Using a community-based rather than a facility-based approach in this study, we found that 90% of the pregnant women had been tested for HIV. Early uptake of the HIV test was more common among women who had attended antenatal care at a primary HF. Likewise, the provision of HIV counselling was also reported to be higher at primary HFs, where more women had received pre- and post-test counselling.
The study population comprised women from Ha Long city in Quang Ninh province, an area in which PMTCT has been widely implemented with support from foreign and international donor agencies. The findings therefore cannot be generalized to Vietnam as a whole since the PMTCT program is not implemented in the same way in all provinces. Yet the study does offer insights into the dynamics of a pilot site where the PMTCT services have been scaled up to community level and the findings may thus be relevant for a more general expansion of PMTCT services in the whole country. Concerning the representativeness of the study population, we were not able to obtain background characteristics of the women who did not participate in the study and were therefore not able to assess whether they differed systematically from the women who participated in the study. However, since 81% of the eligible women were included in the study, it may be argued that the findings represent the vast majority of women who had recently delivered in the study setting. Regarding the internal validity of the study, the information about HIV testing was obtained from the women through questionnaire interviews. The women's answers were not checked against any formal registration of the gestational age at which the women were tested. This lack of cross checking may have affected the results. The uptake of antenatal HIV testing was high; 90% had been tested for HIV, either at the time of antenatal care or at the time of labor. The high rate of HIV testing found in our study is in line with experiences from Thailand, where 93% of pregnant women attending antenatal care were tested for HIV . However, a facility-based study in the neighboring province, Hai Phong in 2005 showed that only 53% of the pregnant women had been tested for HIV . The studies from Hai Phong province (Vietnam) and Thailand were both facility-based, whereas our study was community-based. One of the advantages of a community-based design is that it covers women regardless of whether or not they have had contact with a HF during their pregnancy and labor and may thus, in comparison with studies which rely on a facility-based design, provide a more trustworthy picture of the acceptance of HIV testing in a society .
This study showed that the early uptake of HIV testing and provision of counseling differed depending on HF level. Early HIV testing was more common among women who had had their first antenatal visit at a primary HF, where the women had also more often been provided with pre- and post-test counselling. At the higher level HFs, women were generally tested later in their pregnancy and were not provided counselling. However, when evaluating the HIV testing services offered at different HF levels, the timing of the antenatal visit at the different levels should be taken into consideration. If 'the first antenatal' visit at a higher level facility is actually after 34 weeks of gestation or during labor, while women are attending primary facilities for their first visits for 'normal' antenatal care, this would affect timing of HIV testing and the motivation/ability of healthcare workers to tackle HIV testing. However, no significant difference of the time of the first antenatal care visit was found between primary level and higher level of HFs (Table 2). In addition women who were tested at lower level HFs were more likely to have received counseling in relation to the testing. A likely explanation for the earlier uptake of HIV testing as well as for the higher proportion of women receiving counseling at lower level HFs may be that health staff working in higher level HFs often are preoccupied with many different assignments and do not have sufficient time to spend on HIV counselling and testing; further, they are often unable to offer privacy during counselling [5, 15, 22]. In contrast, primary level HFs have better conditions in terms of both time and space for providing counselling and HIV testing for pregnant women . Hence, the antenatal care nurses who worked at primary level HFs were apparently in a better position to promote HIV testing. This assumption is supported by a number of in-depth interviews showing that pregnant women in Ha Long found the health staff at primary level HFs skilled in tailoring the HIV information and counselling to address the individual women's circumstances and concerns .
Early and voluntary HIV testing is increasingly being challenged. Due to increased availability of antiretroviral treatment, policies on HIV testing have shifted towards routine testing for HIV as a part of antenatal care [24–27]. A recent community-based study from Hanoi has documented that 85% of pregnant women were tested late and received inadequate counseling due to a lack of PMTCT services at the commune level . Hence, many Vietnamese women who are tested for HIV during pregnancy are in a position where they do not get the full benefit of the PMTCT program. Against this background it is encouraging that nearly 71% of the women in our study had been tested for HIV before or at the 34th week of gestation. This high rate of early HIV testing suggests that the antenatal care program in Quang Ninh province is functioning well, an assumption that is supported by the fact that 82% of pregnant women attend antenatal care (90% at urban sites and 80% at rural sites), and that the vast majority come for antenatal care during the first trimester . Thus, excellent conditions exist for an efficient PMTCT service which, in the setting studied, has been backed up by massive investments in both the quality and the quantity of PMTCT.
HIV counselling and testing late or during the time of labor has been advocated to be a rational way to increase PMTCT uptake . However, HIV testing at the time of labor should be treated as the last resort for prevention of MTCT, because the women then miss the opportunity to receive the full prophylactic regime as well as other PMTCT services [5, 6]. Moreover, being confronted with a positive HIV result is associated with great distress [22, 23] and labor is not the optimal time for conveying such information [23, 28]. One way to avoid these problems is to offer women HIV testing at primary HFs, when they have their first antenatal checkup.
A successful scale-up of HIV counseling and testing to lower level HFs has been documented in this study. Other studies in Vietnam have shown that the implementation of HIV testing at lower level HFs or by outreach workers may be an effective way to scale up PMTCT [14, 29]. This approach may especially apply for settings in which the lack of HIV testing at commune level is one of the main reasons for poor quality PMTCT [21, 30]. Moreover, the results of this study are in line with studies in other countries which have shown that the provision of HIV counseling and testing at community level may increase access to the service for vulnerable rural women and place them in a position where they can access and benefit from PMTCT programs [4, 8, 14, 29].