Baseline characteristics of study sites and women enrolled in a three-arm cluster randomized controlled trial: PMTCT uptake and retention (pure) Malawi

Malawi introduced a new strategy that aims at preventing HIV infection from mother to child. There is no cure for HIV, but medication (antiretroviral treatment) is available to slow the progression of the virus and help prevent transmission of HIV to others. The so called “Option B+” strategy means that all HIV-infected pregnant and breastfeeding women start antiretroviral treatment as soon as they are diagnosed. The PMTCT Uptake and REtention (PURE) study evaluated whether support provided by peers (mothers who are HIV-infected and on treatment themselves) can help to increase the number of HIV-infected pregnant and breastfeeding women to start treatment (uptake) and keep them in care and on treatment (retention). The study took place in 21 health facilities, where mothers either received no peer support (7 sites), peer support in the facility (7 sites) or peer support in the community (7 sites). This article describes how the study was successfully set up and managed by 3 organizations. Between November 2013 and November 2014, a total of 1269 HIV-infected women were included in the study in the 21 health facilities. We describe that it is possible to set up a complex study across 3 geographical zones in Malawi.

In 2011, Malawi introduced an ambitious public health strategy known as “Option B+”, providing a standardized lifelong combination antiretroviral therapy (ART) regimen to all HIV-infected pregnant and breastfeeding women, irrespective of their CD4 cell count or World Health Organisation (WHO) clinical stage of HIV infection. This public health approach for the prevention of mother to child transmission (PMTCT) of HIV has since been adopted by several other countries in the region and was included in the updated WHO guidelines in 2013 [1]. Initial data have demonstrated that the Option B+ strategy has resulted in large increases in the number of pregnant women accessing PMTCT [2]. However, concerns have been raised about the level of attrition of women from care [3]. Although the strategy offers an attractive rapid ART scale-up and has the potential to profoundly impact maternal and infant outcomes, services supporting the implementation of Option B + have not been formally evaluated. Several issues were identified at the national level as potential threats to the successful implementation and scale up of the strategy: a) potential suboptimal uptake of ART by asymptomatic pregnant women due to low treatment literacy and stigma; b) low adherence to ART and poor follow-up of HIV exposed infants; and c) lack of psycho-social support for long term retention in this relatively asymptomatic patient population [4]. Hence, it seemed apparent that the identification of interventions to support PMTCT uptake and retention would be the first step towards improvement. A recently published systematic review identified 9 completed studies and 5 ongoing trials which examined initiation of ART in pregnant women [5]. While the authors report several promising interventions for improving ART initiation, the quality of evidence was insufficient to support recommendations. In addition, results for ART initiation in pregnant women were not independently examined, and maternal retention in PMTCT care and exposed infants care were not assessed.

Although randomized controlled trials are promoted as the highest level of evidence for the evaluation of interventions, the implementation of rigorous trials within the context of the public health sector are complex and therefore still uncommon [6]. PURE Malawi embraced the challenge of implementing a multi-site controlled trial within the context of public health clinics under real world conditions in Malawi. PURE Malawi was a three-arm cluster randomized controlled trial that aimed at evaluating whether facility- or community-based peer-support would enhance care-seeking and retention in care by HIV-infected pregnant and breastfeeding women, their HIV-exposed infants, and their male sex partners, and ultimately improve health outcomes in all three populations. The hypothesis was that enhanced support for women and their families within facilities and/or community would result in improved retention in the continuum of PMTCT care. The primary trial outcome is the proportion of women retained in care, 2 years after they initiated antiretroviral therapy under the Option B+ strategy. This paper describes the site selection, recruitment, enrollment process and baseline characteristics of study sites and women enrolled in the trial.

PURE Malawi was a three-arm stratified cluster randomized controlled trial that compared facility-based and community-based models of peer support to standard of care under Option B+. The three arms differed with respect to the presence and location of one-on-one patient education and support, support groups, visit reminders, and missed visit follow-up. The trial was unblinded to study participants and health care providers, but blinded to outcome assessors. A comprehensive description of the study arms was reported elsewhere [7]. The PURE trial was managed by the PURE Malawi consortium comprising of governmental, non-governmental, research, and academic partners. The consortium was led by Lighthouse Trust. Dignitas International, The University of Malawi College of Medicine Malaria Alert Centre, and University of North Carolina Project were each responsible for trial implementation in one Malawian zone. The trial was conducted in the South East, South West, and Central West zones of Malawi, the zones where the implementing partners operate. Selection and responsibilities of all consortium members were described elsewhere [7].

PURE Malawi sites included 2 district hospitals, 2 mission hospitals, 2 rural hospitals, 13 health centers and 1 private clinic. Table 1 shows PURE Malawi sites by health zone, implementing partner and study arm. Each implementing partner managed the implementation of the study across the 7 sites in the respective health zone where the partners operate. The organization Mothers2Mothers provided additional support for training and implementation in all Arm 2 sites. Table 2 shows the number of Ministry of Health staff trained by site and cadre prior to study implementation. Enrollment initiation varied per trial site from November 2013 through January 2014. Recruitment was completed in November 2014. Median enrollment time per site was 31 weeks, ranging from 17 to 51 weeks. Table 3 describes the number of pregnant and lactating women enrolled per site and arm, in addition to the number of eligible women reported in cohort reports at ANC and ART during the recruitment period. Each site and arm reached or surpassed the minimum sample. However, a few participants (4%) were enrolled later than the desired end date of July 31st 2014, making 2 years of follow up impossible by July 31, 2016, the trial end date.

A total of 1269 women were enrolled, 1094 (86%) pregnant and 175 (14%) lactating women.

When comparing number of enrolled pregnant women per site with facility cohort data; 7 sites recruited >100% of the total reported cohort of HIV-infected pregnant women at ANC in the corresponding months. The additional number of pregnant women may have started ART, but not attended ANC in the specific study site. Nine sites recruited >90%, 10 sites 60–90%, and 2 sites 40%–59% of the total reported cohort of pregnant women at ANC in the corresponding months. When comparing the number of women enrolled versus ANC reports, sites recruited a median of 90% (IQR 75–100) of eligible reported women from ANC.

Following quarterly ART cohort report data; in all study sites, a total of 1934 (81%) pregnant and 454 (19%) lactating women had started ART during corresponding quarters. This proportion of pregnant and lactating women was comparable to our study sample.

This paper shows that the PURE consortium was successful in recruiting a complex cohort of pregnant and lactating women across 3 geographical zones in Malawi. Each site and trial arm reached or surpassed the minimum required sample size, and the majority of sites recruited over 90% of the reported eligible women attending ANC during the recruitment period. Although cluster designs are often avoided due to the challenges of enrolling at multiple sites within a reasonable timeline, we show that, despite the absence of remunerated study staff within study site, only few participants were enrolled later than the desired end date; allowing for 96% of the cohort to be followed for at least 2 years. We report that there were HIV test kit stock outs or unavailability of the HTC providers, which may have prolonged the enrolment period. Shortages and stock out of commodities at health facilities is a reality in Malawi, however, national guidelines are set up to improve the supply management [8]. In the majority of participating sites, the ratio of pregnant and lactating women enrolled in the study was similar to the ratio of pregnant and lactating women initiated on ART in the same sites. There was no difference in age between the enrolled pregnant and lactating women. One fifth of the women enrolled were young mothers (15–21), which is consistent with national evaluations that describe women in Malawi’s Option B+ program [3]. Proportionally more lactating women reported being divorced, widowed or having no sexual partner. As male partner support is thought to be an important determinant for uptake and outcomes of PMTCT initiatives [9‐11], follow up of our cohort of mothers with and without partners will add to the evidence on the impact of male partner involvement approaches on PMTCT services. According to Malawi’s demographic health survey, 14% of Malawian women are married to a man with more than one wife. Polygamy is most common in the Northern Region [12], which may explain why we found a lower proportion of women in polygamous marriage (9%). In our study, 98% of pregnant and lactating women were reported clinically asymptomatic. This high proportion of women starting ART in the earliest stages is an indication of the success of Malawi’s PMTCT programming.

Malawi, as the first country to implement Option B+, has the potential to be a leader in developing and documenting highly effective models of service delivery to pregnant and lactating women and enhance retention in lifelong care and treatment. Cluster randomized controlled trials are needed to determine whether promising interventions are indeed effective. However, the implementation of such trials pose a substantial challenge, especially across multiple regions, different facility types, and several implementing partners. Thus, the ability to select 21 comparable clinics, stratify them in a meaningful way, and have each enroll its target sample size is an achievement. This achievement stems from clear leadership, small incentives to clinical staff, and well-coordinated teams.

This paper described the processes for site selection, recruitment, and enrollment of study participants, along with baseline characteristics of study sites and women enrolled in the trial. Follow-up of participants was completed on July 31st, 2016. The final results and outcomes (per protocol) of this study are reported elsewhere [13]. Alongside this trial, a qualitative study has been implemented to explore experiences of patients and health care workers in relation to the implementation of Option B+ and the trial interventions [14].

Malawi, as the first country to implement Option B+, has the potential to be a leader in developing and documenting highly effective models of service delivery to provide needed support to women and enhance retention in lifelong HIV care and treatment approaches. The PURE trial is expected to play a critical role in this process. We show that a consortium of organizations was successful in recruiting a complex cohort of pregnant and lactating women across 3 geographical zones in Malawi within a reasonable timeline.