Progress made in knowledge of HIV infection, and more particularly in the use of antiretrovirals (ARVs), has resulted in considerably reducing mother-to-child transmission (MTCT) risk. Despite the increased availability of ARVs, an unacceptably high number of infants are being infected with HIV every year. In 2007 alone, 370,000 children were infected worldwide, 90% of them through MTCT and the majority in developing countries. [1]

The World Health Organization (WHO) recommends use of efficacious ARV regimens beyond single dose nevirapine (sdNVP). [2] HIV-positive pregnant women eligible for antiretroviral treatment (ART) should be put on combination antiretroviral treatment (cART), and those not eligible should be given a short course ARV prophylaxis. Single-dose nevirapine only is given to those identified late in pregnancy or when the first two options are not feasible. WHO also recommends that all HIV-exposed infants be given an appropriate ARV prophylaxis.

The cornerstone in implementing efficacious ARV regimens is identification of HIV-positive pregnant women in need of cART for their own health. [2, 3] In addition to clinical staging, eligibility for cART is based on absolute CD4 cell count, which is often limited to secondary or tertiarylevel hospitals with capacity for immunological screening. The majority of pregnant women in resource-limited settings seek antenatal and delivery care from midwives and nurses, most often in primary health centers (PHCs). These facilities have limited capacity to perform CD4 count and identify women in need of cART, making implementation of efficacious ARV regimens a challenge. [4, 5]

In developing countries, safety and effectiveness of efficacious ARV regimens have been demonstrated in well-controlled clinical studies or in secondary and tertiary level facilities. [6–10] However, large-scale implementation in PHCs in developing countries has not been widely described. The objective of this review is to determine the uptake of and factors associated with implementation of more efficacious ARV regimens (beyond sdNVP) for prevention of mother-to-child transmission of HIV (PMTCT) among HIV-positive pregnant women in PHCs.

This review demonstrated that implementation of efficacious ARV regimens for PMTCT is possible in PHCs in resource-constrained settings. However, it also highlighted three key bottlenecks in implementing efficacious ARV regimens: (1) capacity to assess CD4 count, (2) availability of CD4 count results at clinics, and (3) capacity to initiate cART.

Most worrisome if the fact that nearly 83% of HIV-positive pregnant women did not have their blood collected for CD4 cell enumeration. This major gap can be explained by the non-systematization of same-day blood collection for CD4 count along with other probable causes found in providers' comments such as (1) limited sensitization for beneficiaries and providers on the importance of efficacious ARV regimens, and (2) a client flow that continues to adjust to this relatively new intervention. Other comments of PMTCT providers suggest that stigma may prevent women from providing a blood sample for CD4 cell enumeration in an HIV/AIDS care and treatment setting. Pregnant women, when determined to be HIV-positive, may be reluctant to be seen in the same waiting area with HIV/AIDS patients. In PHCs providing ART on site, there is usually a separate structure and/or waiting area specific to HIV/AIDS services. On the other hand, at PHCs that do not provide ART onsite, women provide blood samples in the antenatal setting of PHCs and are therefore more likely to provide a sample (Table 1).

The second major bottleneck was the receipt of CD4 count results by PHCs. Test results of 33.5% of blood samples collected for CD4 cell enumeration were never returned to the clinic (Figure 1). Possible reasons mentioned in the one-page comment form include a lack of coordination of identification numbers between PHCs and laboratories, limited efficiency of the sample referral network, frequent breakdown of CD4 count machines, insufficient number of trained laboratory technicians to run CD4 count laboratory equipment, lab fees applied in some facilities for CD4 count, and clerical errors.

The third bottleneck was observed at initiation of cART. We found that 27% of women determined to be eligible did not initiate treatment (Figure 1). Barriers suggested in the comments from providers include the limited capacity of PHCs to offer HIV/AIDS care and treatment onsite and lack of training and sensitization among healthcare providers on implementation of a more efficacious ARV regimen.

Among HIV-positive women whose CD4 count results were available, 47% were eligible for cART. When that percentage (47%) is applied to the 13,135 HIV-positive pregnant women who failed to be assessed for CD4 count, we estimate that 6,173 HIV-positive pregnant women who were eligible failed to be initiated on cART. The MTCT rate is 7% in mothers with indication for lifelong ART that receive cART, compared to 26% in HIV-positive pregnant women with indication for lifelong ART that received a short-course ARV regimen. [7, 8] We therefore estimate that 1,173 pediatric HIV infections failed to be prevented because CD4 cell counts were not available for HIV-positive pregnant women.

Among HIV-positive pregnant women that accessed ARVs for PMTCT, the proportion of those initiated on cART increased from 5.3% the first quarter to 7.1% the fourth quarter (Figure 3). These proportions, although encouraging, remain low compared to global data. UNICEF and WHO reported that in 2007, among HIV-positive women that accessed ARV for PMTCT, 9.0% were initiated on lifelong cART. [15] Moreover, a Rwandan study reported a 16.0% proportion of cART among all PMTCT ARV regimes from health centers and district hospitals. [10] The differences in proportions initiating cART between our review and the UNICEF/WHO report and the Rwanda study may be due to several factors: (1) our review included data only from PHCs and did not include district hospitals; (2) unlike Rwanda PMTCT sites, ZPCT-supported sites do not physically escort HIV-positive pregnant women to ART clinics; and (3) implementation of efficacious (and complex) ARV regimens were being rolled out in ZPCT-supported sites during the period of review. [Personal communication with K. Torpey, 2009]

The increasing trend in uptake of CD4 count assessment, availability of results at clinics, and cART initiation among HIV-positive pregnant women as well as the increase in the overall uptake of ARV regimes are more likely related to the continuous technical support that ZPCT provided [Personal communication with K Torpey, 2009]. ZPCT's technical support consisted of training on the implementation of the efficacious PMTCT regimen and continuous mentoring of healthcare providers. For example, the referral network was reinforced through provision of motorbikes to transport blood samples and establishment of a referral coordinator in each laboratory. Additionally, patient flow was optimized to reduce client waiting time, and the capacity of PHCs to initiate cART (whether on site, through ART mobile clinics, or through referral) was increased through cART training.

The key to improving implementation of more efficacious ARV regimens for PMTCT in PHCs lies in building the capacity to determine CD4 cell count. Looking forward, several approaches may be used to improve this capacity in PHCs: (1) providing same-day blood-draw for CD4 cell enumeration, which requires coordination between antenatal and laboratory departments; (2) building reliable networks for blood-sample transportation and communication of test results; and (3) strengthening the capacity of laboratories to perform CD4 cell enumeration with point-of-care (POC) CD4 count testing. [16–20]

POC CD4 count testing may mitigate the limited capacity of laboratories and inconsistent sample referral networks. Using POC CD4 count testing, PHCs can operate with more flexibility to organize same-day blood-draw for pregnant women determined to be HIV-positive. However, the POC CD4 approach has drawbacks: cost per CD4 count test is higher, quality assurance in testing is more difficult to ensure, and already stretched human resource, logistics, and supply chain management of PHCs would be further burdened. [16] While the current approach-a central laboratory and network of PHCs referring samples-is not optimal, establishing several CD4 count POC hubs that cover a smaller number of PHCs may be an alternative.

To improve implementation of efficacious PMTCT ARV regimens, PHCs should also maximize initiation of cART at PMTCT sites: PMTCT providers should be able to assess eligibility for lifelong ART and initiate cART. This approach requires intensive training and mentoring and should be balanced with the already heavy workload of healthcare providers in antenatal clinic (ANC) settings. Given the human resource shortages experienced by most developing countries, increasing the number of providers at PHCs may not be feasible.

Another option may be to schedule regular mobile ART clinics visits at antenatal clinics on designated maternal and child health days.

One major limitation of this review is that only CD4 count (CD4 ≤ 350 cells/mm³ as per the national recommendations) was considered as the criteria to initiate cART. Although clinical staging is another important criterion, no related data were available. Additionally, analysis by age was not possible since tools developed for this review only captured aggregated data.

Another limitation is that reporting on "uptake of ARVs" refers only to dispensation of ARV drugs and not adherence to the dispensed drugs. Data to ascertain adherence to dispensed ARV drugs was unavailable. In similar resource-limited settings, self-reported adherence to ARVs among pregnant women accessing PMTCT ranged between 90% and 95%. [21–24] However, a study conducted in Zambia that measured the presence of NVP in the cord blood of women that accessed PMTCT found an actual adherence rate of only 68%. [25] In addition, studies in Kenya and Zambia found that women delivering at home were two to three times more likely to be non-adherent. Given that more than 50% of women deliver at home in Zambia, non-adherence could be a major issue. [9, 11, 22] Adherence to complex ARV regimens remains mostly unknown.

The review revealed that implementation of efficacious ARV regimes for PMTCT is possible in PHC in resource-constrained settings. The majority of women identified eligible for ART were initiated on cART; however, only a minority of HIV-positive pregnant women were assessed for CD4 count and had their test results available. Factors associated with implementation of more efficacious ARV regimens included timing of blood-draw for CD4 count and capacity to initiate cART onsite where PMTCT services were being offered. In order to improve implementation, multiple obstacles at the health system level should be addressed. Additional organizational and programmatic research should examine the causes and potential solutions to bottlenecks. Further research should also be conducted to ascertain adherence rates of HIV-positive pregnant women to complex ARV regimens.