HIV Counselling and Testing of Pregnant Women Attending Antenatal Clinics in Botswana, 2001

J HEALTH POPUL NUTR 2005 Mar;23(1):58 _ 65 2005 ICDDR,B: Centre for Health and Population Research ISSN 1606-0997 $ 5.00+0.20 HIV Counselling and Testing of Pregnant Women Attending Antenatal Clinics in Botswana, 2001 Serai Daniel Rakgoasi Department of Population Studies, University of Botswana, Botswana ABSTRACT This study examined the extent to which women accessing antenatal-care services in Botswana were offered HIV-related information and counselling and an opportunity to take an HIV test as part of the Prevention of Mother-to-Child Transmission of infection, and how these women responded. Data for this study were drawn from the Botswana AIDS Impact Survey 2001, a nationally-representative sample survey. The survey successfully interviewed over 4,494 of 4,728 eligible women on various issues relating to HIV/AIDS at both household and individual levels. Frequencies, cross tabulations, and logistic regression were used for data analysis. Over half (57.9%) of the women were offered HIV/ AIDSrelated information, counselling, or testing. Age, education, and residence were important predictors of being offered HIV counselling or testing. Younger and more-educated women and those residing in towns were more likely to be offered both HIV counselling and testing than older, less-educated, and rural women. Seventy-nine percent of the women who were offered HIV testing agreed to undergo the test regardless of their background characteristics. The number of pregnant women who underwent HIV testing during antenatal-care accounted for only a fifth (21%) of all antenatal-care attendees in 2001. Lack of capacity to deliver voluntary counselling and testing services to all pregnant women attending antenatal care is one of the biggest challenges to increased use of voluntary counselling and testing services. Key words: HIV; HIV infections; Acquired immunodeficiency syndrome; AIDS serodiagnosis; Counselling; Disease transmission; Antenatal-care; Botswana INTRODUCTION Correspondence and reprint requests should be addressed to: Serai Daniel Rakgoasi Lecturer, Department of Population Studies University of Botswana Private Bag UB 0022, Gaborone, Botswana Email: rakgoasi@mopipi.ub.bw HIV/AIDS poses a serious socioeconomic development challenge for Botswana. During 1992-2002, the rate of HIV prevalence among antenatal clinic attendees, aged 15-49 years, in Botswana increased from 13.8% to 35.4% (1). By 2003, the prevalence of HIV among pregnant women had increased to 37.4% (2). In all districts, the rates of HIV prevalence among pregnant women remained over 20%, with some districts exceeding 50% (3). Mortality across all age groups is on the rise, and life expectancy has declined and could be as low as 29 years by 2010. Owing to increased morbidity and mortality due to HIV/AIDS, the human development index of Botswana declined from 71 in 1996 to 122 in 1999/ 2000 (3). The Government of Botswana has declared HIV/ AIDS a national emergency and has committed itself to a comprehensive and expanded multi-sectoral response to curb the impact of the HIV/AIDS epidemic. This response is coordinated by the National AIDS Council (NAC) and chaired by the President of Botswana. The Council has representatives from 17 sectors, including civil society, private and public sectors. This multi-sectoral response is guided by the National AIDS Policy of 1998, which is aligned to the National Development Plan 9 (2003-2009) and the Botswana National Strategic Framework for HIV/AIDS (2003-2009), whose aim is to have no new infections by 2016. Currently, all ministries have HIV/AIDS sector committees whose function is to mainstream HIV/AIDS into sector plans and full-time HIV/AIDS coordinators who coordinate, plan,

59 J Health Popul Nutr Mar 2005 Rakgoasi SD and implement sector responses, and monitor and evaluate the impact of sector actions. At the district level, District Multi-Sectoral AIDS Committees (DMSACs) coordinate district-level responses. The programme for the Prevention of Mother-to-Child Transmission (PMTCT) of HIV was introduced in Botswana in 1999. Its main goal is to improve the survival and development of children by reducing HIV-related morbidity and mortality. It aims at reducing the number of new HIV infections in children occurring through mother-to-child transmission by half by 2009. The programme offers HIV counselling and testing services to pregnant women as a gateway to free antiretroviral regimens aimed at reducing the chances of mother-to-child transmission of HIV. The objectives of offering voluntary counselling and testing services to pregnant women include acceptance of the test, provision of care for HIVinfected persons, prevention of mother-to-child transmission, and psychological support for infected individuals. While the PMTCT and voluntary counselling and testing services have been integrated into national healthcare packages, such as the Ante Natal Program, manpower shortages at the national and district levels, inadequate space for counselling at the facility level, and limited HIV testing facilities are the major cons-traints (4). Because of these constraints, a phased approach to scaling up the programme from the original pilot sites of Gaborone and Francistown was adopted. This study examines the extent to which women accessing antenatal-care services in Botswana were offered HIV-related information and counselling and an opportunity to take an HIV test as part the prevention of mother-to-child transmission of infection, and how these women responded. MATERIALS AND METHODS Data for this paper were derived from the Botswana AIDS Impact Survey 2001 (BAIS) conducted by the Central Statistics Office of the Government of Botswana. The survey was designed to provide estimates of health indicators at the national and district levels and provide upto-date information for assessing HIV/AIDS intervention programmes (5). The sample was selected using a two-stage probability sample design. The first stage involved the sampling of enumeration areas as primary sampling units with probability proportional to measures of size where measures of size were the number of households in each enumeration area. At the second stage, 2,126 households were systematically selected from a list of 2,023 occupied households prepared at the beginning of the fieldwork. Of the occupied households, 1,781 (88%) were successfully interviewed. The response rate in urban and rural areas was 90% and 86% respectively. Of 4,728 eligible respondents identified, 4,494 (95%) were successfully interviewed (5). (The detailed methodology is available in the BAIS 2001 [5]). For the purpose of this paper, the BAIS sample was restricted to women with a pregnancy history who attended antenatal care during the study year. Variables Responses to four questions were used for providing a picture of the extent of HIV counselling and testing services. The questions are: Whether a woman was offered HIV counselling/information when she attended antenatal care; Whether she was offered a test for HIV; Whether she agreed to undergo the test for HIV; and finally, whether she obtained the results of the test. For voluntary counselling and testing to be effective in reducing mother-to-child transmission (MTCT) of HIV, the service provider should have the capacity to provide voluntary and PMTCT services, and the clients should be ready to accept and use such services. The main outcome variable in this paper is a proxy for the capacity of health services to provide voluntary counselling and testing and PMTCT services. It is a composite variable combining responses to the first two questions on whether the woman was offered HIV counselling or information and another on whether she was offered HIV testing when she attended antenatal care. The second outcome variable is a proxy for women's willingness to use voluntary counselling and testing and PMTCT services, and it combines responses to questions on whether the woman agreed to undergo HIV testing in cases where it was offered and on whether she came back for the test results. RESULTS Univariate and bivariate results are presented first, followed by results of the logistic regression model. Although both bivariate and logistic regression results are presented, more emphasis is placed on the latter. Where there appears to be an inconsistency between the bivariate and the logistic regression results, those of the logistic model were preferred. This is because the logistic regression model controls for more confounding variables

HIV counselling and testing of pregnant women 60 compared to bivariate analysis. Table 1 shows the sociodemographic characteristics of the sample. Table 2 shows the descriptive statistics on the percentages of women who were offered HIV counselling or testing, while Table 3 presents similar statistics on the percentages of women who agreed to undergo HIV testing and came back for the results. Table 4 shows the results of the logistic regression. Table 1. Sociodemographic characteristics of respondents Variable No. of respondents Percentage Age (years) <20 83 7.1 21-29 454 38.6 30-39 360 30.6 40-49 279 23.7 Education Primary and below 412 40.2 Secondary and over 613 59.8 Marital status Married 261 22.2 Living together 286 24.3 Divorced/widow/separated 63 5.4 Never married 566 48.1 Number of births 1 333 28.3 2 281 23.9 3 199 16.9 4 and over 363 30.8 Locality Town 311 26.4 Urban village 286 24.3 Rural 579 49.2 Facility Government/public 72 6.4 Private/NGO 1,048 93.6 NGO=Non-governmental organization The age of the sample population ranged from 16 years to 49 years, with a median age of 31 years and a mean age of 32 years. Forty-five percent of the sampled women were aged less than 30 years, with teenagers comprising 7% of the sample. Forty-nine percent resided in rural areas, 26% in towns and cities, and just under a quarter (24.3%) in urban villages. More than half (58.8%) had secondary education and over. Twenty-eight percent had one birth, 24% had two births, and 47% had three or more births. Ninety-five percent attended antenatal care at a modern health facility. Ninety-four percent accessed antenatal-care services from the government facilities, while the rest used the private, mission, or NGO health facilities. HIV information, counselling, and offer of HIV test Close to 6 in every 10 women (57.9%) who accessed antenatal-care services in 2001 were offered HIV information and counselling or an HIV test or both. This proportion was significantly higher among younger women and declined to 4 in 10 among women aged over 40 years. A significantly higher percentage of women with secondary education or more and those residing in towns were also offered HIV information or testing compared to women with primary education or less and those residing in rural areas and urban villages. While the private or NGO health facilities catered for a very small percentage of the antenatal-care attendees, they were equally likely to offer women HIV information and counselling or HIV testing. The results of logistic regression indicated that age, education, and residence were significantly associated with the odds of being offered HIV information or testing. Compared to women aged 40-49 years, the teenage antenatal-care attendees and women aged 21-29 years were four times more likely to be offered HIV information or testing, controlling for other variables. Having secondary education also increased the odds of being offered HIV information or testing by 48% compared to having primary education, while residing in a rural area decreased such odds by 32% compared to residing in towns. HIV testing Table 3 shows the percentages of women who agreed to undergo HIV testing when it was offered. Seventynine percent of women who were offered HIV testing agreed to undergo testing. While this proportion varied according to sociodemographic characteristics of women, the variations were not statistically significant. However, despite the high percentage of women accepting HIV testing from among those who were exposed to HIV information and counselling, this percentage represents only 21% of the antenatal-care attendees in 2001. This is because 40% of the antenatal-care attendees in 2001 were not offered HIV information and counselling or HIV testing. The results of the logistic regression indicated that none of the sociodemographic characteristics of women was significantly associated with the odds of accepting HIV testing.

61 J Health Popul Nutr Mar 2005 Rakgoasi SD DISCUSSION Botswana is the first African country to provide antiretroviral therapy to its citizens on a national scale. With an estimated 40,000 deliveries annually and a prevalence rate of 37.4%, close to 15,000 babies would be born to HIV-infected mothers every year, and assuming a 40% mother-to-child transmission (without any intervention), this would result in 6,000 HIV-positive infants annually. The programme for PMTCT was launched in 1999 as pilots in the two main cities of Gaborone and Francistown, and the phased scaling up of the programme beyond the pilot cities was scheduled to be completed in December 2001. Table 2. Percentage of respondents who were offered HIV information/counselling and HIV testing services among antenatal-care attendees in Botswana in 2001 by selected background characteristics Variable No. of respondents % offered HIV information or HIV testing Remarks Age (years) <20 79 62.0 χ 2 =35.185 21-29 433 65.8 p=0.000 30-39 346 58.4 40-49 262 43.1 Education Primary and below 394 53.6 χ 2 =10.565 Secondary and over 593 63.9 p=0.001 Marital status Married 250 57.2 Living together 271 58.3 χ 2 =2.225 Divorced/widow/separated 59 49.2 p=0.527 Never married 540 59.1 Number of births 1 320 56.3 χ 2 =4.898 2 269 63.2 p=0.179 3 188 59.0 4 and over 343 54.8 Locality Town 294 66.0 Urban village 273 58.2 χ 2 =12.242 Rural 553 53.5 p=0.002 Facility Government/public 72 58.3 χ 2 =0.005 Private/NGO 1,048 57.9 p=0.524 NGO=Non-governmental organization While provision of antiretroviral therapy to all HIVinfected Batswana and the prevention of mother-to-child infection remain two of the highest priorities of the Government of Botswana, 43% of the women attending antenatal care in Botswana in 2001 were not offered HIV information and counselling or testing. Manpower shortages at both district and national levels, inadequate space for counselling at the facility level, and limited testing facilities are some of the constraints the programme faced (4). These constraints disproportionately affect certain groups of women more than others, such as rural, older and less-educated women. Rutenberg et al. reported that the mixed success that the African governments have had in integrating HIV-related care into the maternal and child health (MCH) setting is due, among other things, to short-comings in the health systems of the respective countries (6). Despite the existing capacity constraints, there is an encouraging level of acceptance of the HIV test among women who were offered this service. While some studies have associated acceptance of voluntary counselling and testing services to maternal age and socioeconomic status (7), no such relationship was found in this study. A significantly high percentage (79%) of women who were offered HIV counselling and testing services agreed to test, irrespective of their sociodemographic characteristics, such as age, education, or urban or rural residence. The results suggest that, once exposed to HIV information and counselling, the majority of women will

HIV counselling and testing of pregnant women 62 accept HIV testing if it is offered, regardless of their sociodemographic characteristics. Research on voluntary counselling and testing has found that, despite the low use of voluntary counselling and testing services among certain groups of African women, the majority accepted voluntary counselling and testing services when these are offered (8). While close to 8 in every 10 women who were offered information and counselling or HIV testing agreed to take the test; these women represent only 21% of all the necessary conditions for more pregnant women to be exposed to the voluntary counselling and testing services. Botswana has relatively good healthcare infrastructure compared to other countries in sub-saharan Africa, with every citizen estimated to be within 15 km of professional medical help (9), and over 94% of births receiving antenatal care in a modern health facility and 87% of deliveries being medically supervised (10). This good health infrastructure can ensure the successful integration of voluntary counselling and testing and PMTCT Table 3. Percentage of respondents who agreed to undergo HIV test among antenatal-care attendees in Botswana in 2001 by selected background characteristics Variable No. of respondents % who agreed to test/tested and got results Remarks Age (years) <20 22 75.9 χ 2 =0.710 21-29 138 78.3 p=0.871 30-39 89 82.0 40-49 42 78.6 Education Primary and below 87 83.9 χ 2 =0.723 Secondary and over 186 79.6 p=0.249 Marital status Married 64 81.3 Living together 76 81.6 χ 2 =1.861 Divorced/widow/separated 10 90.0 p=0.602 Never married 148 76.4 Number of births 1 104 77.9 χ 2 =1.299 2 67 83.6 p=0.729 3 49 75.5 4 and over 81 79.5 Locality Town 101 80.2 Urban village 66 83.3 χ 2 =1.398 Rural 131 76.3 p=0.497 Facility Government/public 29 89.7 χ 2 =2.134 Private/NGO 269 78.1 p=0.107 NGO=Non-governmental organization antenatal-care attendees in 2001. A significant percentage of the antenatal-care attendees in 2001 were never offered HIV counselling or testing. Such a low use of voluntary counselling and testing services among pregnant women is likely to have exposed more babies of HIV-positive women to HIV infection than would have been the case if more women were offered HIV information and testing. Thus, a major window of opportunity to increase the use of voluntary counselling and testing and reduce mother-to-child transmission of HIV among more women in Botswana seems to lie in creating services into the existing programmes and services. The Government of Botswana, since 2001, has put certain measures and recommendations in place to address the problem of insufficient capacity. Some of these measures include establishing both public and private implementation teams and supporting structures, capacity-building, strengthening of prevention initiatives and uptake monitoring (11). Currently, most health facilities use the services of lay counsellors to alleviate the work load of health staff; all hospitals and clinics in the country have port-a-camp cabins to provide extra space for counselling,

63 J Health Popul Nutr Mar 2005 Rakgoasi SD and they use rapid testing kits to cut down on the waiting time for results (4). Nurses and midwives have also undergone short training courses on HIV/AIDS to allow them to provide counselling and treatment services to eligible patients. There is already evidence that efforts to increase capacity have generated an increased demand for voluntary counselling and testing services. For example, during the third quarter of 2003, 11,329 of the 12,557 (90%) of new antenatal-care attendees were pre-test counselled, before delivery, and 9 in every 10 infants of these mothers were also given AZT after delivery (2). While all districts in the country have established District Multi- Sectoral AIDS Committees (DMSACs) and large-scale vertical programmes, such as PMTCT, Home-based Care, and Orphan and Vulnerable Children, are being implemented, a phased approach to the distribution of antiretrovirals (ARVs) is being implemented (1). It is estimated that most of the 300,000 HIV-positive Batswana in the 15-49-year age group do not know that Table 4. Adjusted odds that a woman will be offered HIV information/counselling and HIV testing during her visit to antenatal-care by selected background characteristics Variable Odds ratio 95% CI for exp (ß) Lower Upper Age (years) <20 4.509*** 2.282 8.909 21-29 4.064*** 2.509 6.584 30-39 2.076*** 1.392 3.094 40-49 1.000 --- Education Primary and below 1.000 --- Secondary and over 1.484*** 1.75 2.048 Marital status Married 1.000 --- Living together 0.839 0.553 1.274 Divorced/widow/separated 1.086 0.563 2.095 Never married 0.890 0.609 1.303 Number of births 0.197 0.552 1 0.330 0.360 0.894 2 0.567 0.489 1.200 3 0.766 --- 4 and over 1.000 Religion --- Christian 1.000 0.26 3.641 Non-Christian 0.310 0.672 1.426 No religion 0.979 Locality --- Town 1.000 0.539 1.121 Urban village 0.777 0.492 0.949 Rural 0.683** Facility 0.612 1.874 Government/public 1.071 --- Private/NGO 1.000 ***Significant at p<0.01; **Significant at p<0.05 The group with 1.000 is the reference category CI=Confidence interval; NGO=Non-governmental organization 8,000 (71%) of whom underwent HIV testing. Of the 8,000 antenatal-care attendees who underwent HIV testing, 2,640 (30%) tested HIV-positive. Of those who tested positive, 1,697 (64%) were given zidovudine (AZT) they have HIV (3) and that, while 110,000 of them are eligible for immediate HIV therapy, most of them are unaware of their HIV status, and it is, therefore, unlikely that the full patient load will present for therapy as soon

HIV counselling and testing of pregnant women 64 as it becomes available (11). The secondary epidemic of fear and stigma associated with HIV/AIDS continues to present challenges to the success of voluntary counselling and testing and PMTCT programmes. Fear and stigma exacerbate the HIV epidemic by preventing people from knowing their HIV status and taking appropriate steps to protect others. There is, therefore, a need for voluntary counselling and testing programmes to address sexual communication and decision-making, stigmatization, negative reactions leading to violence, and involvement of partners of individuals who come for HIV testing services (8,14,15). These factors have been identified as serious challenges to the success of voluntary counselling and testing programmes (16). A study of attitudes towards prenatal HIV testing among pregnant women in India found that, while 86% of women would agree to undergo prenatal HIV testing, just over a fifth would make that decision independently, while 46% indicated that their husbands will have to make that decision (13). Another study found that 30% of HIV-positive women did not disclose their status to their partners for fear of their reaction (13). Thus, while in the short-term strategies available to the African governments to enhance delivery of effective intervention may include group pre-test counselling, universal offer of testing, and mass treatment without testing, long-term strategies require community education and the involvement of men to promote voluntary counselling and testing services (12). The BAIS did not collect information on reasons why women failed to use certain services, or the characteristics and attitudes of voluntary counselling and testing service providers, despite the fact that the characteristics and attitudes of those providers have an impact on women's use of the services. The unavailability of qualitative data means that explanations of certain observations from quantitative results cannot be done without some degree of speculation about the reasons why some women were not offered or chose not to use certain services. The survey did not cover questions on uptake of antiretrovirals among HIV-positive women, or the reasons why certain women declined to use services offered to them. While the healthcare system of Botswana has been overwhelmed by the HIV/AIDS epidemic, the Government and other stakeholders have invested significant resources and shown interest in HIV prevention, treatment, and care. The healthcare system's capacity to deliver voluntary counselling and testing and PMTC services has improved tremendously as witnessed by a significant increase in the number of antenatal-care attendees who were offered voluntary counselling and testing services between 2001 and 2003. However, even with an increased access to voluntary counselling and testing, stigma and discrimination remain serious challenges to increased use of voluntary counselling and testing and PMTCT services. REFERENCES 1. Botswana. Ministry of State President. National AIDS Coordinating Agency. The state of the 2002 national response to the UNGASS Declaration of Commitment on HIV/AIDS. Gaborone: National AIDS Coordinating Agency, Ministry of State President, 2003. 14 p. 2. Botswana. Ministry of Health. Family Health Division. Programme uptake. PMTCT Preven Mother Child Trans HIV Prog (A quarterly newsletter for implementing partners) 2003 Dec;1:1-2. 3. Botswana. Ministry of State President. National AIDS Coordinating Agency. National strategic framework for HIV/AIDS 2003-2009. Gaborone: National AIDS Coordinating Agency, Ministry of State President, 2003:15-16. 4. Botswana. Ministry of Health. Family Health Division. Annual report 2002. Gaborone: Ministry of Health, 2002:20-1. 5. Botswana. Central Statistics Office. Botswana AIDS impact survey 2001. Gaborone: Central Statistics Office, 2002:22-3. 6. Rutenberg N, Kalibala S, Mwai C, Rosen J. Integrating HIV prevention into maternal and child health care settings: lessons learned from horizon studies, July 23-27, 2002, Masai Mara and Nairobi, Kenya: horizons research summary. Washington, D.C.: Population Council, 2002:41. 7. Kowalczyk J, Jolly P, Karita E, Nibarere JA, Vyankandondera J, Salihu H. Voluntary counseling and testing for HIV among pregnant women presenting in labor in Kigali, Rwanda. J Acquir Immune Defic Syndr 2002;31:408-15. 8. Kawichai S, Celentano DD, Chaifongsri R, Nelson KE, Srithanaviboonchai K, Natpratan C et al. Profiles of HIV voluntary counseling and testing of clients at a district hospital, Chiang Mai Province, northern Thailand, from 1995 to 1999. J Acquir Immune Defic Syndr 2002;30:493-502.

65 J Health Popul Nutr Mar 2005 Rakgoasi SD 9. Botswana, Government of. Masa the National Antiretroviral Therapy Program of Botswana (http://www.gov.bw/government/ministry_of_ health.html#masa, accessed on 11 July 2004). 10. Letamo G, Rakgoasi SD. Factors associated with non-use of maternal health services in Botswana. J Health Popul Nutr 2003;21:40-7. 11. Botswana, Government of. Dawn challenges and learnings of ARV implementation in Botswana. Gaborone: Ministry of Health, 2001:8-9. 12. Bassett MT. Ensuring a public health impact of programs to reduce HIV transmission from mothers to infants: the place of voluntary counseling and testing. Am J Public Health 2002;92:347-51. 13. Brown H, Vallabhaneni S, Solomon S, Mothi S, McGarvey S, Jackson T et al. Attitudes towards prenatal HIV testing and treatment among pregnant women in southern India. Int J STD AIDS 2001; 12:390-4. 14. Maman S, Mbwambo JK, Hogan NM, Kilonzo GP, Campbell JC, Weiss E et al. HIV-positive women report more lifetime partner violence: findings from a voluntary counseling and testing clinic in Dar es Salaam, Tanzania. Am J Public Health 2002;92:1331-7. 15. Best K. Counselling of couples facilitates HIV disclosure. Network 2002;21:25-7. 16. Ginwalla SK, Grant AD, Day JH, Dlova TW, Macintyre S, Baggaley R et al. Use of UNAIDS tools to evaluate HIV voluntary counselling and testing services for mineworkers in South Africa. AIDS Care 2002;14:707-26.