The experiences of the women in this study showed that women are apprehensive about HIV testing because of the possibility of a positive HIV result, but the atmosphere created by the health workers has made the women comfortable about testing within the antenatal clinic. At this point, eight years after the introduction of PMTCT at this hospital and four years after introducing routine HIV testing, when the programme has reached maturity, women in general are informed about the procedures before attending. Most women in this study had learned about routine HIV counselling and testing during their past visit to health facilities, through the media and their families and friends, a clear sign of how informal and formal social structures re-enforce each other in creating informed health service users. In part, these findings reveal that women’s acceptance of routine testing is the result of good access to information on the benefits of HIV testing both from informal and formal sources. On the other hand, the general perception by women of routine counselling and testing as ‘compulsory’ for all pregnant women also reinforces women’s acceptance of the test. In general, health workers favoured HIV testing during health education and counselling sessions. In fact, the possibility of not testing (opt-out) was downplayed by the ANC staff to such a point that most women think it is compulsory to test. The apparent acceptance of HIV testing by women could also mirror the inherent power relationships between women and health workers where women may be unable to go against what their health care providers tell them to do. The ongoing discourse, especially in the media, aimed at popularising the PMTCT programme further reinforced women’s perception of HIV testing within the ANC as being compulsory. The perception that HIV testing as part of ANC was compulsory, has also been documented in other African settings [20–22]. The perception by some women that testing is compulsory renders some merit to earlier concerns that RCT might undermine informed consent and patients might feel obliged to test . In the study setting this is however debatable given that most women knew about the test before coming to the health facility and were provided with additional health education within the ANC. Although, there has been speculation that the perceived compulsory HIV testing may negatively affect women’s attendance of ANC, this was not the case in our study setting where the number of new ANC attendees continued to increase even after a shift from VCT to RCT .
Women in this study said that taking an HIV test was a difficult step to make. The narratives of pregnant women in this study revealed that they feared a positive test result which was associated with the fear of death, living with HIV and being blamed for bringing HIV infection to the family. However, the routine provision of HIV testing helped such women to overcome the fear and take the test. On the other hand, most women strongly believed that HIV testing was beneficial, especially by enabling those who test HIV positive to protect their babies from HIV infection through enrolling in the PMTCT programme. In this regard, the high acceptance of HIV test could be a reflection of women’s conformity to the moral imperative of doing ‘good’ for their babies. The women also appreciated that HIV testing was beneficial since those found to be positive could access HIV treatment for themselves. The need to protect their children and the concern for their own health have been documented in other African settings as key reasons for women’s acceptance of HIV testing during pregnancy [11, 13].
Our findings differ from those of a study done in 2008 in the nearby Iganga district, in Uganda, which found that women did not fully understand the benefits of HIV testing . This difference could in part be explained by differences between the two districts in the initiation of PMTCT programmes (Iganga 2004 and Mbale in 2002) and being different study populations.
Whereas HIV testing was in general perceived as beneficial by women, for some, it placed on them an extra burden. For instance, some women were worried about disclosure of HIV status to their partners and the possibility of being blamed by their sexual partners for bringing HIV infection into the family as has been documented elsewhere . The struggles of living with HIV including uncertainty about the future of their marriage and care for the women and their children were common concerns among HIV positive women in our study. These findings are not surprising given the day to day struggles of women to secure their own and their children’s survival amidst poverty, stigma and marginalisation in the African setting . In the study area, most women, depend on their male partners for their care and that of their children .
The benefits of HIV testing for the women who test HIV negative were rarely mentioned by women during interviews. In part, this is a reflection of how HIV counselling has not prioritised the majority of the people who test HIV negative for effective HIV prevention. In line with critical theory, the limited number of health workers in the study setting, as a structural constraint , compounded the inadequacy of the post test counselling and support received by HIV negative women. Human resource constraint as a barrier to effective implementation of provider-initiated HIV counselling and testing has previously been documented in Uganda  and needs to be re-emphasised.
The women in our study expressed a strong desire to protect their babies from HIV infection as a major motivator for undergoing HIV testing. This strong desire to protect children can be explained by the values attached to children in Uganda even in the time of HIV , but also the moral imperative where children are seen as innocent and deserve protection. This strong desire by women to protect their babies has not been taken advantage of by the health system to optimize the PMTCT services in Uganda. Currently, only 51.6% of the estimated HIV positive pregnant women in Uganda receive antiretroviral drugs to reduce the risk of mother-to-child transmission of HIV. This coverage is still insufficient to reach the desired goal of eliminating paediatric HIV infections . However, the strong desire of most women to test for HIV so as to protect their children from HIV infection further complicates the implementation of opt-out option in practice as those who may choose not to test may be perceived or made to feel as ‘not good mothers’ or ‘acting against their own health’. This finding further reveals a real life challenge of making opt-out an option for women.
Some women in our study understood routine HIV testing during pregnancy as a new ‘law’ by the government aimed at protecting babies from HIV infection. For such women, HIV testing can be interpreted to mean embodying the good of the Nation as discussed in another context by Booth Karen . The implication here is that for some women, taking an HIV test may be a reflection of compliance with the state.
Whereas the health education session that doubled as a pre-test counselling session was described by mothers as informative and educative, there was a general need to strengthen post-test HIV counselling for both HIV positive and HIV negative women. A common concern among both HIV positive and HIV negative women was doubt about the test results. Doubt about the positive results among HIV positive women was linked to their limited perception of risk for HIV infection. It could also reflect that they had been faithful to their sexual partners, so they could not imagine how they could have become infected. Such women in doubt may be reluctant to utilize PMTCT drugs or to seek care for their own health.
To our surprise in this study, some women who tested HIV negative did not believe in the results. Doubt about the results among HIV negative women was enhanced when they were told they did not have syphilis, contrary to their locally accepted community ‘diagnosis’ of syphilis based on a skin rash and a pale skin. This is an example of a clash between the socio-cultural definition and understanding of illness and the biomedical understanding of disease. Such women in doubt about their HIV negative sero-status may be reluctant to adopt risk-minimising behaviours, which in turn could have negative effects on fostering primary prevention of HIV as part of the PMTCT programme. This finding depict that the doctor and patient explanatory models often co-exist in a clinic encounter  and calls for more collaborative efforts between service users and providers. In general, such apparent contradictions need to be identified and dealt with in health education and counselling. In this case, it is critical that health workers provide sufficient information to pregnant women about syphilis and other conditions that may lead to skin rash and pale skin and how they can be managed. In addition, mothers should be encouraged to re-test if they doubt the results. Another explanation for women’s doubt of their negative HIV status could be linked to the public health messages which tend to over estimate the likelihood of HIV transmission and thus many people often think testing for HIV will lead to a positive HIV result, as has been documented in Malawi [45, 46]. Suspicions about partners or reflections about one’s own health and sexual history could be other explanations for doubt of test results.
Health workers were generally described by women as caring and supportive though constrained by being few in number. Indeed, women’s experiences of limited post-test counselling owing to few health workers is a reflection of the national and global human resource crisis affecting many sub-Saharan countries. The limited post-test counselling presents missed opportunities for health workers to foster messages for primary HIV prevention. In Mbale District, the situation is worsened by the intermittent provision of HIV testing services and lack of maternity services at lower level health facilities. Some strategies that may help in this are 1) ensuring that HIV counselling and testing services are regular at lower level health facilities in the district, 2) increasing the number of health workers to allow more time for health worker-client dialogue on the interpretation of biomedical results relating to HIV and other sexually transmitted infections, 3) using lay counsellors who may include persons living with HIV as expert clients, or members of support groups who may also help, especially in providing insights on dealing with the challenges and fears related to a positive HIV diagnosis . The need to improve the quality of HIV counselling as part of PMTCT programmes has been documented by other scholars in Uganda  and in Tanzania .
Most HIV positive women in this study interpreted their results to mean betrayal and infidelity by their sexual partners. In addition, most of the women who tested HIV positive strongly believed that their partners were already HIV positive and thus not likely to discuss risk reduction strategies such as condom use with their partners. Contrary to this belief, a study of HIV-infected people receiving antiretroviral therapy in Uganda found that 43% of the spouses of HIV-infected married people were HIV negative .
Women who tested HIV positive were happy with the immediate counselling, but still expressed long term worries about survival with HIV and the care of their children. Women’s fears and concerns about meeting their needs and those of their children are not surprising given dwindling public services and social support systems in Uganda. In this regard, partnerships with other organisations are needed to address the support and survival of such women.
In contrast, most women who tested HIV negative interpreted the test results as a sign of fidelity of their sexual partners. These findings depict a limited understanding and appreciation among women of the likelihood of discordant HIV test results for themselves and their partners. Moreover, such women may not be motivated to encourage their partners to go for HIV testing. In 2008, it was estimated that 43% of the new HIV infections among adults aged 15–49 years in Uganda were among people in discordant monogamous relationships [3, 50]. Thus, the risk for HIV infection even among women who test HIV negative should be emphasized in both pre-test health education and post-test counselling.
The HIV negative women in our study revealed that counsellors advised them to remain faithful to their sexual partners to avoid HIV infection. This is of course good advice but it does not minimize the relevance of the statement that ‘women get HIV in their own bedrooms’, meaning that if only women and not their sexual partners remain faithful, the chance of contracting HIV is still high. More efforts are needed to reach men with HIV testing and HIV prevention messages within and beyond the antenatal clinic.
Strengths and limitations
Use of qualitative interviews and the inclusion of pregnant women and health care providers in the study facilitated an in-depth understanding of women’s experiences with routine HIV testing. However, our findings should be interpreted in view of the following limitations. 1) The study was conducted at a public hospital in the general antenatal care setting characterized by congestion and delays at the clinic; it is possible that women with higher incomes obtain antenatal care from private clinics and may thus be under-represented in this study. 2) Experiences of women who decline the HIV test could not be gathered since at the time of the study all women who attended the antenatal clinic at the hospital tested. 3) The study was conducted at a regional referral hospital (a tertiary health facility) with experienced health workers and relatively better facilities. Thus the gaps documented in this paper could be more pronounced at remote lower level health facilities with fewer staff and recurrent shortage of HIV test kits 4) Besides, health workers at the study site have been exposed to research for a long time, which may also limit applicability of study findings to other settings with no or minimal exposure to research 5) We conducted all interviews at the health facility since it was not easy to identify pregnant women tested for HIV as part of ANC at community level, especially HIV positive women. This might have biased the respondents to report more of their positive experiences. Indeed, some pregnant women might have been reluctatnt to blame health workers who will eventually help them at the time of giving birth. Besides, women were introduced by health workers to the research team, which might have further enhanced this bias. To minimize information bias, the fact that the researchers were independent of the hospital staff was made explicit to the study participants. In addition, qualitative interviews involving use of probes and triangulation of data from different sources helped to improve the trustworthiness of our findings.