For any PMTCT programme to effectively prevent vertical transmission of HIV between mother and baby, a pregnant woman must successfully follow the PMTCT cascade beginning with acceptance of HIV counselling and testing to receiving ARV prophylaxis (if HIV positive) and safe infant feeding practices. Uptake of services provided in this cascade has been shown to be feasible in resource limited settings by studies conducted in Zambia and Ivory Coast where more than 80% of ANC attendees accepted the HIV test and a majority of the HIV positive women commenced ARV prophylaxis even though there were reports of patient attrition between the testing for HIV and commencement of ARV prophylaxis for PMTCT [19
]. In this study, roughly one-third of the mother-baby pairs did not receive any form of chemoprophylaxis for PMTCT. Patient attrition coupled with a relatively lower ANC attendance rate, 58% in Nigeria [21
], is likely to contribute to the suboptimal uptake of ARV prophylaxis for PMTCT observed in this study.
The HIV transmission rate of 4.8% observed for babies between the ages of zero and six weeks where both mother and baby received a pharmacological intervention for PMTCT, and 20% where neither mother nor baby received ARVs, indicates that PMTCT interventions are effective in a resource-limited program setting. These findings compare reasonably with related studies on HIV transmission rates in PMTCT program settings. In Zambia, it was observed that the transmission rates among babies between zero and six weeks old for the two categories of mother-baby pairs mentioned above were 6.5% and 20.9%, respectively [13
]. Similarly a study in South Africa by Mnyani et al recorded an overall transmission rate of 5.8% for HIV exposed babies between four and six weeks of age, where both mother and baby received some form of ARVs for PMTCT [22
In our study, over 80% of the exposed babies screened for HIV were breastfed. This finding is consistent with a related study in Zambia which showed that about 84% of HIV exposed infants whose records were reviewed had ever been breastfed [23
]. In addition, our analysis indicated that regardless of pharmacological intervention received by the mother-baby pairs, there was an increase in the age-specific transmission rates for the sub-population of babies who were exclusively breastfed and for those who had mixed feeding. On the other hand, the age-specific HIV transmission rates of babies who were formula fed showed minor changes over time. Prolonged exposure to breastfeeding is likely to have affected HIV transmission rates for babies taking the DNA PCR test at a later date compared with those who had the test at six weeks. Moreover, the observed age specific transmission rates in this study were considerably higher for babies who had mixed feeding compared with those who were exclusively breastfed; this suggests that exclusive breastfeeding is safer than mixed feeding as a feeding option for HIV exposed infants. This finding is consistent with evidence from the ZVITAMBO project in Zimbabwe, which demonstrated that compared with early breastfeeding, early mixed feeding was associated with a four-fold risk of HIV transmission at six months of age for HIV-exposed babies who had previously tested negative at six weeks of age [24
The 2010, WHO guidelines on HIV and infant feeding recommend that, provided the mother and/or baby is receiving ARVs for their health or as prophylaxis, exclusive breastfeeding should be practiced by HIV-infected mothers for the first six months of life. After the six month period, complimentary feeding should be introduced while continuing with breastfeeding for up to 12 months of age unless replacement feeding is acceptable, feasible, affordable, sustainable and safe for them and their infants before that time [25
]. The PMTCT guidelines in Nigeria endorses the WHO guidelines on infant feeding, however the reality is that pregnant women who test positive face a difficult decision about how to feed their babies which is complicated by poor access to proper feeding counselling support and the influence of family members of culturally and socially accepted feeding methods. This ultimately results in improper infant feeding practices as demonstrated by the high rate of mixed feeding practiced by the HIV positive mothers in this study. Accurate information, clear infant feeding guidance, and ongoing support by healthcare workers and family members will help HIV positive mothers succeed with their chosen strategies.
Early diagnosis of HIV in infants provides a critical opportunity to strengthen follow-up of HIV-exposed children and initiate early treatment for those who are HIV infected. If untreated, HIV infection in children is associated with very high mortality rates. The children with HIV early antiretroviral (CHER) study carried out in South Africa showed that early diagnosis and initiation of antiretroviral treatment reduces early infant mortality and HIV progression by 76% and 75% respectively [26
]. The national PMTCT guidelines in Nigeria stipulate that the first DNA PCR test should be conducted at six weeks of age for all HIV exposed infants [18
]. However, we observed in our study that over 70% of the infants came for the DBS test after six weeks. While acknowledging that the launch of the EID program in these facilities initially resulted in backlog testing of many older age HIV exposed infants, over time it was still observed that majority of HIV exposed infants took the DNA PCR test at ages well over six weeks as stipulated in the PMTCT guidelines. Conducting the DNA PCR tests for early infant diagnosis at six weeks as outlined in the guidelines will offer the health providers the opportunity to strengthen infant feeding counseling provided to the mothers in order to reduce post-natal transmission of HIV.
EID also makes it possible for HIV-exposed infants to receive an early clinical evaluation and be linked to ART care. The finding that only 25% of the 125 babies that tested positive in our study had established evidence of enrolment for paediatric ART at any of the 6 study sites is a cause for concern. The majority of HIV positive babies were not traceable and presumed to be lost to follow up. Given the disease progression and mortality rates in untreated HIV positive infants [10
], it is reasonable to assume that many of the children lost to follow up had died. However, it was not possible in this study to establish how many babies among those lost to follow up had died. Related studies done in western Malawi also reported high loss to follow up rates for HIV exposed babies and their mothers [27
Improved client tracking and strengthening the link between the PMTCT, EID and the paediatric ART, coupled with proper and effective counselling of the mothers of HIV-exposed babies before and after the DNA PCR test, will increase the likelihood of enrolling and retaining HIV-exposed babies in ART programs. One model which can be used to improve follow-up for these babies is an integration of HIV-exposed infant follow-up with routine childhood immunisation services. This model is being piloted in Rwanda and preliminary results appear very promising [28
]. Expanding the availability of PMTCT services in Primary Healthcare Centres (PHCs) by enabling access to CD4 count test through a 'hub and spoke' laboratory referral system and the provision of short course ARVs for PMTCT at PHCs is an option that is worth exploring. The feasibility of this strategy has been demonstrated in Zambia even though completion of the CD4 count assessment at the PHCs remains a challenge [29
]. The proposed strategy also has the potential to minimize patient attrition as PHCs are generally more accessible to patients.
The delay in establishing a diagnosis is another factor that likely contributed to the high attrition rate observed in this study. Though the threshold turnaround time (TAT) for the return of lab results to the DBS collection centers in the national EID program is 28 days, in this study the median TAT for processing of DBS samples was almost double this figure (47 days). A shorter TAT will minimise the frustration and anxiety caregivers face while waiting for the HIV results. This will ultimately reduce loss to follow up and provide an opportunity for early initiation for ART.
The limitations to our study include the purposive nature of the selection of sites that offered EID services in the two states where the study was carried out. In addition, use of facility based records may have led to selection bias (i.e., study population included those who sought services at facilities). These limitations affected our ability to generalize the findings of this study. Additionally, some findings have limited statistical significance due to a limited sample size. Furthermore, we acknowledge that our analysis would have been strengthened by including variables such as baseline CD4 counts and clinical staging, however, the selection of the variables included in our analysis was guided (and thus limited) by what we were able to obtain from the national PMTCT data collection tools used and maintained at the study sites.