Introduction.
Despite the strong evidence for the effectiveness of anti-retroviral (ARV) drugs for improving the health of HIV-infected women and for the prevention of mother-to-child transmission (PMTCT) [1], [2], HIV-related morbidity and mortality among childbearing women and vertical transmission of HIV from mother to child continue to be major health problems in sub-Saharan Africa [3], [4]. In order to improve health outcomes, women need to successfully navigate a cascade of services including antenatal care (ANC) clinic attendance, acceptance of HIV testing, receipt of results, enrollment in HIV care, acceptance of ARV prophylaxis or treatment, adherence with maternal antiretroviral (ARV) prophylaxis or treatment, adherence to infant ARV prophylactic doses, and early infant HIV testing [5].

Unfortunately, a significant proportion of women and their infants drop out at each step along this cascade thus decreasing the effectiveness of PMTCT programs [6]–[9]. The 2009 Expert Consultation on Implementation Science Research sponsored by the NIH Office of AIDS Research identified as a top priority research on how to promote women’s linkage to and retention in care at each step of this PMTCT cascade [10].

One promising approach for improving linkages to and retention in services for women and infants is to fully integrate HIV care into ANC clinics [11]. Although definitions of “integration” vary, experiences in sub-Saharan Africa suggest that integrating ANC and HIV services may result in a variety of benefits for HIV-positive women and their families; including better uptake of services, more women receiving counselling, reduction of the time to treatment initiation, improved quality of care, and reduction of stigma [12]–[17]. Lack of integration of PMTCT into routine maternal and child health services has been identified as one of the major contributors to drop-off of women and infants at various steps in the PMTCT cascade [18]. On the other hand, the potential downsides of integration in low-resource settings include: increased provider workload in an already overburdened system, increased training needs, lack of space and equipment, lack of staff motivation to provide more services, and even “organizational culture clash” [19]–[21]. From the perspective of clients, it is possible that integration of HIV services into ANC clinics could have negative effects–such as increased wait times–for the majority of clients who are HIV-negative [22].

Since 92% of women in Kenya have at least one antenatal care visit during pregnancy, ANC clinics have become prime locations for expansion of HIV testing and PMTCT services in this country [23]. Due to these efforts, rates of antenatal HIV testing have been increasing over time in Kenya – approximately 73% of pregnant women were tested during 2008–2009 [23]; however, only an estimated 72% of pregnant women who tested HIV-positive received antiretroviral medications for PMTCT in 2009 [24]. Antenatal care has been modified to include PMTCT services, but generally has not included comprehensive HIV care and treatment for pregnant women. Comprehensive HIV care and treatment encompasses the clinical and social components necessary for the highest quality of care, including opportunistic infection prophylaxis, tuberculosis (TB) diagnosis and treatment, WHO clinical staging for HIV, highly active antiretroviral therapy (HAART), CD4 count monitoring, relevant laboratory tests (such as complete blood counts, creatinine, liver enzymes, etc.), adherence counseling, peer education, and access to support groups. In most ANC clinics in Kenya, HIV-positive pregnant women are normally referred to a separate HIV clinic for care and treatment (either located elsewhere on the grounds of the health facility or at another health facility), which may operate at different times and days than the ANC clinic.

As elsewhere, the Kenyan national guidelines for PMTCT–including recommendations for infant feeding, CD4 count monitoring, and HAART initiation–have been evolving over time. In August 2010 the Kenyan PMTCT guidelines were updated [25]. These guidelines recommended earlier initiation of HAART for a larger group of HIV-positive pregnant women (WHO clinical stage III or IV regardless of CD4 count OR WHO clinical stage I or II with CD4 count <350/mm3, compared with previous recommendations of WHO clinical stage of III or IV OR CD4 count <200/mm3) to benefit both the health of the mother and prevent HIV transmission to her child during pregnancy and breastfeeding. The ARV prophylaxis regimen guidelines were also changed in 2010, to start at 14 weeks or at first contact thereafter, compared with previous guidelines that recommended initiating prophylaxis at 28 weeks. The new PMTCT guidelines also include provision of ARV prophylaxis (zidovudine and lamivudine) to the mother for 7 days after the delivery and daily infant prophylaxis with nevirapine (NVP) monotherapy until one week after breastfeeding cessation if mother is not on HAART or up to 6 weeks of age if mother is on HAART.

Abstract
Background
Despite strong evidence for the effectiveness of anti-retroviral therapy for improving the health of women living with HIV and for the prevention of mother-to-child transmission (PMTCT), HIV persists as a major maternal and child health problem in sub-Saharan Africa. In most settings antenatal care (ANC) services and HIV treatment services are offered in separate clinics. Integrating these services may result in better uptake of services, reduction of the time to treatment initiation, better adherence, and reduction of stigma.

Methodology/Principal Findings
A prospective cluster randomized controlled trial design was used to evaluate the effects of integrating HIV treatment into ANC clinics at government health facilities in rural Kenya. Twelve facilities were randomized to provide either fully integrated services (ANC, PMTCT, and HIV treatment services all delivered in the ANC clinic) or non-integrated services (ANC clinics provided ANC and basic PMTCT services and referred clients to a separate HIV clinic for HIV treatment). During June 2009– March 2011, 1,172 HIV-positive pregnant women were enrolled in the study. The main study outcomes are rates of maternal enrollment in HIV care and treatment, infant HIV testing uptake, and HIV-free infant survival. Baseline results revealed that the intervention and control cohorts were similar with respect to socio-demographics, male partner HIV testing, sero-discordance of the couple, obstetric history, baseline CD4 count, and WHO Stage. Challenges faced while conducting this trial at low-resource rural health facilities included frequent staff turnover, stock-outs of essential supplies, transportation challenges, and changes in national guidelines.

Conclusions/Significance
This is the first randomized trial of ANC and HIV service integration to be conducted in rural Africa. It is expected that the study will provide critical evidence regarding the implementation and effectiveness of this service delivery strategy, with important implications for programs striving to eliminate vertical transmission of HIV and improve maternal health.