Despite the notable scale up of antiretroviral therapy (ART) coverage among pregnant and postpartum women living with HIV (PWLWH) globally, reductions in new HIV infections in children have plateaued and significant numbers of PWLWH are still not able to achieve optimal health outcomes. In sub-Saharan Africa, emerging evidence indicates some vulnerable sub-groups of PWLWH are at higher risk of poor ART adherence, disengagement from HIV care, and unsuppressed viral loads. In order to eliminate mother-to-child transmission and protect maternal health, it is critical to identify these PWLWH at risk of suboptimal outcomes early and provide effective interventions that address key challenges. There is significant evidence that psychosocial stressors—HIV-related stigma, intimate partner violence, and lack of ability to disclose HIV status—are major drivers of treatment lapses in this population. Problem Management Plus (PM+) is a lay health worker-delivered intervention that has high potential to address psychosocial stressors and promote treatment engagement for high risk PWLWH. PM+ combines problem-solving therapy and behavioral strategies to address mental distress and has been adapted and implemented in other populations in Kenya. In the current study, we propose to develop and validate a multivariable prediction tool (i.e., “risk calculator”) using socio-demographic, clinical, and psychosocial data from our recent longitudinal studies with large samples of PWLWH in southwestern Kenya, the same setting where our current study is proposed (Aim 1). We will establish risk prioritization scores (critical, moderate, minor) that predict the combined outcome of a missed visit (> 30 days) or treatment failure (elevated viral load >400 copies/ml). Next, we will utilize a human centered design (HCD) approach to work collaboratively with PWLWH and key stakeholders to tailor modifiable components of the PM+ intervention to address challenges to achieving optimal HIV treatment outcomes in this population (Aim 2). We will collaborate with PWLWH and lay health workers to develop both mobile phone (mHealth) and in-person PM+ delivery formats to be able to deliver the intervention safely and efficiently to a mobile population. Finally, we will pilot our tailored intervention in a hybrid type 2 effectiveness-implementation trial in which pregnant women identified as critical or moderate risk are randomized to one of three study arms 1:1:1 standard of care, in-person PM+, or mHealth PM+ and followed through 6 months postpartum (N=120). Feasibility, acceptability, intervention satisfaction, and preliminary effects on mental health and HIV outcomes will be assessed. Findings from this research will prepare us for a large-scale trial in which we intend to use an adaptive design to test which interventions (including adapted PM+), combination of interventions, and sequence of interventions are the most effective and efficient to support the most vulnerable PWLWH in low-r...