Failure to test children of HIV-infected mothers in South Africa: Implications for HIV testing strategies for preschool children

Summary

Objectives

To assess the uptake of HIV testing among preschool children with HIV positive mothers in a peri-urban population-based study in KwaZulu-Natal (KZN), South Africa, an area of high HIV prevalence.

Methods

All children four to six years old and their primary caregivers from the area were invited to participate. All participants were asked about prior HIV testing and were offered counseling and voluntary HIV testing irrespective of previous testing. 27 HIV-infected mothers were interviewed to identify barriers to testing their children.

Results

1583 children (88% of eligible children) and their caregivers participated. Of the biological mothers, 86% were previously tested for HIV (27% tested positive). Among the surviving 244 children born to an infected mother only 41% had been tested for HIV (23% tested positive). Subsequently, 90% of previously untested children of infected mothers underwent HIV testing (9.3% were positive). Overall seroprevalence among study children was 4.9%. All infected mothers interviewed endorsed the belief that children of HIV-infected women should be tested for HIV. Women who missed opportunities for antenatal HIV testing reported no systematic testing of their children at later ages.

Conclusions

In this community with high HIV prevalence, HIV testing of children is infrequent despite high testing coverage among caregivers. The low proportion of children tested for HIV, particularly those of infected mothers, is of great concern as they are at high risk for morbidity and mortality associated with untreated childhood HIV infection. HIV testing programs should strengthen protocols to include children, especially for those who missed PMTCT opportunities in infancy.

Introduction

Identification of HIV-infected individuals is an essential step for initiation of life saving prophylaxis and therapy as well as reduction of virus transmission. Even in rich nations, however, as many as 20–30% of individuals infected with HIV are unaware of their status (CDC 2004). In sub-Saharan Africa, where the burden of disease is largest, it is estimated that only 12% of men and 10% of women have been both tested for HIV and received their test results (WHO 2007, 2009). During childhood, early diagnosis of HIV infection is critical for survival, as the disease course in young children is often rapidly fatal if untreated: more than half of perinatally infected children succumb by 2 years, and 80% by 5 years of age (Newell et al. 2004).

The expanded access of anti-retroviral medications over the past decade has brought unprecedented opportunities to extend survival and improve the quality of life for children with HIV. Nonetheless among the more than estimated 2 million children under the age of 15 infected with HIV, only a fraction of those in need (35%) are receiving treatment (WHO 2009). While incomplete identification of HIV-infected children may be an important reason for the low treatment coverage of children, few data evaluating completeness of testing of children in high prevalence countries are available (Kellerman & Essejee 2010). Reported experience of testing in children surviving infancy is particularly limited as HIV testing programs in resource-limited nations have concentrated heavily on testing among pregnant women and young infants (Rollins et al. 2009; Chersich et al. 2008).

We describe findings from a community-based epidemiologic study of pre-school children conducted in a peri-urban population of KwaZulu-Natal, South Africa. The community studied had undergone community-based HIV testing campaigns conducted by local organizations in the 6 months before our epidemiologic study began. The aim of this analysis was to assess the uptake of HIV testing among adult care-givers and their preschool-aged children, as well as the caregiver characteristics associated with consent for testing of children in an area with a high prevalence of HIV. We sought to evaluate the pattern of testing, attitudes and past experiences related to HIV testing of children of HIV-infected women.

Methods

This study was part of the Asenze Study, an on-going epidemiologic study to determine the prevalence of developmental disabilities among preschool children and predictors related to school enrollment and performance. We conducted a household survey in 5 tribal areas in a peri-urban area in outer west district of Durban, KwaZulu-Natal, South Africa. All children aged 4–6 years and their parent or primary caregiver were interviewed with a standardized health questionnaire and invited to a detailed health and psychosocial assessment at the project research center. All children and caregivers who enrolled in study were offered HIV testing during the study visit at the project research center irrespective of prior testing. Consent for HIV testing was obtained by trained counselors in accordance with local guidelines (National Department of Health, South Africa 2010). A separate consent was obtained for participation in the main research study. In order to learn about attitudes and prior experiences related to HIV testing of their children, we also conducted in-depth interviews with 27 HIV-infected mothers whose children had not been previously tested.

This study was approved by the Research Ethics Committee of University of KwaZulu Natal, Durban, South Africa and the Institutional Review Board of Columbia University, New York, USA.

Results

Between September 2008 and May 2010, 1583 preschool children, aged 4–6 years of age, and their 1438 caregivers completed assessments. The characteristics of the study participants are summarized in Table 1. The primary caregiver of the child was the biologic mother for 977 (68%), the father for 32 (2%), the grandmother for 286 (20%) and a sibling or other relative in the remainder.

Table 1.

Characteristics of study children and their caregivers

	Children (N = 1583)	Mother (N = 977)	Caregiver (N = 461 )
Characteristics
Age (mean, SD)	4.9 (0.57)	31 (7.4)	45 (15.5)
Range	4.0 – 6.0	18 – 50	15 – 90

	N (%)	N (%)	N (%)
Sex-Child
Male	799 (50.5)
Female	784 (49.5)
Primary Caregiver
Mother	977 (67.9)
Father	32 (2.2)
Grandmother	286 (19.9)
Another relative	143 (9.9)
HIV testing
Prior HIV test	291 (18.4)	842 (86.2)	254 (55.1)
Prior positive HIV test	33 (11.3)	223 (26.5)	66 (26.0)

Among the 461 primary caregivers who were not mothers, 254 (55%) reported being previously tested for HIV and 242 (95%) reported knowledge of their results. Of these 242 adults, 66 (27%) reported testing positive for HIV and 39 (59%) of those testing positive were already on antiretroviral therapy (ART). In contrast, 842/977 (86%) of the biological mothers enrolled in the study had been previously tested for HIV and 223 (27%) reported a positive test. Among the HIV-positive mothers, 175/223 (78%) were tested during pregnancy and 72/223 (32%) were on ART. In total 792 mothers (81%) reported having been offered HIV testing during pregnancy as part of routine antenatal care.

Of 1583 children, 291 (18%) had reportedly been tested for HIV in the past and 1238 (78%) had not been tested; the prior HIV testing history of 54 (3.4%) children was unknown. The proportion of children tested born to HIV-infected mothers is shown in Figure 1. Among the 244 children of HIV-infected mothers only 101 (41%) had undergone HIV testing in the past, of whom 30% had been tested by 1 year and 58% cumulatively by 2 years of age. Only 24% of children whose mothers were deceased had undergone HIV testing. Caregivers reported a positive result among 33/291 (11%) of previously tested children.

Figure 1.

Testing among children with HIV-infected mothers.

While all children were offered HIV testing, among the 1238 children previously untested, consent for testing was obtained for 1012 (82%) during the study visit. Consent for child HIV-testing during the study visit was associated with caregivers' relationship to the child: if the mother was present at the study visit, 90% consented to child testing, whereas of other caregivers only 63% did. For children of HIV-infected mothers who were previously untested, consent was obtained and testing was performed in 129 (90%) during the study visit and 12 previously untested children of HIV-infected mothers had a positive antibody test. The overall seroprevalence among all 4–6 year-old children who underwent testing, either in the past or during our evaluation, was 4.9% (65/1325).

In the qualitative interviews, all HIV-infected mothers endorsed the belief that children of HIV-infected women should be tested for HIV. Some reported not having been advised to do so in the past, others did not test their children out of fear of learning their child was infected, or stated they had waited as they were not emotionally ready to find out. Opinions varied with respect to whether mothers or healthcare providers carried the responsibility for initiating testing of children of HIV-infected women. A majority of mothers also agreed that children of HIV-infected women should be tested for HIV, even if a mother was first found to be HIV infected long after pregnancy. Mothers acknowledged that HIV was treatable if children received early diagnosis but this did not reduce their feelings of fear. Feelings of guilt also emerged as a factor contributing to delays in HIV testing of children.

Discussion and Conclusions

Results from this survey of households with preschool children indicates that apart from childbearing women, uptake of HIV testing for adults in this high seroprevalence region of South Africa remains low, similar to previous reports (Mkwanazi et al. 2008; Menzies et al. 2009).

Overall, testing children for HIV infection was uncommon. Of greatest concern is that less than half of children of HIV-infected women in this high prevalence community had undergone prior testing despite the fact that many of the HIV-infected women were engaged in care and receiving ART. This is a group at high risk for HIV whose HIV status should be determined as early as possible in order to benefit from early management. A recent randomized clinical trial demonstrated that initiating treatment of children with HIV prior to one year of age reduces mortality by 76% (Violari et al. 2008).

While there are many potential reasons for the low rate of testing for children, acceptability does not appear to be one of them. Our results suggest that given the opportunity, most caregivers readily agree to have children tested. Acceptability of HIV testing of children was quite high in this study: 90% of mothers consented for their children to be tested. In particular, all of the HIV-infected mothers in our qualitative sub-sample endorsed the belief that their children should be tested for HIV. These findings argue for enhancing opportunities for HIV testing of children of women who were not tested during or immediately after pregnancy, including maternal orphans. Despite great efforts to scale up PMTCT programming throughout sub-Saharan Africa, universal coverage has not been achieved; thus strategies for identifying children as well as women in other health services or in community settings are warranted.

Provider-initiated HIV testing and counseling (PIHTC), where healthcare workers advise that children undergo voluntary testing at all points of contact at health care facilities, is one of the strategies advocated for improving identification of HIV-infected children in countries with generalized epidemics where treatment services are available (WHO 2007). Our data indicate that adult HIV treatment programs have an essential role to play by extending PIHTC for children of HIV-infected parents enrolled in their services. Results of pilot programs that encourage HIV-positive adult patients attending ART clinics to refer their family members including children are promising (Jonathan et al. 2010; Sheehy et al. 2009).

A number of measures are required to achieve equitable access to HIV testing and treatment for children. National guidelines for PIHTC that address issues specific to children must be adopted. Policies that provide a framework for consent procedures for children are also necessary. The South African Children's Act and National Strategic Plan for Children provides a legal framework and makes provisions for critical issues such as who may consent to testing a child when a parent is not readily available, an essential provision in a nation where between 1.5 and 3 million children have lost one or both parents to AIDS (UNGASS 2010). Health systems treating adults must develop a protocol to offer testing to the children of adults testing positive. Policy development must be accompanied by structured dissemination, training, implementation and monitoring plans. Health care workers, counselors, educators, community partners and other stakeholders play a crucial role in closing the treatment gap for HIV-infected children.