Christine Kiruthu-Kamamia

This dissertation focuses on mobile health and differentiated HIV service delivery in Malawi. HIV remains a leading cause of morbidity in sub-Saharan Africa despite substantial progress in the scale-up of antiretroviral treatment and viral load suppression. In high-burden countries such as Malawi, the expansion of lifelong antiretroviral treatment has transformed the HIV epidemic by substantially decreasing mortality and increasing the number of people alive on ART, thus simultaneously increasing pressure on health systems to sustain long-term engagement in care across growing client populations. Differentiated service delivery has been adopted to address these pressures by tailoring the intensity of care, visit frequency, and service location to client needs, including for individuals newly initiating antiretroviral treatment and those who are clinically stable. However, as differentiated service delivery models expand, especially into decentralized and community settings, persistent gaps in routine monitoring and evaluation, documentation, and workload management increasingly constrain efficiency and long-term sustainability. Mobile health interventions have been proposed as tools to address these gaps by strengthening communication, data capture, and coordination of care. However, evidence on how mobile health performs when embedded in routine differentiated service delivery models and its implications for health system efficiency and sustainability in LMICs remains limited.

Chapter 1 details this background and outlines the aim of this dissertation, which is to examine whether and how integrating mobile health strengthens differentiated service delivery models for HIV care in a resource-constrained health system. Using Malawi as a case study and drawing on routine implementation within the Lighthouse Trust HIV clinic, the dissertation evaluates two mobile health interventions: an interactive two-way text messaging system for clients newly initiating antiretroviral treatment and a mobile, community-based electronic medical record system supporting antiretroviral treatment delivery for clinically stable clients. The research is guided by two overarching questions; 1) How do mobile health interventions influence efficiency and healthcare worker workload within differentiated service delivery, including implications for economic performance? and, 2) How do mobile health interventions affect client engagement in care and healthcare worker experiences when implemented as part of routine differentiated service delivery programs? Together, these questions examine whether mobile health strengthens differentiated service delivery by enabling efficiency gains and improving engagement in routine HIV care.

Furthermore, Chapter 1 situates the dissertation within the global and Malawian HIV epidemic, tracing the historical evolution of the response and the emergence of differentiated service delivery models. The chapter reviews evidence on differentiated service delivery and mobile health, identifying persistent gaps in monitoring and evaluation systems, documentation processes, workload measurement, and economic evaluation. It situates the research within the scale-up of differentiated service delivery as a response to expanding lifelong antiretroviral treatment cohorts, persistent health workforce constraints, and increasing pressure to deliver HIV care efficiently under constrained donor financing. The chapter further justifies choosing Lighthouse Trust as a case study, given its leadership in HIV service delivery and digital health integration in Malawi. Chapter 1 articulates the overarching study aim and research questions guiding the dissertation.

Chapters 2 and 3 jointly examine whether mobile health alters early engagement in HIV care among newly initiated clients and whether such changes can be justified economically under routine implementation conditions. Chapter 2 evaluates the effect of an interactive two-way text messaging (2wT) intervention on early antiretroviral treatment engagement using a randomized controlled trial conducted in a high-volume public HIV clinic in Lilongwe. The study compares visit adherence, retention, and viral load suppression outcomes between clients receiving two-way text messaging support and those receiving standard care. The findings show that the intervention improves early engagement in care, particularly through improved on-time attendance at scheduled visits during the initial treatment period. However, no statistically significant differences are observed in six-month retention or viral load suppression. These results indicate that interactive text-based communication can strengthen early engagement behaviors during a high-risk period for disengagement, but does not, on its own, produce short-term differences in clinical outcomes.

Chapter 3 extends this analysis by assessing the cost-effectiveness of the 2wT intervention relative to standard care from the healthcare provider’s perspective. The study compares costs and retention at 12 months post-ART initiation, as well as incremental cost-effectiveness ratios. The findings indicate that the intervention incurs modest additional costs, largely driven by personnel time required to manage client communication. However, in scenario analyses, the cost-effectiveness improves substantially as fixed costs are distributed across larger client populations. Together, Chapters 2 and 3 show that mobile health can improve early engagement processes without compromising clinical outcomes, while economic efficiency depends on implementation at a larger scale.

Chapters 4 and 5 shift the analytical focus from client behavior to provider workload and system efficiency within a community differentiated service delivery model. Chapter 4 examines the acceptability and feasibility of a mobile electronic medical record system, the Community-based ART Retention and Suppression (CARES) app, designed to support a nurse-led community-based antiretroviral treatment delivery for clinically stable clients. Using a rapid qualitative analysis, the study explores healthcare workers' experiences with the CARES application guided by the Technology Acceptance Model. Findings indicate high perceived usefulness and acceptability related to improved access to client records, reduced reliance on paper-based documentation, and enhanced continuity of care. Healthcare workers also identified feasibility challenges, including training demands, device management, system performance limitations, and incomplete integration with existing health information systems. Despite these implementation challenges, healthcare workers perceived the application as valuable for documentation and continuity of care.

Chapter 5 evaluates the cost–benefit of the CARES application from the payer perspective. Drawing on time–motion observations and costing estimates, the study quantifies changes in healthcare worker time allocation associated with digital documentation during community antiretroviral treatment delivery. The findings demonstrate substantial reductions in time spent on routine documentation and data management tasks, particularly among nurses. When monetized over a ten-year horizon, the benefits associated with reduced documentation time and improved data management exceeded system development and maintenance costs, resulting in a positive net present value. Collectively, Chapters 4 and 5 demonstrate that, over the long term and with mature implementation, mobile health can strengthen the delivery of community service delivery.

Chapter 6 is the final chapter of the dissertation and synthesizes findings across the four empirical studies to address the overarching research aim and research questions. Drawing on evidence from Chapters 2 through 5, the chapter integrates results across effectiveness, economic, workload, and qualitative analyses to explain whether and how mobile health strengthens differentiated service delivery models for HIV care in a resource-constrained setting. Chapter 6 synthesizes findings across the four studies and advances four cross-cutting statements that capture the dissertation’s main contributions. The first statement demonstrates that mobile health strengthens differentiated service delivery by likely reducing the routine monitoring and evaluation burden on healthcare workers, primarily by digitizing documentation and tracking processes within routine care. The second statement clarifies that potential efficiency gains from mobile health are not automatic but depend on alignment with existing health system capacity, workflows, and organizational structures. The third statement shows that, in Malawi’s donor-dependent HIV program, mobile health–enabled differentiated service delivery shows promise to support maintaining the quality of HIV service delivery by improving efficiency and resource use rather than by reducing costs or clinical inputs. The fourth statement situates the dissertation’s findings within the broader goal of HIV epidemic control, showing that mobile health–enabled differentiated service delivery supports achieving epidemic control by strengthening early engagement during antiretroviral treatment initiation and supporting continuity of care among clinically stable clients.