Addressing Barriers to Antiretroviral Adherence Among Postpartum Women Living with HIV in Rural South Africa: A Participatory Action Research Approach

The paper addresses persistent challenges in antiretroviral (ARV) adherence among people living with HIV (PLHIV) in South Africa, noting that most health service interventions are derived from health services research and prioritize supply-side or biomedical metrics (e.g., viral suppression, ARV uptake). The authors pose the question: What if PLHIV—specifically postpartum women, a group with high non-adherence—led the research and designed interventions based on their lived realities? With evidence that up to 50% of women are lost to HIV care by 6 months postpartum in South Africa, the study aims to test a collaborative, person-centered, participatory action research (PAR) approach using Photovoice to understand and address barriers to ARV adherence among postpartum women in a rural Eastern Cape municipality. The overarching purpose is to reframe success toward outcomes that matter to service users, not solely biomedical measures.

Background literature highlights high poverty rates among women—especially in the Eastern Cape—with widening gendered poverty gaps and structural disadvantages in employment and education. Studies link poverty (food insecurity, housing precarity, transport costs) to poor ARV adherence among women. Fear of severe side effects without food often leads to missed doses, and extreme scarcity can halt adherence entirely. Poverty also heightens vulnerability to intimate partner violence (IPV) and depression, further undermining adherence. Stigma—particularly anticipated stigma (fear of negative reactions)—is repeatedly implicated in both initiation and continuation of ART; it correlates with alcohol abuse, depression, decreased HIV knowledge, and non-disclosure. For mothers, anticipated stigma complicates disclosure within families and to children. Conversely, postpartum women possess strong motivators for adherence: belief in ARV efficacy (e.g., preventing vertical transmission) and the imperative to remain healthy to care for their children. The paper draws on calls for person-centered, community-based participatory research (Pantelic et al.) where PLHIV act as co-researchers and co-designers, with outcomes measured by what matters to them.

Design: Participatory Action Research (PAR) employing the Photovoice method. PAR elements included participation of those most affected, systematic research, action, and a goal of social change. Photovoice was chosen to enable participants to document and reflect on barriers and motivators via photographs. Ethics: Approved by the Biomedical Research Ethics Committee, University of KwaZulu-Natal (BREC BFC253/17). Informed consent obtained in isi-Xhosa. Emphasis on co-researcher ownership, welfare, confidentiality, and anonymization of data (photo codes; no identifiable faces). Setting: Ngqushwa Municipality, Amathole District, Eastern Cape, South Africa—rural, high poverty, limited infrastructure and employment, high HIV prevalence among pregnant women. Participants: Ten isi-Xhosa-speaking women living with HIV who had given birth within the previous 3 years and were diagnosed and initiated on ART during pregnancy. Identified via clinic records from four clinic catchment areas. From 37 eligible, 10 were randomly selected (one replacement after a withdrawal). All infants of co-researchers were HIV-negative. Research team: Ubunye Foundation staff (local field officer/CHW and programme coordinator/professional nurse) facilitated data collection and analysis. The lead author (Ubunye Programmes Manager) managed protocol, analysis (NVivo), and dissemination. Procedures: Preparatory training for staff covered study aims, Photovoice facilitation, and debrief procedures. Co-researchers received training on camera use, ethical photo-taking (no identifiable faces), and their role as researchers. Over 4 weeks, each co-researcher photographed motivators/barriers to ART adherence and participated in weekly debrief sessions with Ubunye staff. To protect confidentiality, debriefs and final data review were conducted individually rather than in groups. Co-researchers could exclude photos or content at any time. Data management and analysis: Audio debriefs were recorded, transcribed, and translated into English by external translators; checked by staff and co-researchers. Photos and transcripts were coded in NVivo to identify themes. Iterative validation occurred through individual reviews with co-researchers to confirm themes and interpretations. PAR Action component: Findings were disseminated through conferences and local health forums for advocacy. Co-researchers participated in designing a community-led programme (Lelethu) through Ubunye Foundation’s governance structure (Siyakhana) and a Programme Working Group (PWG). Steps included agreeing on key findings, co-designing interventions, and mobilizing resources/funding for implementation.

- Primary barriers to ARV adherence: anticipated stigma and poverty. - Anticipated stigma: Co-researchers described pervasive fear of status disclosure, leading to secrecy and daily vigilance. ARVs were seen as conspicuous markers of HIV, with multiple stigma “hot spots” across the treatment continuum: clinic pickup (visibility of HIV-designated rooms; staff handling/labeling; pill bottle noise); transport and public settings; home storage; and taking pills in the presence of others. Women reported avoiding doses to prevent exposure risk. Interim clinic measures (e.g., fast pickup points, alternative venues, discreet packaging) did not fully alleviate fears. - Poverty: Extreme material hardship (food insecurity, overcrowded/poor housing, unemployment) directly impeded adherence (e.g., avoiding pills on an empty stomach; lack of private space to take/store medication). Dependence on social grants and others for basic needs increased vulnerability to control and abuse. Economic precarity exacerbated isolation and constrained health-seeking behaviors. - Postpartum-specific dynamics: Three distinct features intensified the stigma–poverty–adherence nexus: (1) total child dependence heightened motivation to survive and remain healthy; (2) strong belief in ARV efficacy reinforced by experiences preventing infant infection; (3) profound guilt and fear of being judged as a “bad mother,” intensifying concealment, social isolation, and withdrawal from employment opportunities. Women constantly balanced the fear of not taking ARVs against the fear of being exposed by taking them. - Action and advocacy outcomes: Findings were presented at International AIDS Conference (2018), Durban AIDS Conference (2019), PHASA (2018), and local/district health programme meetings. The team won a 2018 Community Chest Impumelelo Social Innovation Award. - Programme development (Lelethu): Co-researchers led design and implementation of interrelated interventions: (1) individualized Life Plans with CHWs/Ubunye staff addressing family dynamics, disclosure, clinical access, and economic empowerment steps; (2) stigma reduction via women openly living with HIV engaging schools, clinics, and community organizations; (3) life skills and economic empowerment support (business development, job search, training/education access); (4) confidential peer support via WhatsApp groups with data/phones provided for information sharing, advocacy, and opportunities. - Early programme outcomes: Among 118 women aged 16–24 enrolled, 31 obtained jobs, 11 started businesses, and 14 pursued further study. Nine participants began community advocacy against HIV/TB stigma; one received commendation at a GBV seminar (French Embassy, Cape Town, 2022).

The study demonstrates that a person-centered PAR approach, with postpartum women as co-researchers, can generate nuanced, context-specific insights into adherence barriers and catalyze targeted community-led interventions. By reframing success toward improvements in quality of life (QOL)—economic stability, psychosocial support, reduced stigma—rather than narrowly on biomedical adherence metrics, the approach aligns interventions with what matters to PLHIV. An intersectionality lens clarifies how anticipated HIV stigma and gendered poverty interact to shape women’s concealment strategies, isolation, and constrained agency. Postpartum identity adds unique pressures (child dependence, moral judgments of motherhood, heightened fear of disclosure) that amplify adherence challenges. The PAR process—ensuring ownership, collaboration, and action—helped women articulate shared challenges, advocate locally and nationally, and co-design the Lelethu programme to address intersecting determinants. While biomedical measures were not the primary focus, improvements in QOL are expected to support sustained adherence indirectly. The findings support calls to mainstream person-centered, participatory models in HIV services and research.

Participatory research and community-led programme development effectively identified and addressed core barriers to ARV adherence among postpartum women living with HIV. Empowering women as co-researchers revealed anticipated stigma and poverty as central, intersecting barriers that also degrade overall well-being. The resulting Lelethu interventions target social, psychological, and economic determinants, prioritizing outcomes meaningful to participants. This work challenges reliance on adherence/viral load as sole success metrics and advocates for QOL-focused, person-centered evaluation. The model offers a replicable approach for NGOs, health services, and researchers to co-create sustainable, locally relevant solutions. Future work should scale and rigorously document PAR-based, community-led programmes, evaluate long-term impacts on both QOL and clinical outcomes, and adapt the model to diverse settings and populations.

- Small sample size (n=10) and focus on postpartum women limit generalizability. - Single rural municipality in Eastern Cape; findings may not transfer to other regions or urban contexts. - Confidentiality needs necessitated individual rather than group Photovoice sessions, potentially altering group dialogue dynamics typical of Photovoice. - Action evaluation is ongoing; early programme outcomes are preliminary and lack controlled comparison. - Authorship and dissemination constraints (co-researchers declined named publication) may limit external validation of participatory contributions.