers found increasing HIV prevalence from 1988 through 1994-95 for women ages 15-24 across all strata. Between 1994-95 and 2000, prevalence declined significantly in all strata for this age group in Mbeya region, with diverse patterns in infection spread along with a significant decrease in positive syphilis serology, high rate of condom use, significant delay in sexual début among primary school pupils, and high treatment rate for STIs. However, the increasing trend in HIV prevalence in Rukwa region continued. The study concluded that declining trends in HIV prevalence among women ages 15-24 may correspond to reduced incidence, partly attributable to behavior change and reduction in a biologic factor influencing HIV transmission¾outcomes to which Mbeya's large-scale HIV prevention activities (implemented since 1988) may have contributed.[9] Reductions in HIV incidence trends would provide the most convincing evidence of a decrease in epidemic size, but large, long-term, longitudinal studies (cohort studies, which can measure both incidence as well as prevalence) are needed to obtain such evidence.[10] Mwanza Cohort Study The Mwanza Cohort Study was a community-randomized trial to investigate the impact of an STI treatment program on HIV incidence (for detail on the trial, see, inter alia, Grosskurth et al. "Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomized controlled trial." Lancet 1995 Aug 26;346(8974):530-6.) In a retrospective study nested within the Mwanza trial, researchers followed a cohort of 1,802 couples for two years, with HIV status of each couple assessed at baseline and follow-up. At baseline, 96.7 percent of couples were concordant-negative, 0.9 percent were concordant-positive, 1.2 percent were discordant with the male partner being HIV-positive, and 1.2 percent were discordant with the female partner being HIV-positive. Individuals living with an HIV-positive partner were more likely to be HIV-positive at baseline (women: OR: 75.7, 95% CI: 33.4-172; men: OR: 62.4, CI: 28.5-137). Seroincidence rates in discordant couples were 10 per 100 person-years (py) and 5 per 100 py for women and men, respectively (RR: 2.0, CI: 0.28-22.1). In concordant-negative couples, seroincidence rates were 0.17 per 100 py in women and 0.45 per 100 py in men (RR: 0.38, CI: 0.12-1.04). Individuals living in discordant couples were at a greatly increased risk of infection compared with individuals in concordant-negative couples (RR: 57.9, CI: 12.0-244 for women; RR: 11.0, CI: 1.2-47.5 for men). It appeared that men were more likely than women to introduce HIV infection in concordant-negative partnerships. In discordant couples, incidence in HIV-negative women was twice as high as in men.[11] Police Officers Cohort Study in Dar es Salaam HIV prevalence among females was 18.0 percent, compared to 13.3 percent among males. The incidence of HIV was 19.6 and 22.4 per 1,000 person years at risk for males and females, respectively.[3] Blood Donors Blood donor screening was introduced on the Tanzanian mainland in 1987. Initially, screening was done at regional and referral hospitals. In 1990, this activity was extended to all hospitals providing blood transfusion services. Donors are screened using either rapid tests in peripheral hospitals or the ELISA test in regional and referral hospitals. Test results are filled in blood donor HIV register forms made available to the hospitals from the MOH through the regional medical office. Copies of completed forms are returned to NACP for processing and reporting.[3] In 2000, a total of 128,595 individuals donated blood on the mainland. Persons under age 15 and those whose information on HIV test results was missing were excluded from analysis, resulting in 128,366 individuals as the basis for analysis. Of them, 107,593 (83.8 percent) were males, 20,619 (16.1 percent) were females, and for 153 individuals, sex was not specified. Overall prevalence among blood donors was 9.9 percent (95% CI: 9.7-10.1). As in past years, prevalence among males was significantly lower than among females: 9.2 percent (95% CI: 9.0-9.4) versus 13.3 percent (95% CI: 12.8-13.7), respectively.[3] In Zanzibar, there are eight hospitals that provide blood transfusion services, three in Unguja and five in Pemba. They screen blood donations and report results to ZACP. A questionnaire is usually administered to potential donors, almost all male, to exclude those with self-reported high-risk behavior, thus reducing reported prevalence among this group. Prevalence increased from 0.5 percent in 1987 to 1.5 percent in 1998. During the first half of 1999, prevalence was 1.4 percent.[6] AIDS Cases AIDS cases diagnosed by hospitals are reported quarterly to NACP. Reporting is done using forms distributed to all hospitals through regional medical officers (RMOs). Information collected includes name of reporting hospital, district of usual residence, sociodemographic characteristics of the diagnosed case, case definition criteria used to make the diagnosis, possible source of infection, and whether an HIV test was conducted.[3] Since 1983 through December 2000, there have been a cumulative total of 130,386 AIDS cases reported to NACP. During 2000, 11,673 AIDS cases were reported to NACP from the Tanzania mainland. Because of underutilization of health services, underdiagnosis, underreporting, and delays in reporting, NACP estimates that only 20 percent of AIDS cases are reported to it.[3] In 2000, among reported AIDS cases, 44.2 percent were married, and 24.2 percent were single. The remaining cases were divorced (6.6 percent), separated (4.2 percent), cohabiting (1.9 percent), or widowed (1.3 percent). For about 12.6 percent of cases, marital status was not stated. Ninety-four percent of cases were above age 15. Among women, AIDS cases peaked in the 25-34 age group; among men, they peaked in the 30-39 age group, suggesting a large portion of HIV transmission from older males to younger females.[3] (See the Age Mixing section below.) According to the 2000 data, the region with the highest cumulative case rate was Mbeya, followed by Dar es Salaam and Coast. The region with the lowest case rate was Mara.[3] However, these data should be viewed with caution, as Mara has a high rate of underutilization of health services compared with other regions.[12] AIDS Mortality According to UNAIDS, during 2003, there were 160,000 AIDS deaths (adults and children) in Tanzania.[13] Health facility-based data compiled in the Tanzania Health Statistics Abstract in 1999 indicated that the leading causes of mortality among those age 5 and above were malaria (22.0 percent), AIDS (17.0 percent), TB (9.0 percent), pneumonia (6.5 percent), and anemia (5.5 percent).[2] See the Impact section for more detailed discussion. U.N. Estimates At the end of 2003, UNAIDS estimated that 1.6 million Tanzanians were living with HIV/AIDS (estimate range: 1.2 million to 2.3 million). Of them, 1.5 million were adults (ages 15-49), and adult HIV prevalence was 8.8 percent. Of adults living with HIV/AIDS, UNAIDS estimated that 840,000 (or 56.0 percent) were women.[13] The U.N. Population Division believes that HIV prevalence in Tanzania peaked in 1997 at 8.8 percent. It projects that prevalence in 2050 will be 1.9 percent. [14] Sexually Transmitted Infections During 2000, a total of 1,974 women attending ANCs for the first time for a particular pregnancy were screened for syphilis. Across sites, prevalence of syphilis ranged from 0.0 to 44.5 percent, compared with 0.4 to 32.6 percent in 1999. Some sites reported declines in prevalence, whereas others reported dramatic increases. As discussed above, however, irregularities in screening and testing procedures likely explain at least part of these fluctuations. Generally, however, data from various sites have indicated a decreasing trend in syphilis prevalence from 1990 to 2000.[3] In 2002, a total of 7,201 ANC attendees were tested for syphilis (under improved surveillance procedures, as discussed above). A total of 590 women tested positive, for an overall syphilis prevalence of 8.2 percent (95% CI: 7.6-8.8). Syphilis prevalence ranged from 3.0 (95% CI: 1.9- 4.1) in Kilimanjaro to 12.3 percent (95% CI: 10.1-14.4) in Dodoma. Across Dar es Salaam sites, prevalence was 4.8 percent (95% CI: 3.8 -5.9).[1] Women living in rural areas had higher syphilis prevalence than those in urban areas (p<0.0001). Marital status did not appear to influence syphilis prevalence. Women ages 25-34 were more likely to have syphilis than women less than 25 years (p=0.015). In contrast to the 2002 HIV sentinel surveillance findings discussed above, women with no education were more likely to be infected with syphilis than were women with some education (p<0.0001).[1] In 2002, overall, 12.4 percent of ANC attendees were coinfected with syphilis and HIV. Of those coinfected, 89 percent lived in an urban area, 60 percent were ages 25-34, 86 percent were married, and 86 percent had some education. The highest proportion of coinfected attendees was observed in Mtwara (24.6 percent).[1] The above STI data refer to the Tanzanian mainland. In Zanzibar, ZACP reports that among STI patients, HIV prevalence has fluctuated from 22.0 percent in 1994 to 31.4 percent in 1996, falling to 25.8 percent in 1997 and rising to 55.5 percent in 1998.[6] Data Quality Issues ANC data currently serve as Tanzania's primary sentinel surveillance of HIV. Limitations of ANC data (see box 1) include that comparative studies have shown that the HIV prevalence among pregnant women in sub-Saharan Africa underestimates prevalence in women of reproductive age because fertility among HIV-positive women is substantially lower than among uninfected women.[15] For example, Gregson et al. have found 25 to 40 percent lower fertility in women with HIV in high-prevalence African countries; they attribute about half of this "subfertility" directly to HIV infection.[16] A team comprising researchers from LSHTM, the Tanzania-Netherlands Project to Support HIV/AIDS Control in Mwanza, and the University of North Carolina at Chapel Hill examined the association between HIV and fertility in Kisesa. (Kisesa is a rural ward in the Mwanza region of northern Tanzania. The ward's population is about 20,000 and is located along the main road to Kenya. The Kisesa cohort study began in 1994.[17]) They used data from a demographic surveillance system in Kisesa during 1994-98 and two large serosurveys of all residents in 1994-95 and 1996-97. HIV-associated fertility reduction among women was investigated by estimating fertility rates by HIV status and HIV prevalence rates by fertility status. A substantial reduction (29 percent) was observed in fertility among HIV-infected women compared with HIV-uninfected women. The fertility reduction was most pronounced during the terminal stages of infection, but no clear association with duration of infection was observed. Use of modern contraception was higher among HIV-infected women. However, among both contracepting and noncontracepting women, a substantial reduction in fertility was seen among HIV-infected women.[18] In a study using several Tanzanian datasets, Ainsworth and her World Bank colleagues found that the death of an adult female household member and the death of a sibling or a husband were associated with lower recent fertility for surviving women. The reasons that surviving women have lower fertility in areas of high adult mortality might include reduced long-run economic benefits of higher demand for women's time; lower income following an adult death; and the need to absorb orphaned children, which might reduce a household's own demand for children.[19] Another issue is reliability of HIV testing, standardization of data collection procedures, and quality assurance. As discussed above, prior to the introduction of a standardized HSS system in 2001, sites used a variety of methods to test for HIV and report findings; quality assurance was often lacking. Thus, comparison of findings and discernment of trends are rendered very difficult. Compounding this scenario is that there has been no national behavioral survey since 1999 (and that itself was an interim study) that could aid in validating prevalence dynamics. Political Economy and Sociobehavioral Context In a paper prepared for the WHO Commission on Macroeconomics and Health, David Bloom and his colleagues note that: "Existing data provide some indication that the relationship between poverty and HIV is growing stronger over time, both between and within continents. But it is not possible to infer causality from these data. That is, it is difficult to tell whether poverty causes AIDS or vice versa¾or whether another variable, such as war, inadequate health, or poor education, explains the relationship....In sum, the link between economic status and AIDS is complex."[20] Håkan Björkman, senior adviser on HIV/AIDS to UNDP's Bureau for Development Policy, states that: "HIV/AIDS is not strictly speaking a "disease of poverty" as it affects people at all income levels. But evidence from some countries at advanced states of the epidemic shows that new HIV infections disproportionately affect poor people, unskilled workers, and those lacking literacy skills¾especially young women in each of these categories. The relationship among poverty, gender, and HIV vulnerability has important policy implications."[21] According to UNFPA: "For Tanzania and most developing countries, the last quarter of the 20th century was characterized by the latest round of a series of sharp and often long-term development shocks. Three related shocks are frequently cited: oil price hikes, debt crises and structural adjustment programs (SAPs). In addition, many countries suffered droughts and/or flooding as well as armed conflicts. The majorities in many affected countries saw their lives deteriorate in every way, while local and international elites enriched themselves ever more. Indeed, disillusionment and despair became widespread in contrast to the period of optimism when many nations won their independence from colonial rule. While no one would argue that these shocks "caused" the AIDS epidemic, they did help create an environment highly conducive to its spread into the general population of an STI, and did so just at the time when HIV was entering into societies around the world."[22] This section analyzes key political economy and sociobehavioral contextual elements to highlight the range of sectoral policies and interventions that may affect or be affected by HIV/AIDS. (In addition to the comparative table of key HIV/AIDS and socioeconomic indicators that accompanies this analysis, readers may also want to consult the 2003 indicators related to Tanzania's progress on achieving the Millennium Development Goals, which are published by UNDP .) The section examines individual as well as community characteristics, as increasingly, risk of HIV infection is recognized as related to individual as well as community variables. Shelah Bloom of the University of North Carolina at Chapel Hill and her colleagues investigated the effect of community characteristics on HIV prevalence and incidence using data from the Kisesa cohort study. Kisesa ward is composed of six villages, divided into 47 subvillages-a subvillage is the smallest administrative unit of the government (kitongoji). The ward includes a trading center (also called Kisesa) located on the main road to Kenya. The population of subvillages ranges from 58 people in the more rural areas to 461 in the trading center. Using subvillages as the unit of analysis, community factors investigated were level of social and economic activity, ratio of bar workers per male population ages 18-59, level of community mobility, and distance to the nearest town. Logistic and Cox regression models were used to estimate community effects, controlling for multiple individual factors. All four community factors had strong effects on risk of HIV. Men who lived in subvillages with the highest level of social and economic activity were about five times more likely to be HIV-positive (OR: 4.71, 95% CI: 2.89 to 6.71) than those in places with low levels of activity; women in these subvillages were twice as likely to be HIV-positive HIV (OR: 1.92, 95% CI: 1.27 to 2.92). After controlling for community effects, the effects of some individual factors on the risk of HIV¾education, male circumcision, type of work, and number of household assets¾changed notably. The association between HIV incidence and community factors was in the expected direction, but did not reach statistical significance (RR = 2.07, p = 0.10). A relatively robust and mobile economy appeared to elevate a community's risk of HIV because the local population comes into contact with a larger sexual network, some of whom may already be infected. Although the researchers found that community risk factors did not displace individual risk factors as an explanation for HIV infection, they concluded that individual risk does rise and fall with community risk.[17] Overview Formed in 1964, the United Republic of Tanzania is a union between Tanganyika and Zanzibar. The country is divided into 26 regions: 21 on the mainland and five on Zanzibar.[1] Tanzania has the largest land area of any country in East Africa, occupying 945,087 km2. It borders eight countries: Kenya and Uganda to the north; Rwanda, Burundi, and Democratic Republic of Congo to the west, and Zambia, Malawi, and Mozambique to the south.[1] Surrounded by many nations in conflict, Tanzania is one of the most politically stable countries in Africa.[23] Population distribution in Tanzania is extremely uneven. Density varies from 1 person per km2 highlands and 134 per km2 on Zanzibar.[24] Seventy-eight percent of the population is rural.[25] Dar es Salaam is the capital and largest city; Dodoma, located in the center of Tanzania, has been designated the new capital, although action to move the capital has stalled. [24] Mwanza is the second-largest city in Tanzania.[17] Islam and Christianity each account for 45 percent of the population, with the remaining 10 percent adhering to indigenous beliefs. The Asian community, including Hindus, Sikhs, Shi'a and Sunni Muslims, and Goans, has declined by 50 percent in the past decade to 50,000 on the mainland and 4,000 on Zanzibar. The African population comprises over 120 ethnic groups, of which the Sukuma, Haya, Nyakyusa, Nyamwezi, and Chaga have over 1 million members.[24] Colonial Era In 1886 and 1890, Anglo-German agreements were negotiated that delineated the British and German spheres of influence in the interior of East Africa and along the coastal strip previously claimed by the Omani sultan of Zanzibar. In 1891, the German Government took over direct administration of the territory from the German East Africa Company and appointed a governor with headquarters at Dar es Salaam. German colonial domination of Tanganyika ended after World War I, when control of most of the territory passed to the U.K. under a League of Nations mandate. After World War II, Tanganyika became a U.N. trust territory under British control.[24] In 1954, Julius K. Nyerere, a school teacher who was then one of only two Tanganyikans educated abroad at the university level, organized a political party, the Tanganyika African National Union (TANU). TANU-supported candidates were victorious in the Legislative Council elections of September 1958 and February 1959. In December 1959, the U.K. agreed to the establishment of internal self-government, with general elections to be held in August 1960. Nyerere was named chief minister of the subsequent government. In May 1961, Tanganyika became autonomous, and Nyerere became prime minister under a new constitution. Full independence was achieved in December 1961. Nyerere was elected president when Tanganyika became a republic within the Commonwealth a year after independence. [24] Zanzibar was an early Persian trading center. The Anglo-German agreement of 1890 made Zanzibar and Pemba a British protectorate. In July 1957, elections for six nongovernment members to the Legislative Council were held. Two parties were formed: the Zanzibar Nationalist Party (ZNP), representing the dominant Arab and "Arabized" minority, and the Afro-Shirazi Party (ASP), led by Abeid Karume and representing the Shirazis and the African majority. The ASP won three of the six elected seats, with the remainder going to independents. Following the election, the ASP split; some of its Shirazi supporters left to form the Zanzibar and Pemba People's Party (ZPPP). In December 1963, Zanzibar became independent from the U.K., as a constitutional monarchy under the sultan. In January 1964, the African majority revolted against the sultan and a new government was formed with the ASP leader, Abeid Karume, as president of Zanzibar and chairman of the Revolutionary Council..[24] (President Karume was assassinated in April 1972.[12]) The Tanganyika union with Zanzibar to form the United Republic of Tanganyika and Zanzibar occurred in April 1964; in October of that year, it was renamed the United Republic of Tanzania. Under the terms of its political union with Tanganyika in April 1964, the Zanzibar government retains considerable local autonomy.[24] Postcolonial Context One of Africa's best-known elder statesmen, Nyerere was also one of the founding members of the Non-Aligned Movement. Under Nyerere, Tanzania was a one-party state, with a socialist model of economic development. In 1977, Nyerere merged TANU with the Zanzibar ruling party, the ASP, to form the CCM as the sole ruling party in both parts of the union. In 1985, President Nyerere was succeeded by Ali Hassan Mwinyi. Nyerere retained his position as chairman of the ruling party for five more years and was influential in Tanzanian politics until his death in October 1999.[24] Under Ali Hassan Mwinyi, Tanzania undertook a number of political and economic reforms. In early 1992, the government decided to adopt multiparty democracy. Legal and constitutional changes led to the registration of 11 political parties. Two parliamentary byelections (won by CCM) in early 1994 were the first multiparty elections in Tanzanian history. Tanzania's current president, Benjamin Mkapa of the ruling CCM, was elected in 1995. In October 2000, Mkapa was reelected, and CCM won 202 of the 232 elected seats in Parliament. Zanzibar President Salmin Amour was elected in single-party elections in 1990. In 1995, he was named the winner of Zanzibar's first multiparty elections, a victory widely deemed to have been tainted by fraud. In the 2000 Zanzibar presidential election, Abeid Amani Karume, the son of former President Abeid Karume, defeated CUF candidate Seif Sharif Hamad. The election was marred by irregularities, and subsequent political violence claimed at least 23 lives in January 2001, mostly on Pemba island. In October 2001, the CCM and the CUF parties signed a reconciliation agreement that called for electoral reforms and establishes a commission of inquiry to investigate the deaths that occurred in January 2001 on Pemba.[24] Tanzania is helping to broker peace talks to end conflict in Burundi and supports the Lusaka agreement concerning the conflict in the DRC. Tanzania is the only country in East Africa that is also is a member of the Southern African Development Community (SADC).[24] Economy Agriculture accounts for 80 percent of the workforce and 48 percent of GDP. Industry accounts for 8.3 percent of GDP and includes textiles, agribusiness, light manufacturing, oil refining, and construction. Natural resources include coal, iron, gold, natural gas, nickel, diamonds, and other gemstones. Main exports are coffee, cotton, tea, sisal, diamonds, cashew nuts, tobacco, flowers, seaweed, fish, and cloves.[24] In the 1990s, Tanzania's export economy changed dramatically. During the first half of the decade, coffee was the most significant export; by the end of the decade, however, it had been replaced by gold and diamonds. The growing importance of mining in the Tanzanian economy brings in significant resources from abroad and provides employment; however, mining also entails separation from family and exposure to new sexual networks, which may facilitate risk of HIV acquisition.[26] (See Mobile and Displaced Populations section below.) Macroeconomic Reform In 1986, the government of Tanzania embarked on an adjustment program to dismantle state economic controls and encourage more active participation of the private sector in the economy. The program included a comprehensive package of policies to reduce the budget deficit, improve monetary control, liberalize trade regime, remove most price controls, ease restrictions on the marketing of food crops, free interest rates, and initiate restructuring of the financial sector.[27] The recent performance of Tanzania's macroeconomy has been impressive. GDP growth rose from 5.7 percent in 2001 to 6.2 percent in 2002, related to relatively strong performances in agriculture, mining, manufacturing, as well as wholesale and retail trade. Inflation declined from 4.5 percent in June 2002 to 4.3 percent in April 2003.[27] Improvements in fiscal performance have enabled the government to steadily increase budget allocations to education, health, water, agriculture, and rural roads. The Household Budget Survey 2000/01 indicated improvements in housing conditions; increased possession of consumer durables; and decreased distance to markets, shops, public transport, etc.[28] However, transforming impressive macroperformance into microlevel benefits remains a major challenge for the government.[28] Tanzania remains one of the world's poorest countries. In 2002, gross national income (GNI) per capita (terminology that has replaced GDP per capita) was US$290. This figure is lower than that for the SSA region (US$450) and for all low-income countries (US$430).[29] A 1997 ILO workshop on gender and structural adjustment in Tanzania found that structural adjustment appeared to have worsened gender disparities in access to employment and economic opportunities, particularly in the formal sector. Although both men and women are affected by the retrenchment and personnel reduction that structural adjustment involves, these processes had a greater impact on women. [30] Luisa Ferreira of the World Bank notes that structural adjustment appears to have benefited many poor households; those near the poverty line benefited the most. However, those with extremely low incomes appear to have become somewhat poorer. She notes how increases in the inequality of income distribution eroded some of the potential for poverty reduction that would have otherwise resulted from growth. Moreover, structural adjustment rewarded those with education, excluding those with little education. [31] Poor pricing and unreliable cash flow to farmers continue to constrain the agricultural sector. Tanzania's industrial sector is one of the smallest in Africa. It has been heavily affected by persistent power shortages caused by low rainfall in the hydroelectric dam catchment area, a condition compounded by years of neglect and poor management at the state-controlled electric company. Foreign exchange shortages and mismanagement continue to deprive factories of much-needed spare parts and have reduced factory capacity to less than 30 percent. Despite Tanzania's past record of political stability, an unattractive investment climate has discouraged foreign investment. Zanzibar's economy is based primarily on the production of cloves, the principal foreign exchange earner. Exports have suffered with the downturn in the clove market.[24] Poverty Using the Household Budget Survey 2000-01, the 2000-01 Integrated Labor Force Survey, and other selected studies finds that 18.7 percent of Tanzanians live below the national food poverty line and 35.7 percent live below the national basic needs poverty line.[28] The overall poverty level in urban areas (particularly in Dar es Salaam) is substantially lower than that in rural areas. Among the total poor population, the proportion of the urban poor is only 13 percent, compared to 87 percent in rural areas.[28] Using international poverty markers, 19.9 percent of the population lived below US$1 a day during 1990-2001, and 59.7 percent lived below US$2 a day during the same period.[32] A team from Tulane investigated the concentration of poverty at the community level in Tanzania and its association with availability and quality of primary health care services, utilization of primary health care services, and health outcomes among both poor and nonpoor households. It found that on average, both poor and nonpoor households living in low poverty concentration areas had better health outcomes and service utilization rates than their counterparts living in high poverty concentration clusters. Consistent with these finding, high poverty concentration areas were found to be farther way from facilities offering primary health care than were low poverty concentration areas. Moreover, the facilities closest to high poverty concentration areas had fewer doctors, medical equipment, and drugs. Although the study did not directly measure quality, the characteristics that differentiate high poverty from low poverty concentration clusters indicated that quality was more important than physical access among the study population. Therefore, the team surmised, one possible way to improve the health status of the population in poor clusters is to provide better quality services from existing health facilities; distance to the health center did not appear to be the critical factor.[33] Debt Tanzania qualified for debt relief under the Heavily Indebted Poor Countries (HIPC) Initiative.[34] However, William Easterly of the Center for Global Development states that, "The IMF and World Bank declaring a country eligible for debt relief is an admission that past loans, including adjustment loans, did not bring enough current account adjustment and export and GDP growth in that country to keep debt ratios within reasonable bounds."[35] Tanzania reached the HIPC completion point in November 2001. It received debt relief equivalent to US$2.026 billion in net present value (NPV) terms.[36] The Bank and IMF project that HIPC assistance, in conjunction with comparable action by other creditors, will allow Tanzania to redirect resources to priority poverty reduction efforts. Part of HIPC debt relief was used to abolish user fees for primary education. As mentioned above, expenditures on health, nontertiary education, basic sanitation, and certain rural development and urban development programs have increased. These social expenditures as a percent of government revenue were projected to increase from 42.8 percent in 1999 to 58.0 percent in 2001. Although this percentage is projected to increase to 84.9 percent by 2003, it is projected to begin declining in 2004, to 82.9 percent in 2004 and 77.9 percent in 2005. And as a percentage of GDP, these expenditures, although rising to 10.6 percent in 2003, will stagnate at that level through 2005.[37] HIPC is not debt cancellation; rather it is a restructuring of debt repayment through provision of grants. Tanzania is required to continue servicing its debt. Moreover, HIPC does not preclude that a country will have to continue to borrow indefinitely. Tanzania, for example, is and will continue to be heavily dependent on donors and foreign creditors.[37] The country also faces a scenario in which some donors predicate their own assistance on continuation of IMF support as a sign of commitment to macroeconomic reform. As the Bank and IMF note: "While Tanzania has achieved commendable progress in implementing macroeconomic and structural reforms in the last seven years, the country still faces a substantial reform agenda and other policy challenges that will require continued Fund engagement. The exit from Fund financial support may also be constrained by the linkage of donor support to a Fund-supported program."[37] Many NGOs have argued that the assumptions underlying HIPC are highly unrealistic. Jubilee Plus, for example, a U.K.-based social justice NGO, notes that HIPC assesses whether a country can afford to pay its debts by looking primarily at its export earnings and often making very optimistic assumptions about them. Countries such as Tanzania are highly vulnerable to external shocks such as changes in the price of and demand for export commodities as well as climatic fluctuations.[38] Indeed, a 2003 study by the World Bank's Operations Evaluation Department found that Tanzania's "long-run debt sustainability prospects are tenuous."[39] In a September 2003 paper, Fedelino and Kudina of the IMF presented a model to examine the impact of fiscal policies on debt sustainability in African HIPC countries. They found that: "The nonsustainability of the four post-HIPCs is quite telling: as these countries have made considerable progress in macroeconomic stabilization, they are now 'allowed' to increase their expenditure levels to address poverty reduction needs. For example, Tanzania is projected to increase its expenditure level by more than 4 percent of GDP, to above 22 percent of GDP, in 2002/03 relative to the previous fiscal year. However, based on our framework, this may result in this country's swinging back into unsustainable debt levels."[40] Governance Corruption According to Transparency International's Corruption Perceptions Index 2003, Tanzania ranked as the 36th-most corrupt country in the world, scoring 2.5 (on a scale of 0 [highly corrupt] to 10 [highly clean], based on perception of the degree of corruption as viewed by businesspeople, academics, and risk analysts).[41] Tanzania's dependence on foreign aid has subjected it to pressure from donors to introduce effective anticorruption bodies. Tanzania now has a national anticorruption strategy and a minister for good governance. However, anticorruption bodies are generally underfunded and lack effective powers.[42] Conflict provides wide opportunities for corruption in the region. In 2002, Tanzania was added to the list of governments involved in covert activity in DRC, with reports of close business ties between Tanzanian officials and army officers in the supply of military equipment to rebel factions in Burundi and Rwanda.[42] Law Enforcement and Judiciary According to a recent World Bank report, the quality of legal and judicial services in Tanzania remains low, as reflected in: * a legal and regulatory framework that is fragmented, excessively bureaucratic, and outdated * inordinate delays in resolving disputes and dispensing justice * limited access to legal services for the majority of citizens * limited public trust in the legal and judicial system * weak management and coordination of legal sector institutions * low competence, morale, and integrity of public sector legal personnel * inadequate numbers of professionally trained legal personnel * constrained administrative independence of the judiciary[43] Population Dynamics Tanzania's miid-2003 population was 35.4 million. The population is projected to increase to 52.0 million by 2025 and to 73.8 million by 2050.[25] According to the U.N. Population Division, Tanzania's population growth rate will fall from 1.93 percent during 2000-05 to 1.81 percent during 2010-15, and to 0.79 percent during 2045-2050.[14] Tanzania has a very young population. In 2000, the median age in the country was 16.8.[14] Forty-five percent of the population is below age 15 (compared to a median 44 percent in sub-Saharan Africa and 33 percent in all developing countries).[44] The percentage of the population ages 15-59 will rise from 50.3 percent in 2000 to 65.5 percent by 2050.[14] The Tanzanian government has noted that although ongoing reforms have resulted in economic growth, such growth does not seem to have made a substantive improvement in enhancing access to productive and quality employment among the majority of Tanzanians, especially youth.[28] Human Development One method of tracking human development is to analyze trends in a country's Human Development Index. The HDI was created by UNDP to measures average achievements in life expectancy at birth; adult literacy and combined primary, secondary, and tertiary gross enrollment ratios; and gross national income (which may be thought of as average income). An HDI of 0.800 or above = high human development; 0.500 - 0.799 = medium human development; less than 0.500 = low human development.[32] In 2001, Tanzania's HDI value was 0.400, ranking it 160 out of 175 countries for which UNDP calculated an HDI. The country's HDI is below that for the sub-Saharan Africa region (0.468) as well as all low-income countries (0.561). More worrying, Tanzania's HDI value, already very low, has been decreasing since 1990, when it stood at 0.408.[32] Although the government's spending on health rose during the 1990s, clearly the enormous impact of AIDS mortality (see Impact section) drastically reduced the life expectancy component of the HDI value. Moreover, as discussed above, despite notable macroeconomic gains, the country's GNI per capita remains very low. A critical indicator of the well-being of children is the under-five mortality rate. Since independence, Tanzania has made great strides in improving child health. In 1960, its under-five mortality rate was 241 per 1,000 live births; by 2002, it had fallen to 165, below the rate for sub-Saharan Africa (174) though higher than for all least-developed countries (158). Of the 193 countries for which UNICEF provided under-five mortality rates, Tanzania had the world's 24th-highest under-five mortality rate. Infant mortality, another key human development indicator, fell from 142 in 1960 to 104 by 2002, again below the rate for the region (106) though somewhat higher than for all least-developed countries (99).[45] Another critical human development indicator is the maternal mortality ratio (MMR), the number of deaths to women per 100,000 live births that result from conditions related to pregnancy, delivery, and related complications. According to the most recent estimates by WHO, UNICEF and UNFPA, Tanzania accounted for 3.97 percent (21,000) of the world's 529,000 maternal deaths in 2000. Tanzania had the world's 6th-highest number of maternal deaths during that year. Its MMR was 1,500 (estimate range: 910-2,200). A Tanzanian woman's lifetime risk of maternal death was 1 in 10.[46] In its State of World Population 2003, UNFPA estimated that Tanzania's MMR was 1,408.[44] Mobile and Displaced Populations Tanzania's mobile and displaced populations include: * refugees and internally displaced persons * people affected by drought, flood, and other natural disasters * migrant workers * military personnel * transport workers * tourism-related workers * bar workers * sex workers * miners * merchants/traders/vendors * orphans (who may be sent to live with relatives residing in other regions of the country) and other vulnerable children (e.g., street children, child laborers) * humanitarian and relief workers * prisoners There is high mobility among urban, rural, and mining areas, within Tanzania as well as the continent. Much of this movement is dominated by men and, in eastern and southern Africa, has been facilitated by a well-developed transport infrastructure. Another factor is the retrenchment associated with structural adjustment, which has often pushed young men out of rural areas searching for income opportunities in urban areas.[30] Migrant labor separates men from their families, places them in close proximity to "high-risk" sexual networks, and may result in their having an increased number of sexual contacts. Concurrently, it may also lead to women's reliance on sex to supplement their incomes while their male partners are away for long periods.[47] Below are data for several of these subpopulations. See also the Sex Work section below. Mining Areas As mentioned above, mining is becoming increasingly important in Tanzania, both in terms of employing large numbers of workers and providing a significant amount of Tanzania's foreign exchange. Recent changes in Tanzanian law allow foreign businesses to invest in Tanzanian mines.[26] Tanzania has a long history of artisanal gold mining. As the country has opened up its largely unexploited gold reserves, "gold rushes," and the influx of money and people they entail, have increased.[48] Researchers from the African Medical and Research Foundation (AMREF) used cross-sectional surveys among adults ages 16-54 from different sectors of communities neighboring two newly opened, large-scale gold mines near Lake Victoria. Mine workers, men, women, and female food and recreational facility workers (FRFWs) from the community were randomly selected for interviews and HIV and STI testing. The study enrolled 207 male Tanzanian mine workers, 206 FRFWs, and 202 other male and 205 female community members. Overall, 42 percent of FRFWs were HIV-positive, compared to 6 percent of male mine workers, and 16 percent and 18 percent of other community men and women, respectively. HIV prevalence in FRFWs was significantly associated with alcohol consumption (AOR: 2.5, 95% CI: 1.1 to 5.5), past or present syphilis (AOR: 2.7, 95% CI: 1.4 to 5.1), and single status (AOR: 3.8, 95% CI: 1.2 to 11.9). Among FRFWs, 24 percent had active syphilis, 9 percent chlamydia, and 4 percent gonorrhea. Overall, 50 percent of FRFWs and 50 percent of community men reported never using condoms during sex; 55 percent of mineworkers and 61 and 20 percent of male and female community members, respectively, reported receiving/giving payment for sex during the previous year.[49] Women Working at Truckstops In 1997, AMREF published results of a study involving 1,330 women (mean age: 27.3 years) recruited from seven highway truckstops Most study participants were local brew sellers (47.2 percent), bar/guest house attendants (27 percent), or petty traders (21.1 percent). Overall HIV prevalence was 50 percent.[50] Refugees Most refugees in Tanzania are from the Great Lakes region.[51] During April 1994, in the wake of genocide, an estimated 300,000 Rwandan refugees settled in two large camps in Ngara district in the northwest.[52] Wars in DRC and Burundi led to citizens of these countries to seek refuge in Tanzania as well.[53] Ascertaining the exact number of refugees in Tanzania is difficult. According to UNDP, at the end of 2001, the country hosted 647,000 refugees.[32] According to the U.N. Standing Committee on Nutrition, at the end of 2002, Tanzania had 519,373 refugees.[51] In April 2003, Refugees International estimated that Tanzania had over 700,000 refugees, 500,000 of whom were concentrated in camps along its western border. The remaining 200,000, inhabitants of the Rukwa region since 1972, are from Burundi and live self-sufficiently, although without permanent citizenship in Tanzania.[53] Refugee camps suffer from serious food shortages, and health and safety problems. The transportation system to the camps, both road and rail, is poor, adding to the cost of delivering aid. Moreover, international support to aid refugees remains insufficient, and Tanzania's refugee policy limits refugees' ability to work and till arable land. Violence in camps is a serious problem, including rape and domestic violence.[53] (Also see the Sexual Violence section below.) Between1994 and 1996, AMREF managed an HIV/AIDS project in the Rwandan refugee camps in Ngara district as well as the surrounding Tanzanian communities. Among the key findings of rapid STI and behavioral surveys undertaken among Rwandan refugees: * About 31 percent of women attending ANCs were infected with trichomoniasis; 27 percent with candida; and 16 percent with bacterial vaginosis. Three percent of women were infected with gonorrhea; no chlamydial infection was detected. * The prevalence of biological urethritis was about 10 percent in men, of whom one-third had a gonorrheal and/or chlamydial infection. * The prevalence of active syphilis was 4 percent among women and 6 percent among men * Men reported risky sexual behavior before the exodus from Rwanda, with 10 percent indicating that they had paid for sex at some time in their lives. There was, however, a marked reduction in sexual activity during the actual period of exodus and establishment of the camps in Tanzania. * Despite a high level of knowledge of HIV transmission and ways to prevent it, only 16 percent of men reported using condoms during casual sex. This finding appeared to be partly related to various myths surrounding the use of condoms. * Although AIDS was generally acknowledged as a serious problem in the community, it was also perceived as a stigmatizing condition.[52] People Affected by Drought Food shortages can spur population dislocation, and large movements of people also entail exposure to new sexual networks and thus may heighten vulnerability to HIV. The World Food Program (WFP) reports that over 40 percent of Tanzanians live in chronic food-deficit regions, where irregular rainfall patterns cause repeated food shortages.[54] In September 2003, WFP estimated that from December 2003 to April 2004, 1.9 million Tanzanians would be facing food shortages because of drought.[55] Child Laborers See also the Gender section below. The 2000-2001 National Labor Force and Child Survey found that 4.7 million children ages 5-17 were engaged in economic activities. Of them, an estimated 1.2 million were engaged in commercial agriculture, mining, sex work, and domestic service.[56] A multiagency government task force coordinates anti-child labor programs. There are public awareness campaigns regarding the dangers of child labor and exploitation. Tanzania is one of three countries participating in an ILO-supported, U.S.-funded pilot program to eliminate the worst forms of child labor. The program brings together government agencies, trade unions, and legal and social welfare organizations to combat child labor, including trafficking.[57] Trafficking Tanzania is a source country for women and children trafficked for sexual exploitation and forced labor. Tanzanian girls are internally trafficked for forced domestic servitude and, to a lesser extent, for sex work in the Middle East, South Africa, and Europe. Tanzania is a destination country for women and children from India, Kenya, Burundi, and DRC who are trafficked for forced agricultural labor and forced sex work. According to the U.S. Department of State, the Government of Tanzania does not fully comply with the minimum standards for the elimination of trafficking, although it is making significant efforts to do so. Tanzanian law criminalizes trafficking for sexual purposes, but the country lacks a comprehensive antitrafficking law that addresses trafficking for the purposes of forced labor. Forced labor is, however, prohibited by the Constitution.[58] Prisons Tanzania's Legal and Human Rights Center notes lack of funding, poor living conditions, and disrespect for human rights in prisons. UNDP also highlights the problems of overcrowding and the lack of facilities for remand prisoners, who in some nonremand facilities comprise 90 percent of the prisoner population.[59] The Bugando Medical Center in Mwanza treats inmates in the Butimba prison. The center examined the extent of smear-positive TB among prisoners, using a retrospective cohort study of 501 prisoners from January 1994 to December 1997. The proportion of smear-positive TB in this study was high¾204 prisoners (40.7 percent). Among them, 25.9 percent were coinfected with HIV. The mean length of imprisonment at the time of diagnosis was 19 months.[60] Tanzania's National Policy on HIV/AIDS states that: "Prison inmates have the right to basic HIV/AIDS information, voluntary counseling and testing, and care, including treatment of STIs."[61] The policy does not mention making condoms available to prisoners. Tanzania's Penal Code criminalizes homosexual acts between men[62]; thus, distribution of condoms in prisons is prohibited. Education See also the accompanying table of indicators as well as the Gender section below. Public spending on education fell during the 1990s, from 3.2 to 2.1 percent of GDP.[32] The abolition of user fees for primary education and subsequent increased enrollment ratios are straining the education system and threatening the quality of services.[28] School attendance ratios are low in Tanzania, with only slightly more than half of primary-school-age children attending primary school. Attendance is higher for children in Zanzibar than for those on the mainland (67 versus 53 percent). It is also higher for urban than for rural children and for older children. Educational attainment is higher in Zanzibar than on the mainland. (Compulsory education in Zanzibar incorporates three years of secondary education.)[63] Health System Tanzania has 4,961 government health facilities and 1,926 facilities owned by NGOs, parastatals, voluntary agencies, and the commercial sector. Approximately 65,000 people are involved in health care delivery, 70 percent of whom are in the public sector. About 64 percent of the recurrent public health budget is spent on human resources.[2] At the primary level, there are dispensaries, health centers, and district hospitals. The secondary level comprises regional hospitals, which provide both basic and specialized services. There are six tertiary hospitals in the country.[2] Access to health care is constrained by distance to facility, poor road infrastructure, and lack of vehicles for transportation.[64] Most public dispensaries do not have sufficient funds to provide proper facilities and services, and the poor cannot afford private facilities.[65] (See also the Tulane study on poverty and health services discussed above.) Schellenberg of the Ifakara Health Research and Development Center and her colleagues found that health care-seeking behavior was lower in poorer than in relatively rich families in a rural community in southern Tanzania that might be assumed to be uniformly poor. Using cluster samples of 2,006 children under age 5 in four rural districts (Kilombero, Morogoro Rural, Rufiji, and Ulanga), they found that 1,026 (52 percent) of 1,968 children reported having been ill in the two weeks prior to the survey. Wealthier families were more likely to bring their sick children to a health facility (p=0.02). Their children were more likely than poorer children to have received antimalarials and antibiotics for pneumonia (p=0.0001 and 0.0048, respectively).[66] In January 2004, Tanzania's Global Fund Country Coordinating Mechanism reported that: "The human resource shortage in the health sector is a major concern in Tanzania. It was identified during the recent Health Sector Review as a key issue for the next year. The Care and Treatment Plan estimates that 10,000 additional Full time equivalents (FTEs), most of the Counselors of various types, will be needed. Part of this shortage has been created by the hiring freeze that was in effect in all sectors as part of the IMF structural adjustment measures. The hiring freeze has been partially relaxed for the health sector, but major recruitment has not yet begun; recent classes of medical graduates are still without jobs. A comprehensive HR study is planned for Q2 of 2004, and will better inform current capacity, while also exploring ways to increase capacity (through hiring and retaining higher numbers, and improving productivity)."[8] Health Status See also the accompanying comparative table of key indicators. During the 1990s, there was an increase in the use of improved water sources in rural areas; in Dar es Salaam, however, the proportion of households using improved water sources fell, and other urban areas reported little change.[28] Since the early 1990s, overall vaccination coverage has declined slightly.[63] Little progress was achieved during the 1990s with regard to improvement of nutrition rates for children. There are significant disparities in undernutrition between rural and urban areas, and between children from poorer and richer households. For example, children from the poorest quintile of households are four times more likely to be severely underweight than children from the richest quintile.[28] During the 1990s, there was no substantial progress in reducing infant and under-five mortality. There are even indications of slight increases in recent years, likely related to the HIV/AIDS epidemic.[28] (See also the MMR data above.) Health facility-based data compiled in the Tanzania Health Statistics Abstract in 1999 indicated that the leading causes of mortality among those age 5 and above were malaria (22.0 percent), AIDS (17.0 percent), TB (9.0 percent), pneumonia (6.5 percent), and anemia (5.5 percent).[2] Reform and Decentralization In the early 1990s, Tanzania launched a process of health sector reform to address capacity and performance problems such as: * shortage of trained personnel * low morale because of a downward spiral in real incomes * heavy political influence in appointments * poor management systems and controls[43] Reforms focus on decentralizing the administration of district hospitals to local councils, shifting resources from curative to preventive care, and creating insurance schemes for employees. Overall, the government hopes to redefine its role as provider of health care to that of "facilitator."[43] In a 1999 survey of district health management teams (DHMTs) in Tanzania, Hutchison of the University of North Carolina at Chapel Hill sought to monitor the extent of the decentralization process and to collect information on decentralization's achievements and limitations to date. His survey found that although decentralization had been ongoing for over a decade, less than half of DHMTs reported that decentralization was under way in their districts and that the actual transfer of administrative and fiscal responsibilities was still limited for the majority of districts. Most districts were heavily reliant on external funding and reported that they had control over only a small proportion of their budgets. For those districts in which decentralization was ongoing, the decentralization process was reported to be associated with improvements in a variety of areas: availability of district funds, coordination with donors, ability to attract and retain staff, and utilization of government health services.[67] In 1995, the MOH introduced a cost-sharing policy for health services. Previously free health care services subsidized by the government now require fee payment. ANC and family planning services are exempt from user fees. Often, no fees are imposed on pregnant women, children under five, or the very poor [7,19]. However, other areas of reproductive health (e.g., treatment of STIs) are not exempt.[28]. According to the U.S.-based Center for Reproductive Rights, this policy has made it difficult for the poor to acquire medical treatment. Many are unable to afford doctors' consultation fees of Tsh500, or more expensive diagnostic tests, which can cost up to Tsh120,000.[65] Expenditures During the 1990s, public expenditures on health rose from 1.6 to 2.8 percent of GDP.[32] During 2000-01, the government spent US$5.88 per capita on health. Total health spending was US$178.6 million, of which 55 percent was provided by external donors.[2] Although the absolute amount of public spending on health care has increased, the percentage of the government budget allocated to health has declined, from 15 percent in 1996-97 to 11 percent in 1999-2000.[2, 32] In 2000, private health expenditure represented 2.5 percent of GDP.[32] Sexual & Reproductive Health UNFPA ranks Tanzania a Category "A" Country, meaning that it is furthest from achieving the goals of the International Conference on Population and Development (ICPD), held in Cairo in 1994, and has low levels of development. Group A countries have the greatest need for external assistance and the lowest capabilities for mobilizing domestic resources to close this gap.[68] According to the U.N. Population Division: "In the late 1990s, the cumulative impact of HIV/AIDS, the influx of Rwandan refugees, the burden of debt servicing and deteriorating socio-economic conditions resulted in a general deterioration of the sexual and reproductive health of women and adolescents. Illegal abortion and maternal mortality as a result of complications from abortion are reportedly on the rise."[69] Antenatal Care and Delivery (The last Tanzania Demographic and Health Survey [TDHS] was undertaken in 1996. [The 2003 DHS is currently in process.]. In 2000, an interim DHS was published, Tanzania Reproductive and Child Health Survey 1999 (TRCHS). The 1999 TRCHS sampled 4,029 women ages 15-49 and 3,542 men ages 15-59.) The 1999 TRCHS found that almost all pregnant women in Tanzania (98 percent) receive antenatal care. Health aides provide 44 percent of ANC, nurses and midwives 43 percent, doctors and medical assistants 6 percent, and birth attendants 1 percent. There is significant variation in ANC between rural and urban areas. Urban women are more likely than rural women to receive ANC from a doctor, nurse, or midwife (76 versus 41 percent). Half of pregnant women in rural areas receive ANC from a less-trained rural medical aide or maternal and child health aide, likely because rural populations receive most of their health care services from dispensaries that are run by MCH aides. On the mainland, a greater percentage of ANC is provided by nurses and midwives than in Zanzibar (44 versus 14 percent); in Zanzibar, three-quarters of antenatal care is provided by health aides.[63] The 1999 TRCHS found 44 percent of births are delivered at some type of health facility, whereas 56 percent are delivered at home. The proportion of births delivered in health facilities has been declining.[63] According to the most recent data from UNFPA, 36 percent of births are attended by a skilled health attendant.[44] Births to younger women, first births, and births to urban women are much more likely than others to take place in a health facility.[63] Maternal Mortality As mentioned above, WHO, UNICEF and UNFPA estimate that Tanzania accounted for 3.97 percent (21,000) of the world's 529,000 maternal deaths in 2000. Tanzania had the world's 6th-highest number of maternal deaths during that year. Its MMR was 1,500 (estimate range: 910-2,200). A Tanzanian woman's lifetime risk of maternal death was 1 in 10.[46] (The 1996 TDHS calculated the MMR using the reported survivorship of sisters. It found that between 1987 and 1996, the MMR was 529 per 100,000 live births.[70]) Abortion Abortion legislation in Tanzania is based on the English Offences Against the Person Act of 1861 and the Infant Life (Preservation) Act of 1929. Under the Revised Penal Code of Tanzania (chapter 16, sections 150-152), the performance of abortions is generally prohibited. An abortion may be performed to save the life of a pregnant woman. In addition, Tanzania, as do a number of Commonwealth countries whose legal systems are based on English common law, follows the holding of the 1938 English Rex v. Bourne decision in determining whether an abortion performed for health reasons is lawful; the decision set a precedent for future abortion cases performed on the grounds of preserving the pregnant woman's physical and mental health. Although abortion is restricted by law, there is overwhelming evidence that it is widely practiced. The government has expressed concern about the high incidence of illegal abortion because of its effect on maternal morbidity and mortality. Studies show that illegal abortion is one of the major causes of maternal mortality. A study conducted in the Southern Highlands in 1983 estimated that 17 percent of maternal deaths were directly associated with abortion. Another study carried out in the Kilimanjaro region suggested that about 21 per cent of maternal deaths were related to abortion. In a study undertaken in 1987 at Muhimbili Medical Center, the teaching hospital in Dar es Salaam, it was determined that in a random sample of 300 women admitted to the hospital for early pregnancy loss, 31 percent had had an induced abortion.[69] Fertility Tanzania's total fertility rate for 2002 was 5.2, somewhat lower than that for the sub-Saharan African region (5.5). (TFR = the average number of children a woman would have assuming that current age-specific birth rates remain constant throughout her childbearing years, usually considered to be ages 15 to 49).[71] UNFPA estimates that the country's TFR for 2000-05 is 5.11.[44] The 1999 TRCHS found that on average, rural women have three more children than their urban counterparts.[63] Contraception According to UNFPA, during the 1990s, contraceptive prevalence among women in a union was 25.0 percent for any methods and 17.0 percent for modern methods. These rates are much lower than those for all developing countries (59 percent and 54 percent, respectively) and slightly higher than those for the East Africa region (21 percent and 16 percent, respectively).[44] The 1999 TRCHS found that the modern methods most commonly used by women were the pill (16 percent), injectables (12 percent), and male condoms (11 percent); traditional methods were withdrawal (11 percent), periodic abstinence (8 percent) and lactational amenorrhea (4 percent). Overall, 17 percent of women have an unmet need for family planning, of which 12 percent is for spacing and 6 percent is for limiting births.[63] Adolescents See also the Population Dynamics section above. Young Tanzanian women bear a large burden of fertility. According to UNFPA, in 2003, there were 120 births per 1,000 women ages 15-19. The comparable figure for East Africa was 117 and for all developing countries, it was 53.[44] Most Tanzanian women become mothers before they reach the age of 20. The median age at first birth has increased slightly from around 18 or 19 for older women to over 19 for women in their early 20s.[63] Sexually Transmitted Infections See also the Sex Work section below for several studies on bar workers and HIV/STIs. Mwanza Cohort Study In 1995, researchers reported that improved syndromic management of STIs in Mwanza had reduced HIV incidence by 38 percent in intervention compared to control communities.[72] (For detail on the Mwanza trial, see, inter alia, Grosskurth et al. "Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomized controlled trial." Lancet 1995 Aug 26;346(8974):530-6.) Two community randomized controlled trials in Uganda (Masaka and Rakai) did not find that STI interventions affected HIV incidence.[73] Various hypotheses have been offered to explain the Mwanza , Masaka, and Rakai findings, including differences in the epidemic's maturity, STI patterns, and effectiveness of the interventions assessed.[74] A study undertaken by Tanzanian and Ugandan researchers sought to determine the extent to which the higher impact of STI treatment on HIV incidence in Mwanza could be explained by baseline differences among the trial populations. The researchers reanalyzed baseline data from the three trial populations comparing demography, sexual risk behavior and HIV/STI epidemiology. Data were compared after age-standardization and adjustments for sample selection where necessary. STI rates were also adjusted for the sensitivities and specificities of the diagnostic techniques used. The researchers found that apart from the effects of AIDS on fertility and mortality (including widowhood) in Uganda, demographic patterns were similar across populations. Higher sexual risk behaviors, including younger age of sexual début, higher numbers of recent partners, and lower frequency of condom use, were apparent in Mwanza compared to Masaka and Rakai. High-titre serological syphilis, gonorrhea, chlamydia infection, and trichomoniasis were all more prevalent in Mwanza, except for chlamydia infection in males. There was little difference between sites in the seroprevalence of herpes simplex virus type-2 (HSV-2). Age patterns in the prevalence of short-duration STI and current risk behaviors were similar across sites, although all-titre serological syphilis was more prevalent among older participants in Rakai and Masaka than in Mwanza. The study team concluded that differences among trial populations included higher reported risk behavior and higher rates of curable STI in Mwanza compared to Rakai and Masaka. The team posited that these differences probably related to previous reductions in risk behavior in Uganda and may explain, at least in part, the contrasting results of these trials.[75] The Mwanza research team sought to quantify the association between prevalent or incident HSV-2 and the incidence of HIV seroconversion among adults in the general population of Mwanza. Participants included 127 cases that seroconverted to HIV during the two-year follow-up period and 636 randomly selected controls that remained HIV-negative. After adjusting for confounding factors, a strong association between HSV-2 infection and HIV seroconversion was observed in men (test for trend: P < 0.001), with AOR of 6.12 (95% CI: 2.52-14.9) in those HSV-2 positive at baseline, and 16.8 (95% CI: 6.06-46.3) in those acquiring HSV-2 infection during follow-up. A weaker association was observed in women (tests for trend: P = 0.14), with AOR of 1.32 (95% CI: 0.62-2.78) and 2.36 (95% CI: 0.81-6.84), respectively. Population attributable fractions of incident HIV infection due to HSV-2 were estimated as 74 percent in men and 22 percent in women.[76] The Mwanza researchers also examined the prevalence of HIV and chlamydia trachomatis (CT) infections among adolescents. They enrolled 9,445 15- to 19-year-olds. HIV prevalence was 0.6 percent (95% CI: 0.4-0.8) in males and 2.4 percent (95% CI: 2.0-2.8) in females, and increased steeply with age (trend: P < 0.006 and P < 0.001, respectively). After adjustment for age, risk of HIV infection was significantly associated with female sex (OR: 4.3), never having been to primary school in males (OR: 2.7), and current symptoms of genital discharge (OR: 2.3) or genital ulcer (OR: 5.3) in females. The prevalence of CT was 1.0 percent (95% CI: 0.8-1.4) in males and 2.4 percent (95% CI: 2.0-2.9) in females. After adjustment for age, CT infection was associated with female sex (OR: 2.4), reported current symptoms of STI (males OR: 2.5, females OR: 1.9), and positive leucocyte esterase test (males OR: 3.1, females OR: 2.6). Eighty-two percent of males and 79 percent of females with CT were asymptomatic. There was no association between CT and HIV infection in either sex. [77] The Mwanza team also the prevalence and incidence of syphilis. Two unmatched case-control studies nested within the cohort provide information on potential risk factors. The prevalence of active syphilis (TPHA positive and RPR positive any titre) was 7.5 percent in men and 9.1 percent in women; among those ages 15-19, prevalence among women was 6.6 percent and in men, 2.0 percent. The incidence of TPHA seroconversion was highest in women ages 15-19 at 3.4 percent annually, and about 2 percent per year at all ages among men. A higher prevalence of syphilis was found in those currently divorced or widowed (men: OR: 1.61, women: OR: 2.78), and those previously divorced or widowed (men: OR: 1.51, women: OR: 1.85). Among men, prevalence was associated with lack of circumcision (OR=1.89), traditional religion (OR: 1.55), and reporting five or more partners during the past year (OR: 1.81), whereas incidence was associated with no primary education (OR: 2.17), farming (OR: 3.85), and a self-perceived high risk of STI (OR: 3.56). In women, prevalence was associated with no primary education (OR: 2.13), early sexual début (OR: 1.59), and a self-perceived high risk of STI (OR: 3.57), whereas incidence was associated with living away from the community (OR: 2.72).[78] Other Studies * A team comprising researchers from the University of Oslo and Kilimanjaro Christian Medical Center and College sought to determine the seroprevalence of HSV-2 and to identify clinical, demographic, and behavioral correlates among women attending primary health care clinics. They used a cross-sectional survey of 382 randomly chosen women ages 15 to 49. They found that seroprevalence of HSV-2 was 39 percent. HSV-2 was associated with antibody to HIV-1 (OR: 2.3 CI: 1.1-4.7), syphilis (OR: 4.7 CI: 1.4-4.7), and genital ulcers (OR: 9.7 CI: 2.5-36.9). Age, sexual début, number of sex partners, and history of spontaneous abortion were found to be significantly associated with HSV-2. Eighty-two percent of the women with genital ulcers were HSV-2-seropositive, whereas syphilis accounted for 6 percent of cases. The researchers posited that HSV-2 may be the most common cause of genital ulcers in this population.[79] * A study undertaken by researchers from the University of Goteborg, Muhimbili University College of Health Sciences, and Dar es Salaam Infectious Diseases Clinic examined genital ulcer disease (GUD) and prevalence of HIV infection in patients with GUD in urban areas of Tanzania. A total of 102 clinical specimens were collected from 52 and 50 patients with GUD in Dar es Salaam and Mbeya, respectively, and from 93 patients with genital discharge in a cross-sectional study. Overall, 9 percent of the 102 patients with GUD were infected with HSV-2. Among HIV-seropositive GUD patients, 71 and 46 percent (P<0.003) were coinfected with HSV-2 in Dar es Salaam and Mbeya, respectively. Women with HSV-2 in Dar es Salaam were significantly more likely to be HIV-infected than men (60 versus 39 percent; P<= 0.006).[80] Tuberculosis According to WHO, in 2002, Tanzania had the world's 14th highest burden of TB in terms of new cases. The TB incidence rate was 363 (all) cases per 100,000 population. An estimated 1.2 percent of new cases were multidrug-resistant.[81] Geographically, all of Tanzania is covered by DOTS. The National Tuberculosis and Leprosy Program has achieved a treatment success rate of 81 percent (2001).[81] However, up to 50 percent of TB patients in Tanzania are coinfected with HIV (WHO put this figure at 34 percent for 2002). This has led to increased death rates among coinfected TB patients, rendering it difficult for the NTLP to reach the WHO cure rate target of 85 percent. TB is the leading cause of death among AIDS patients.[2, 81, 82] Gender UNDP measures gender inequality by using the unweighted average of three component indices: life expectancy, education, and income. Its Gender-related Development Index (GDI) value ranges from 0 (lowest gender equality) to 1 (highest gender equality). UNDP calculated Tanzania's GDI value at 0.396 for 2001, ranking it 130 out of 144 countries for which UNDP calculated a GDI.[32] The country's high MMR (see above) is also an indication not only of poor reproductive health, but also of women's low status and poor access to basic health services. Female-headed Households The 1999 TRCHS found that 23 percent of women ages 15-49 and 36 percent of men ages 15-59 have never married; 66 percent of women and 58 percent of men were currently in unions; and 11 percent of women and 5 percent of men were divorced, separated, or widowed. The survey found that 27 percent of urban and 21.7 percent of rural households are headed by women.[63] van Vuuren of the Dutch institute Afrika-Studiecentrum found that in Tanzania: * For female-headed households (FHH), the value of their harvested crops is less than that for male-headed households (MHH). * FHH own less livestock than MHH. * About 40 percent of MHH have members with paying (cash) jobs; for FHH, this figure is only 20 percent. * More MHH receive gifts from relatives than do FHH (56 vs. 34 percent). However, for FHH, these gifts are of greater importance to household income than they are for MHH. For 24 percent of FHH, kin gifts are the main source of income (the corresponding figure for MHH is 5 percent). * MHH have greater ability to remit money to other households than do FHH.[83] Polygyny According to the 1996 Tanzania DHS, 29 percent of married women and 15 percent of all men are in polygynous unions. The practice of polygyny increases with age among women, from 22 percent among teenagers to 38 percent among those ages 45-49. Overall, older women are more likely to be in polygynous unions than younger women. The proportion of women in polygynous unions is slightly higher on the mainland than in Zanzibar (29 vs. 28 percent), whereas the proportion of men in polygynous unions is higher in Zanzibar than on the mainland (21 vs. 15 percent). The highest level of polygynous unions is found in the Southern Highland zone. Nationally, the proportion of currently married women in a polygynous union decreases from 39 percent among women with no formal education to 22 percent among those who have completed primary education.[70] Education See also the table of indicators attached. There is a strong differential in educational attainment between the sexes, especially as age increases. The 1999 TRCHS found that overall, 40 percent of women in Tanzania have never been to school, compared with 31 percent of men. The proportion with no education increases with age. For example, the proportion of women who have never had any formal schooling increases from 17 percent in the 20-24 age group to 88 percent among those age 65 and older. For men, the proportion increases from 11 percent (age group 15-19) to 66 percent (age group 65 and older). The median number of years of schooling is 1.1 for women and 2.6 for men. Thirty-six percent of women are illiterate, compared with only 22 percent of men.[63] In the past, girls who became pregnant were expelled from school. A1996 law permits pregnant adolescents to continue their education following maternity absences. In practice, however, pregnant adolescents are often still forced out of school. No specific law or policy has been enacted to combat the practice of expelling pregnant adolescents from school, although the Minister of Education has stated that it is not proper for educators to do so. The government has thus far failed to issue a binding pronouncement to schools to stop the practice of expelling pregnant girls.[65] Employment A 1997 ILO workshop found that women's wages continue to be lower than those of men; this is partly due to women's employment in jobs with lower remuneration levels, but is also related to gender discrimination within job categories that employ both men and women.[30] The 1999 TRCHS found that 24 percent of women report being unemployed. The proportion not working is higher among younger women and those residing in urban areas. Most women who work do so on a seasonal basis. Thirty-seven percent of working women are self-employed, 9 percent work for others, and 54 percent work in a family business. Most working women (73 percent) earn cash for their work. Rural working women are more likely to work in a family business, whereas urban women are more likely to work for others or for themselves. Urban women who work are also more likely than rural women to receive cash earnings (88 vs. 68 percent). Women in Zanzibar who work are more likely to be self-employed or to work for an employer and are less likely to work in family businesses than women on the mainland.[63] Seventy-two percent of employed women are involved in agricultural activities, mostly working on their own or family-owned land. Twenty percent of working women are involved in unskilled manual jobs. Only 2 percent of women are doing professional, managerial, or technical jobs.[63] The Tanzania Media Women's Association conducted a survey of 737 housemaids from 14 Regions: Mtwara, Mbeya, Arusha, Kigoma, Kagera and Mwanza, Urban West, Unguja West, Unguja South, Pemba South, Dar es Salaam, Kilimanjaro, Iringa, and Singida. The findings indicated that 60 percent of maids reported having sex with male members of the families employing them; reasons included being promised gifts and being threatened with nonpayment of wages or termination of employment. About 30 percent of respondents reported that their male employees (or their sons or male relatives) had forced them to have sex.[84](NB: These results were reported in the popular press and should therefore be viewed with caution.) Widows The 1999 TRCHS found that 3.2 percent of Tanzanian women are widowed. Property "grabbing" is a phenomenon wherein relatives forcefully take possession of the deceased's household goods, land, livestock, clothes, and other assets. This scenario exacerbates the already precarious economic (and social) situation of widows and their children. It is not known how prevalent this practice is in Tanzania, although the World Bank-financed AIDS project discussed below does single out widows as key project beneficiaries.[85] Among organizations that have been active in addressing women's land rights and inheritance issues are the Tanzania Women's Legal Aid Center, Tanzania Women Lawyers Association, and Environmental, Human Rights Care and Gender Organization. More detail on their activities may be found in: * Richard S. Strickland. To Have and to Hold: Women's Property and Inheritance Rights in the Context of HIV/AIDS in Sub-Saharan Africa. Washington, DC: International Center for Research on Women and Global Coalition on Women and AIDS, 2004 * FAO and Oxfam. Report Of The FAO/Oxfam GB Workshop On Women's Land Rights In Southern And Eastern Africa, held in Pretoria, June 17-19, 2003. Pretoria: Southern African Regional Poverty Network, October 2003 Female Genital Mutilation According to the 1996 TDHS, 18 percent of Tanzanian women were circumcised. Younger women (ages 15-19), women living in Zanzibar, and those living in urban areas on the mainland were less likely to be circumcised than other women. A higher proportion of circumcised women lived in the Arusha (81 percent), Dodoma (68 percent), and Mara (44 percent) regions. Twenty to forty percent of circumcised women were found in the Kilimanjaro (37 percent), Iringa (27 percent), Singida and Tanga (25 percent), and Morogoro (20 percent) regions. In the rest of the regions, less than 5 percent of women were circumcised.[70] (The Sexual Offences Special Provisions Act of 1998 outlaws the act of female genital mutilation, terming the offense "cruelty to children." The act stipulates that "any person who, having the custody, charge or care of any person under eighteen years of age...causes female genital mutilation" is subject to five to 15 years imprisonment. Offenders may also be liable for a fine of up to Tsh300,000 as well as compensation for injuries caused.[65]) Sexual Violence The Sexual Offences Act of 1998 criminalizes rape and sexual assault against minors. The act considers rape to include sexual intercourse with a girl or woman "with or without her consent when she is under eighteen years of age." If the perpetrator is married to the victim and she is over the age of 15, the act is not applicable.[65] According to Tanzania's report on the implementation of the Beijing Platform for Action, the government acknowledges that despite severe punishment for offenders instituted with the passing of the 1998 act, high incidence of reported violence against women persists.[65] In 1999, researchers from Muhimbili University College of Health Sciences in Dar es Salaam and Johns Hopkins University investigated the attitudes and experiences related to partner violence and HIV serostatus disclosure of women who seek VCT at the Muhimbili Health Information Center (MHIC), a VCT clinic in Dar es Salaam. The study first collected qualitative data from women, men, and couples (n=67) who were MHIC clients. In the second phase, researchers enrolled 340 women after pretest counseling and prior to collection of test results; 245 of women were followed and interviewed three months after enrollment and testing. Nearly one-third of the sample was HIV-positive, almost half were married, and 50 percent were between the ages of 18 and 29 and had less than seven years of education.[86] Among key findings: * The major reason for nondisclosure of HIV test results among all women, regardless of HIV serostatus, was fear of a partner's reaction, principally fear of abuse or abandonment. * Over 25 percent of women interviewed agreed with the statement, "Violence is a major problem in my life." Male and female informants described violence as a way to "correct" or "educate" women, and reported that violence that does not leave a physical mark on a woman is justifiable. * When asked about lifetime violence by an intimate partner, 38.5 percent of women reported at least one partner who had been physically abusive and 16.7 percent had had at least one partner who had been sexually abusive. Physical violence by a current partner was also commonly reported. * Nearly one-third of women had experienced at least one physically violent episode perpetrated by a current partner, such as slapping, twisting an arm, grabbing, punching, and kicking, in the three months period prior to testing. * Women's HIV status was strongly associated with partner violence. Without adjusting for other variables, HIV-positive women were 2.68 times more likely than HIV-negative women to have experienced a violent episode by a current partner. Examining the interaction between women's age and HIV status and controlling for other sociodemographic variables, young HIV-positive women (ages 18-29) were ten times more likely to report partner violence than young HIV-negative women. * Most women reported that partners showed support and understanding when told test results. However, the proportion of women who reported this positive reaction was significantly greater among HIV-negative women compared to HIV-positive women. * Many women lack autonomy with regard to decisionmaking about HIV testing. Male and female informants frequently referred to the need for women to "seek permission" from partners prior to testing. Men, however, generally made the decision to test on their own without soliciting prior consent.[86] Refugees Women in refugee camps face particular risks. They are vulnerable to rape, sexual assault, and other forms of sexual violence. Levels of domestic violence are also high in many refugee communities, as pressures regarding housing, food, security, and other resources often strain domestic situations and erupt in violence. Extended networks of family, neighbors, and community leaders that may have acted as a deterrent to abuse under normal circumstances no longer exist in refugee camps. Generally, female refugees have limited¾or no¾legal remedies against sexual and domestic violence.[87] Human Rights Watch began monitoring the situation of Burundian refugees in Tanzanian camps in 1997. They have found that Burundian women refugees in the Tanzanian camps were subject to high levels of sexual and domestic violence. They were vulnerable to rape by both male refugees and local Tanzanian nationals. [87] In May 1999, Refugees International estimated that one in four Burundian refugee women in Tanzania had been the victim of rape or serious sexual harassment. In 1998, the International Rescue Committee documented 122 cases of rape and 613 cases of domestic violence in four camps: Kanembwa, Mkugwa, Mtendeli, and Nduta; in 1999, these figures were 111 and 764, respectively. Some of the rape cases had been referred to police for investigation.[87] In 1999, UNHCR began to address violence against women in these four camps more systematically.[87] Stigma and Discrimination In January 2004, Tanzania's Global Fund Country Coordinating Mechanism reported that: "Stigma and discrimination are still major barriers to prevention and care in Tanzania. Often stigma is internalized so that people do not seek diagnostic or treatment services, nor the means to protect themselves. Stigma also is continuing to limit or crush the circulation of information about the epidemic, about options for care, and for communication within couples about risks. Fear of discrimination deters workers from seeking VCT even when it is encouraged by employers, similarly fear of discrimination has greatly limited disclosure and witnessing of PLHAs as compared to Uganda or Zimbabwe."[8] Among women who participated in the 1999 TRCHS, 63 percent replied that they knew someone with HIV/AIDS or someone who had died of AIDS (an increase from the 1996 figure of 48 percent); for men, these figures were 68 and 52 percent, respectively.[63] A part of a three-year study on HIV/AIDS-related stigma in Tanzania, Ethiopia, and Zambia conducted from April 2001 to September 2003, the International Center for Research on Women collected data in Kimara and Bunju wards, Dar es Salaam, and as well as at a stand-alone VCT center in Dar es Salaam. Structured text analysis of 730 qualitative transcripts (650 interviews and 80 focus group discussions) and quantitative analysis of 400 survey respondents from rural and urban areas in the three countries found that: * The main causes of stigma involve incomplete knowledge, fears of death and disease, sexual norms, and a lack of recognition of stigma. The combination of insufficient and inaccurate knowledge and fears of death and disease perpetuate beliefs in casual transmission and thus avoidance of those with HIV. The knowledge that HIV can be transmitted sexually and an association of HIV with socially "improper" sex lead to scenarios in which PLWHA are stigmatized for their perceived immoral behavior. Moreover, people often do not realize that their words or actions are stigmatizing. * Socioeconomic status, age, and gender all influence the experience of stigma. Although the poor are blamed less for their infection than the rich, they face greater stigma because they have fewer resources with which to hide HIV-positive status. Youth are often blamed for spreading HIV through their perceived highly risky sexual behavior. Although both men and women are stigmatized for breaking sexual norms, women are blamed more easily. Concurrently, the consequences of HIV infection, disclosure, stigma, and the burden of care are higher for women than for men. * Many PLWHA face physical and social isolation from family, friends, and community; gossip, name-calling, and voyeurism; and a loss of rights, decisionmaking power, and access to resources and livelihoods. PLWHA internalize these experiences and consequently feel guilty, ashamed, and inferior. As a result, they may isolate themselves and lose hope. Those associated with PLWHA¾particularly family members, friends, and caregivers¾face many similar experiences in the form of secondary stigma. * PLWHA and their families develop various strategies to cope with stigma. Decisions around disclosure depend on whether disclosing would help to cope (through care) or render the situation worse (through added stigma). Some cope by participating in PLWHA networks and actively working in the field of HIV or by confronting stigma in their communities. * Stigma hinders various programmatic efforts. Although testing, disclosure, prevention, and care & support for PLWHA are advocated, they are all often impeded by stigma. Available care & support are often accompanied by judgmental attitudes and isolating behavior, which can lead to delaying care. * The study also found numerous positive aspects regarding how people deal with HIV and stigma. Many people recognize that their limited knowledge has a role in perpetuating stigma and are eager to learn more about HIV/AIDS. Many families, religious organizations, and communities provide care, empathy, and support for PLWHA. Finally, some PLWHA overcome the stigma they face to challenge stigmatizing social norms.[88] Other studies have found that TB patients face high stigma, as they are often automatically labelled as also having AIDS.[2, 82] Awareness and Knowledge of HIV/AIDS As mentioned above, the last Tanzania DHS was undertaken in 1996. The 2003 DHS is currently in process. According to Tanzania Reproductive and Child Health Survey 1999, 97 percent of women and 99 percent of men had heard of HIV/AIDS. Ninety-four percent of women and 96 percent of men believed that AIDS can be avoided. (In the 19996 TDHS, these figures were 88 percent and 90 percent, respectively.) Among women, 98 percent of urban residents, 92 percent of rural residents, and 96 percent of those residing in Zanzibar replied that AIDS can be avoided; among men, these figures were 98, 96, and 97 percent, respectively.[63] Using findings from the 1996 TDHS and 1999 TRCHS indicates that knowledge about ways to avoid HIV/AIDS increased significantly. For example, in 1996, only 39 percent of women spontaneously mentioned condom use as a means of HIV prevention; in 1999, this figure was 56 percent. Among men, 71 percent spontaneously mentioned condoms in 1999, up from 55 percent in 1996. Almost half of women (47 percent) and men (48 percent) mentioned that having only one sexual partner helps to prevent contracting HIV. Other means of prevention spontaneously cited by sizeable proportions of both women and men were abstinence (28 percent of women; 31 percent of men), limiting the number of partners (19 percent; 17 percent), and avoiding injections (11 percent; 10 percent). Notable differentials in knowledge of prevention methods included urban versus rural residence: 73 percent of urban women mentioned condoms, compared with 49 percent of rural women. Among men, these figures were 82 and 67 percent, respectively. Fifty-six of women on the mainland mentioned condoms, versus 32 percent in Zanzibar; among men, these figures were 72 and 44 percent, respectively. For women with no formal education, 32 percent mentioned condoms; 68 percent of those who had completed primary education did so; the comparable figures for men were 49 and 80 percent, respectively.[63] With regard to prompted questions on the three main ways to avoid HIV/AIDS, 49 percent of women interviewed in the 1999 TRCHS knew all three methods, 81 percent knew at least one method, and 17 percent knew of no method. Seventy-one percent of women had correct knowledge of abstinence, 68 percent of having only one partner, and 66 percent of condoms. (Comparable data for men were not included in the 1999 TRCHS report.)[63] In the 1999 TRCHS, 79 percent of women knew that HIV can be transmitted from mother to child (90 percent of urban women vs. 75 percent of rural women); 61 percent knew that MTCT can occur at delivery, 70 percent through breastmilk, and 74 percent during pregnancy. Fifty-five percent of women knew all three transmission modes. Among men, 81 percent knew that HIV can be transmitted from mother to child (88 percent of urban men vs. 78 percent of rural men). (More detailed data on men's knowledge of MTCT were not included in the 1999 TRCHS report.)[63] With regard to personal risk assessment, 52 percent of women and 63 percent of men believed that they had no or a small risk of contracting HIV. Among women, reasons for this perception included has only one partner (64 percent), abstains from sex (23 percent), partner has no other partner (12 percent), and uses condoms (6 percent); for men, these figures were 60, 19, 0, and 22 percent, respectively. Twenty-six percent of women and 23 percent of men believed that they had a moderate to great risk of contracting HIV. Among women, reasons for this belief included partner has other partner (55 percent), doesn't use condoms (34 percent), and has multiple sex partners (percent); for men these figures were 25, 43, and 34 percent, respectively.[63] Misconceptions Among women, 59 percent knew that HIV cannot be transmitted by sharing food (73 percent of urban women vs. 51 percent of rural women), 54 percent knew that it cannot be transmitted by mosquito bites (71 vs. 48 percent), and 69 percent knew that it is possible for a healthy-looking person to be infected with HIV (86 vs. 63 percent). Differentials by level of education were high: 39 percent of women with no formal education knew that HIV cannot be transmitted by sharing food, versus 69 percent of women who had completed primary school. The comparable figures for transmission via mosquito bites were 34 vs. 65 percent, and for knowledge on whether a healthy-looking person to be infected with HIV, 48 vs. 80 percent. (Comparable data for men were not included in the 1999 TRCHS report.)[63] Sexual Behavior Age at First Intercourse According to the 1999 TRCHS, median age at first intercourse for women was just under 17 years. By age 15, about 20 percent of women had had sexual intercourse and by age 18-the legal age of marriage-68 percent of women had had sexual intercourse, whereas 46 percent of them had married. By age 20, 85 percent of women had had sexual intercourse, with 67 percent married by that age.[63] The median age at first intercourse among men was about 18. On average, men entered into marriage six years later than women, but they started sexual relations only about one year later than women. Although the median age at first intercourse had increased slightly from 16.2 years among women ages 45-49 to 16.8 years among those ages 25-29, that of men declined from 18.1 years among those ages 55-59 to 17.5 years among those ages 20-24 years.[63] Irrespective of age, the median age at first intercourse among urban women was slightly higher than that of rural women. There was no apparent urban-rural differential among men. Although the median age at first sexual intercourse was slightly higher among women in Zanzibar than for those on the mainland, the difference was much larger among men; men in Zanzibar initiated sex about three years later than men on the mainland.[63] Age at First Marriage According to the 1999 TRCHS, overall, almost half of women married before age 18 and two-thirds married before age 20. Although the median age at first marriage appeared to have risen from 17 among women ages 45-49 to about 19 among women ages 20-24, much of this increase could be because of recall error on the part of older respondents. Compared with the 1991-92 and 1996 TDHS results, the median age at first marriage for women remained almost the same at slightly over 18 years.[63] Men married considerably later than women. The median age at first marriage for men ages 25- 59 was 24, almost six years later than the median of 18 for women. Only 19 percent of men ages 25-59 were married by age 20, compared with 69 percent of women ages 25-49. Compared with the 1996 TDHS results, the median age at first marriage for men had declined by one year, from 25 to 24.[63] Recent Sexual Activity In the four weeks prior to the 1999 TRCHS, 59 percent of women ages 15-49 years were sexually active, 12 percent were practicing postpartum abstinence, 17 percent were abstaining from sex for reasons other than having recently given birth, and 12 percent had never had sexual intercourse. The proportion of women who were sexually active varied little by age, except that those in the youngest age group were far less likely to be sexually active. The proportion of women who were sexually active was higher on the mainland than in Zanzibar. Two-thirds of men interviewed were sexually active in the four weeks prior to the TRCHS. As with women, men in Zanzibar were less likely than men on the mainland to have had sex in the four weeks before the survey.[63] Multiple Partners The 1999 TRCHS found that that among married women, 7 percent had had two or more partners in the year prior to the survey, including their husbands. Among unmarried women, 11 percent had had two or more partners, whereas almost half were not sexually active at all in the prior 12 months. Men reported having more sexual partners than women; 29 percent of married men and 25 percent of unmarried men reported having had two or more partners in the 12 months before the survey. The proportion of men with two or more sexual partners was higher in rural areas than in urban areas; it was also higher among men on the mainland than in Zanzibar.[63] The 1999 TRCHS notes that: "Although it appears as if there has been an increase since 1996 in the percentage of both women and men who have had two or more partners during the year before the survey, the differences could be due to a change in the line of questioning about sexual behavior. In the 1996 TDHS, married respondents were first asked about the last time they had had sex with their spouse. They were then asked if they had had any sexual partner other than their spouse, a particularly sensitive question. In the 1999 TRCHS, all respondents were asked about the last time they had had sex, the type of relationship they had with that partner (spouse, girlfriend/boyfriend, casual acquaintance, etc.), and then whether they had had sex with anyone else in the previous 12 months. It is likely that this 'softer' series of questions that did not directly inquire about extramarital relationships elicited more honest reporting of the number of partners."[63] Given that in population-based surveys on sexual behavior, men consistently report higher numbers of sexual partners than women¾which may be associated with male exaggeration or female underreporting or with issues related to sampling, such as exclusion of female sex workers¾a study involving researchers from Tanzania's National Institute for Medical Research, the University of North Carolina at Chapel Hill, and LSHTM analyzed data collected in the context of the Mwanza cohort study. The study design included all men and women of reproductive ages and did not involve sampling. The team used these data to compare the consistency of aggregate measures of sexual behavior between men and women living in the same villages. Analysis indicated that nonmarital partnerships were common among single men and women; about 70 percent of unmarried men and women reported at least one sexual partner in the last year. However, 40 percent of married men also reported having nonmarital partners, but only 3 percent of married women did so. Single women reported about half as many multiple partnerships in the last year as men. Underreporting of nonmarital partnerships was much more common among single women than among married women and men. Furthermore, women were more likely to report longer duration partnerships and partnerships with urban men or more educated men than with others. When a woman reported multiple partners, biologic data indicated that she was at high risk of contracting HIV. For men, however, there was a weak association between number of partnerships and risk of HIV, a finding that led the research team to posit that men¾especially single men¾may exaggerate their number of sexual partners.[89] See also the AMREF study on gold mining communities discussed above.[49] Condom Use Overall, the 1999 TRCHS found widespread knowledge of condoms: 92 percent of women and 96 percent of men who had ever had sex knew about condoms. However, urban-rural differentials with regard to ability to obtain condoms were high: among women who had ever had sex and had heard of condoms, 51 percent in rural areas did not know a source for obtaining condoms, compared with 21 percent in urban areas. The comparable figures for men were 28 and 9 percent, respectively.[63] Of respondents who were sexually active during the year prior to the survey, only 16 percent of women and 37 percent of men had ever used condoms, primarily for family planning rather than disease prevention.[63] Eight percent of women and 16 percent of men reported that they had used a condom at last sex. Figures were considerably higher for sexual relations outside marriage: Whereas only 4 percent of women and 5 percent of men reported that they had used condoms at last sex with their spouse, these figures rose to 22-24 percent and 34 percent, respectively, for condom use at last sex with either a regular partner (boyfriend or girlfriend) or someone else.[63] Since 1996, women have reported increasing use of condoms. In 1996, only 5 percent of women reported condom use at last sex; in 1999, this figure had increased to 8 percent. The proportion of women who reported condom use at last sex with someone other than their husband increased from 17 percent in 1996 to about 23 percent in 1999. The proportion of men who reported condom use at last sex had not increased since 1996.[63] In 1999, 49 percent women believed that in general, it was all right for a woman to ask a man to use a condom; 55 percent of women believed that it was all right for a wife to ask her husband to use a condom or to refuse to have sex with him if he had an STI. The level of acceptance in both scenarios was high among those who were formerly in a union, better educated, and living in urban areas.[63] In general, 56 percent of men believed that it was acceptable for a woman to ask a man to use a condom; 58 percent believed that it was all right for a wife to ask her husband to use a condom or to refuse to have sex with him if he had an STI.[63] Researchers from the University of North Carolina at Chapel Hill examined the relationships between individual-, household-, and community-level variables and condom use to prevent HIV infection in women and men in Tanzania (and Uganda) using multilevel modeling. Analyzing data from the 1996 TDHS (and 1995 UDHS), as well as data collected by the MEASURE Evaluation Project at the Carolina Population Center for Tanzania (1996 and 1999), the study found heterogeneity in condom use among different clusters for both women and men; women and men living in clusters with higher indicators of development were more likely to use condoms to prevent HIV infection. In addition, condom use was much more prevalent in areas where health care services were nearby (0-5 km). Condom use was more common among women (but not men) who lived in clusters where HIV/AIDS testing, counseling, and treatment were provided. The researchers found that although education improved condom use, the effect of education was considerably reduced in models that included HIV/AIDS knowledge and cluster-level variables. The positive effect of household wealth on condom use diminished after controlling for the effects of the knowledge and cluster-level factors. Knowledge about HIV and perceiving oneself to be at risk for contracting HIV improved condom use.[90] Age Mixing Using data from the Kisesa (Mwanza region) cohort study, the University of North Carolina at Chapel Hill found that overall, men were about seven years older than their wives, and about five to six years older than their nonmarital sexual partners. Girls under 20 reported sexual partners who were on average 5.3 years older than themselves; 9 percent had a nonmarital partner age 30 and over, and 20 percent had a spouse age 30 and older. Older men (30 and over) tended to have nonmarital partners who were younger than their wives.[91] A study on cross-generational and transactional sex in sub-Saharan Africa undertaken by the International Center for Research on Women and Population Services International, highlighted several studies from Tanzania: * Two studies found that between 27 and 40 percent of girls' (ages 15-19) most recent sexual partners were over 25 years old. Two additional studies reported that the median age difference between girls and young women and their most recent sexual partners was approximately six years. * Two projects conducted in-depth interviews with all girls who presented for induced abortions at urban hospitals in Dar es Salaam. One project reported that 73 percent of girls' (ages 15-19) partners (most often the partner by whom they had became pregnant) were over age 30, and 28 percent were over age 40. In the other study, 31 percent of girls' (ages 14-17) partners (who were responsible for the pregnancy) were over age 45.[92] Transactional Sex The same ICRW-PSI study highlighted that a study in rural Tanzania found that 52 percent of female primary school students and 10 percent of female secondary school students reported the reason for having sex was for money or presents.[92] A study by the National Institute for Medical Research in Mwanza examined premarital sexual behavior among out-of-school adolescents residing in rural communities from farming and lakeshore settings in Magu district. The study found that large numbers of out-of-school adolescents were sexually experienced, that the period from acquaintance or dating to sexual relations was typically short, and that sexual encounters were typically risky. The exchange of money and gifts for sex was reported by both female and male adolescents, although related perceptions and interpretations differed widely. Males perceived that females engage in sex largely for material gain, whereas females view sex as a display of a partner's love or commitment.[93] Sex Work See also the STI section above. A study (the findings of which were published in 1991) conducted by the Kilimanjaro Christian Medical Center and College in Moshi among 106 sex workers (ages 17-70) in Arusha and Moshi found that HIV prevalence was 73 percent. Fifty-one percent of women had evidence of gonorrhea infection, 74 percent had a positive TPHA, and 27 percent had a positive RPR. Of 47 women tested, 25 percent had chlamydia. No significant statistical association was found between the presence of any of the STIs investigated and HIV seropositivity.[94] There have been several studies among bar and hotel workers in Tanzania. Researchers from the Royal Tropical Institute in Amsterdam undertook a study among female bar workers in Magu district in the northwest; they published their findings in 1997. The study team highlights that: "Although the women are still partially dependent on the financial support provided by sexual partners and sexual relations tend to be based on exchange, bar workers cannot simply be equated with prostitutes. Some have a regular partner and the odd casual partner while others may have large numbers of casual contacts. Regular partners are almost always married and often itinerant."[95] In 1993, the team interviewed or enrolled in focus group discussions 33 of the approximately 80 bar, brew shop, and guest house workers in Magu town. In 1994, follow-up interviews were conducted with 27 of the women. The average age of study respondents was 25 years. Bar workers in Magu are predominantly divorced or unmarried and have a high degree of geographic mobility. They select bar work as an alternative to financial dependence on their families. However, because bar workers earn an average of only Tsh 3,000 a month (versus the national average monthly income of Tsh 10,000), they remain partially dependent on the financial support of sexual partners. The difference between regular and casual partners was expressed by respondents primarily in financial terms. Regular partners, who are usually married, support a woman over an extended period and provide assistance in times of acute need (e.g., illness, school fees), whereas casual partners exchange a predetermined amount of money for a single sexual encounter. The distinction between regular and casual partners is based on the nature and extent of financial support; it is also related to condom use and therefore to risk. Women report being able to insist on condom use with casual partners but not with regular partners. Although women report that their regular partners "can be trusted," they do admit feeling at risk of HIV and STIs with these partners..[95] Researchers from Harvard School of Public Health and Kilimanjaro Christian Medical Center and College conducted a study to determine the prevalence and risk factors for HIV infection among women (n = 312) working in bars and hotels in Moshi. They assumed that some percent of study participants were also engaged in sex work with males whom they met in the course of their bar and hotel jobs. Their findings, published in 2002, indicated that HIV prevalence was 26.3 percent (95% CI: 21.4-31.2). In multivariate analyses, the risk of HIV increased with increasing age (p value, test for linear trend <.001) and the number of sexual partners during the last five years (p value, test for linear trend <.03). Another significant predictor was consuming alcohol more than two days each week (AOR: 2.56; 95% CI: 1.12-5.88). The risk of HIV also significantly increased in women with bacterial vaginosis (AOR: 2.37; 95% CI: 1.09-5.13) and in those with HSV-2 (AOR, 2.48; 95% CI: 1.24-4.98).[96] The Harvard-Kilimanjaro Christian Medical Center and College team interviewed bar and hotel workers (n = 519) to obtain information about potential predictors of condom use (consistent condom use was defined as always using condoms with sexual partners in the past five years). The findings, published in October 2003, indicated that overall, consistent condom use among those interviewed was 14.1 percent. In multivariate analyses, consistent condom use was inversely associated with low condom self-efficacy (AOR: 0.20; 95% CI: 0.06-0.71), low condom knowledge (AOR: 0.11; CI: 0.01-0.80), and having more than three children (AOR: 0.23; 95% CI: 0.09-0.54). Other significant predictors included perceived condom acceptability and using condoms at last exchange of sex for money or gift.[97] A study undertaken by researchers from LSHTM, Mbeya Regional Medical, Mbeya Consultant Hospital, Ludwig-Maximilians-University in Munich, and St. George's Hospital in London. determined baseline prevalence of STIs and other reproductive tract infections (RTIs) and their association with HIV as well as sociodemographic and behavioral characteristics in a newly recruited cohort o