Background

Identifying pregnant women living with HIV (WLHIV) and initiating them on antiretroviral therapy (ART) is critical for maintaining their health and decreasing the risk of vertical transmission [1]. Expansion of HIV testing for all pregnant women at the first antenatal care (ANC) visit with immediate ART initiation has led to significant reductions in new pediatric HIV infections [2]. In 2010 across UNAID’s 21 focus countries (those with the largest populations of pregnant WLHIV), less than half of all pregnant WLHIV received ART and there were more than 250,000 new pediatric infections, whereas in 2020, 85% of pregnant women received ART and approximately 110,000 children newly acquired HIV [2]. Despite high coverage of HIV testing and ART among pregnant women there are still many children newly acquiring HIV each year [1, 3, 4].

There are several factors driving vertical transmission of HIV, including poor retention of WLHIV in care and on ART [5]. Loss to follow-up from prevention of mother-to-child transmission (PMTCT) and ART services during pregnancy and in the postnatal period increases the risk of vertical transmission and compromises women’s health [2]. It is estimated that almost a third of women in high HIV burden countries who initiate treatment during pregnancy are not retained by 6 months [5]. In addition, incident HIV infections occurring during pregnancy and in the postnatal period are contributing to an increasing proportion of new pediatric HIV infections [2, 6]. When women newly acquire infection during pregnancy or breastfeeding, there is higher risk of vertical transmission [7] and these women are often missed for interventions, including ART initiation, infant postnatal prophylaxis and early infant diagnosis (EID) [8].

The World Health Organization (WHO) recommends repeat HIV testing for pregnant and postpartum women in countries with generalized HIV epidemics to identify women with incident infections occurring during pregnancy and breastfeeding [9]. Repeat testing in ANC, postnatal and maternal child health (MCH) services, such as family planning and pediatric immunization clinics, are recommended. Repeat testing done in these settings can also be used to identify and reengage WLHIV who previously tested positive but were lost to care, as well as to engage women who did not deliver in health facilities [4]. While many countries have adopted national guidelines for retesting pregnant and breastfeeding women [10], there is limited information about how effectively these guidelines are being implemented in high HIV burden, low resource settings, including information about strains on the healthcare system from increased testing volume [11, 12]. A study of routinely collected data from Kenya, which has had repeat testing guidelines since 2012, found that only 27% of pregnant women who tested HIV-negative at the first ANC visit and returned for a follow-up visit when eligible for retesting (>12 weeks) received a test [13]. In Tanzania, where national guidelines also call for retesting, among women who had delivered in a health facility and had tested HIV-negative during pregnancy, only 30% were retested at the time of delivery; among those tested, 13.3% were found to be HIV-positive [14]. These data suggest significant challenges to implementation of repeat testing for pregnant and postnatal women and emphasize the critical role of retesting in detecting incident HIV infections.

UNAIDS data for Mozambique show that in 2020, most of the infections acquired by children during pregnancy were because mothers dropped off treatment, and among children acquiring infection during breastfeeding, more than 40% were the result of mothers with incident infections [2]. In order to support full implementation of national HIV testing guidelines for pregnant and breastfeeding women, ICAP at Columbia University, in collaboration with the Mozambican Ministry of Health (MOH) and with funding from the US Centers for Disease Control and Prevention (CDC), provided technical assistance at 10 health facilities in 2019 [15]. Using routinely collected data we report testing coverage and positivity from ANC services, test-positivity among postnatal women in MCH services, and test-positivity and referral for treatment among infants of mothers tested in child welfare clinics (CWCs).

Methods

Project implementation

The project was conducted at 10 health facilities in Nampula province, Mozambique from April through November 2019. The health facilities were purposively selected based on high volumes of women attending both pre- and postnatal care. The aim of the project was to provide support to the facilities to fully implement existing national HIV testing guidelines for pregnant and postnatal women (summarized in Fig 1). Per the guidelines, pregnant women should receive HIV testing at the first ANC visit with retesting every 3 months until cessation of breastfeeding. All pregnant and breastfeeding women coming for care are to be assessed for retesting eligibility (using facility registers or hand-held health cards) in the following settings: ANC, maternity wards (labor and delivery), postnatal clinics (where woman should have postpartum visits at 48 hours, 7, 21 and 28 days), family planning clinics and CWCs. CWCs are clinics within the health facilities where infants and young children from ages four weeks to five years of age are brought to receive weight and growth monitoring checks, immunizations, and nutritional and developmental assessments. Infants and children who are malnourished or sick or have health problems are referred to the at-risk child clinic (CCR) and under-5 clinic consultation (CCD), respectively. All health facilities selected for the project provide all of the services listed.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Maternal testing and retesting points of care per Mozambique’s national guidelines (2015) [15].

https://doi.org/10.1371/journal.pone.0283558.g001

Project support to the health facilities included monthly data reviews to help improve documentation and ensure eligible women received retesting at all points of care, as well as clinical mentoring for health facility staff focused on identifying and overcoming gaps in testing coverage. Support and clinical mentoring was provided by two trained ICAP nurses and a data manager. Due to high patient volumes in CWCs, one mentor mother (peer educator/mother living with HIV) and two additional trained counselors were hired to screen maternal health cards to identify women eligible for retesting and counselors conducted HIV testing in the CWC at each health facility. Mentor mothers served as patient navigators for women testing HIV-positive. At some facilities, as a result of limited space, it was necessary to erect tents within courtyards to provide private spaces for testing in the CWC. HIV-exposed infants (HEI) (children of women who tested HIV-positive) in CWC were escorted to the CCR for point of care (POC) HIV testing with polymerase chain reaction (PCR). Results were returned to the mother the same day, and children who tested positive were started on ART.

Results

Over eight months in 2019, across 10 health facilities in Nampula, Mozambique, a total of 84,845 HIV tests were provided to pregnant and postnatal women in maternal and child health services, including 55,101 that were likely repeat testing. Data from antenatal clinics show that 29,808 pregnant women with unknown HIV-status attended a first visit from April through November 2019 and 99.8% of those women were tested (Fig 2). Among the 28,233 women who were HIV-negative at the first visit, only 11,504 (40.7%) returned for a follow-up visit >3 months after the first visit and were thus eligible for HIV retesting. Most women (n = 9,759, 84.8%) who returned for a follow-up visit when they were eligible for retesting received testing and 26 (0.3%) tested positive for HIV.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Antenatal clinic HIV retesting cascade, 10 health facilities in Nampula, Mozambique, April-November 2019.

https://doi.org/10.1371/journal.pone.0283558.g002

During this time period, 4,468 women were retested in maternity at the time of delivery and 1.2% (n = 52) tested HIV-positive; 697 women were retested in the postnatal clinic and 1.0% (n = 7) tested HIV-positive; 678 were retested in the family planning clinic and 2.1% (n = 14) tested HIV-positive; and 39,499 women were retested in the CWC, where 1.0% (n = 397) tested positive for HIV. Highest test-positivity was observed among young adolescent women 10–14 years: in maternity 10.3% (4 out of 39 tested) and in family planning clinics 16.7% (3 out of 18 tested) tested HIV-positive. Only three women 40 years and older tested HIV-positive in CWC (none tested positive in other settings) (Table 1). Test-positivity was similar for older adolescents (15–19 years) across all settings and was highest in the family planning clinics among 30-34-year-old women (3.6%) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Women retested in maternity, postnatal, family planning and child welfare clinics and proportion HIV-positive by age, Nampula, Mozambique, April-November 2019.

https://doi.org/10.1371/journal.pone.0283558.t001

Among 26,503 women retested in the CWC from July through November of 2019, 212 (0.8%) tested HIV-positive and 157 (74.1%) of these women were referred to ART services (Fig 3). One hundred and forty-five (68.4%) infants of mothers who newly tested HIV-positive in the CWC received a PCR test and 19.3% (n = 28) tested PCR-positive. All 28 (100.0%) infants testing PCR-positive were recorded as being referred for ART. In the CWC, mothers of infants 1–3 months of age were the largest group who were retested, whereas few mothers of children who were 10 and 11 months of age received retesting (Fig 4). Maternal test positivity according to the infant’s age was lowest for mothers whose infants were 5 months and 10 months at the time she was retested and was highest among mothers of infants >12 months of age (0.6% and 1.6% respectively).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Maternal and HIV-exposed infant testing in child welfare clinics, Mozambique July-November 2019.

https://doi.org/10.1371/journal.pone.0283558.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Number of mothers retested for HIV in child welfare clinics and test-positivity among mothers according to age of infant, Nampula, Mozambique, July-November, 2019.

https://doi.org/10.1371/journal.pone.0283558.g004

Discussion

While many countries have adopted guidelines calling for retesting of pregnant and breastfeeding women, our study is one of the first to provide a perspective on the scale of repeat testing. In addition, our analysis also showed that in ANC, lack of return for antenatal care led to poor HIV retesting coverage in pregnant women. It is noteworthy that almost all women returning for care when eligible received repeat testing. In other service points, we were limited to examination of test yield by age group, which was highest in family planning clinics and among young adolescents (10–14 years) in maternity wards. Finally, our data on HIV testing for mothers in child welfare clinics identified several critical issues, including very large volume of women in need of retesting, poor linkage to testing for HIV-exposed infants, and low test positivity among mothers but high positivity among HEI.

In these ten health facilities, coverage of HIV testing was high at first ANC visit and among those returning for follow-up care during pregnancy. Retesting yield was low with <1% of returning pregnant women determined to have HIV infection. These indicators suggest fidelity to national guidelines in ANC and low rates of incident HIV infections among pregnant women retained in ANC. However, the fact that only 41% of pregnant women returned for a follow-up visit ≥3 months after a first ANC visit is alarming. It is possible that some women went to other clinics after their initial ANC visit and had repeat testing at those facilities [16]. Unfortunately, Mozambique does not have unique health identifiers that would allow for the tracking of women across different health facilities so we were unable to estimate how many women sought follow-up care and received follow-up testing in other facilities. Better information systems that allow for tracking of women across health facilities are needed to measure retesting and to document outcomes of mother and children. Furthermore, some women may have attended their first ANC visit at >6 months’ gestation and therefore did not reach retesting eligibility during pregnancy. However, in a 2019 study conducted in Nampula, 40% of pregnant women attended their first ANC visit by 16 weeks [17] indicating that most women tested at the first ANC visit should have had a least one additional ANC visit prior to delivery where they could have received retesting.

Poor ANC attendance is a significant problem globally [18], however, in high HIV burden settings, there is increased risk to the health of women, who may not start or continue ART, and to their infants. Of particular concern are infants of mothers who acquire HIV during pregnancy and breastfeeding who are at higher risk for vertical transmission and may have delayed ART initiation due to missed diagnosis [7]. Data from UNAIDS showed that in 2020 in the 21 focus countries, 31,000 new pediatric infections resulted from women not remaining in ante- and post-natal care and 27,000 resulted from newly acquired infections during pregnancy or breastfeeding [2]. In order to retest women who initially test HIV-negative in pregnancy, women must remain engaged in care during pregnancy and throughout breastfeeding. Unfortunately, there has been little research on interventions to improve adherence to care among pregnant and breastfeeding women in sub-Saharan Africa, as a recent systematic review reported [19]. While retaining HIV-positive women remains an important focus, effective strategies for improving retention of all pregnant and postnatal women are urgently needed, including efforts to prevent women from acquiring HIV [20]. Key components of enhanced ANC and MCH services for pregnant and postnatal women should include male partner engagement with HIV testing and immediate linkage to ART for anyone positive, as well as expanded access to pre-exposure prophylaxis (PrEP) which is safe for pregnant women and recommended by the WHO to prevent HIV acquisition in mothers and infants [21–23].

Outside of the ANC, our data were limited to test-positivity, however, they provide information that can be used to target services. We found that family planning clinics had the highest testing yield across all age groups, suggesting that the post-partum period is a critical point for repeat HIV testing and that family planning clinics should be targeted to ensure all women are tested and have easy access to PrEP. Although relatively few young adolescents were tested, the highest testing yield in both maternity and family planning clinics was among 10–14 year-olds. Due to the small sample size, we are cautious in interpreting these findings however they are consistent with previous observations that adolescents are less likely to enroll in ANC and often enroll late, making them more likely to miss repeat testing opportunities during pregnancy [24] which may explain why they were identified in maternity and post-natal clinics (rather than ANC). In addition, adolescents are more likely to drop out of services after testing HIV-positive [25]. It is possible that some adolescents at these facilities who tested HIV-positive at delivery and in the postnatal period had previously tested HIV-positive and subsequently dropped out of care. These findings underscore the importance of offering repeat testing for adolescent pregnant and postnatal women across the MCH cascade and the need to offer services targeted to their specific psychosocial needs to improve retention in this vulnerable group [26, 27].

Unfortunately, our findings from child welfare clinics were not encouraging. Among mothers who tested HIV-positive, linkage to testing for infants was poor and almost 20% of those tested were also found to test PCR-positive. We do not have data on whether and when these women previously tested HIV-negative but the high rate of positivity among infants was likely due to the combination of incident infections during pregnancy or breastfeeding, and from women with established infections who did not receive ART during pregnancy. It is very concerning that one third of HEI identified in child welfare clinics were not linked to HIV testing despite efforts to immediately test all HEI on the same day using POC testing. Greater efforts are needed to ensure that all HIV-exposed infants receive diagnostic testing and have a final HIV status determination after the cessation of breastfeeding. These services are critical so that both mothers and infants who are HIV-positive can be initiated on treatment immediately to reduce the risk of morbidity and mortality [28].

Finally, it is important to note the scale of this project as an indicator of magnitude of the retesting mandate for pregnant and postnatal women per Mozambique’s national guidelines. At the 10 project sites, based on facility data, we estimate that each month almost 4,000 pregnant women attended antenatal care services, 3,500 of whom were HIV-negative at the first visit, there were more than 2,700 deliveries and 11,000 children were brought to child welfare clinics. While we do not have data on what proportion of women attending CWC were eligible for retesting, over the eight months, 40,000 women were tested in CWCs across the 10 health facilities (500 women tested per site per month). Fortunately, most mothers attending these services were found to be HIV-negative, however retesting thousands of women each month along the MCH cascade requires staff to conduct tests as well as consistent test kit supplies. It is clear that strategies to target testing to those most at risk and making testing more feasible are urgently needed if repeat testing guidelines are to be fully implemented. Screening tools and other approaches that can effectively identify high risk women are needed. One approach that should be explored is self-testing which has been used effectively in ANC [29, 30] and could be expanded to family planning clinics and other MCH services.

While our data provide important new information about implementation of repeat testing for pregnant and postpartum women in Mozambique, it also has limitations. The lack of data on gestational age at first ANC visit may have led to an overestimate of the proportion of pregnant women who did not return for retesting. While we were able to look at testing and retesting in the same women in the ANC, in other care settings we were unable to do this. A further limitation is that we did not have a denominator for the number of unique women who were tested outside of ANC so we cannot say how many women who were eligible for retesting received it and can also not identify how many unique women were retested (the same women may have been tested multiple times during the project period). In addition, other than in the ANC, we could not distinguish women tested for the first time versus those receiving repeat testing, in particular in family planning clinics. However, given the very high rate of HIV testing at first ANC visit, it is likely that most of the testing that occurred among women in the postpartum period was repeat testing. Finally, the project did not collect process indicators to measure improvement in services prior to and during implementation nor could we compare uptake of testing at these sites to other health facilities.

Conclusions

Supporting the full rollout of HIV retesting guidelines may be challenging given the high volume of HIV-negative women in need of repeated testing. Innovative strategies and enhancements to the MCH service platform will be needed to ensure the feasibility of full-scale implementation.

Supporting information

Acknowledgments

We thank the Ministério da Saúde de Moçambique for their support and collaboration to implement this project. We also thank the health facility staff and clients at the participating health facilities.