Title : A participatory system dynamics approach to modelling implementation science-informed HIV drug resistance Interventions in low- and middle-income countries

Background: HIV drug resistance (HIVDR) threatens progress toward sustained viral suppression and long-term treatment effectiveness in sub-Saharan Africa. In Tanzania, challenges related to delayed viral load monitoring, treatment switching, health system constraints, and behavioral factors interact dynamically to influence HIVDR outcomes. Traditional linear approaches may inadequately capture the complexity of these interacting determinants. System dynamics offers a means of understanding feedback mechanisms and simulating how changes within the system may influence HIVDR outcomes over time. On the other hand, implementation science complements systems-based analyses by identifying and designing contextually appropriate, feasible, and scalable implementation strategies that can facilitate the adoption, integration, and sustainability of interventions within real-world health system settings. Therefore, this study applies a participatory system dynamics approach to support the design and evaluation of implementation science-informed interventions for HIVDR management in Tanzania.

Methods: This study builds on prior participatory system dynamics work conducted in the Dar es Salaam Urban Cohort Study (DUCS) area, Tanzania, which identified key drivers of HIV drug resistance (HIVDR) and potential leverage points for intervention. The 5-Step Framework for Participatory Systems Modeling(5FPSM) was used to guide the process. Two stakeholder workshops involving HIV clients, clinicians, researchers, policymakers, implementing partners and community representatives were conducted. The first workshop focused on causal loop diagram validation and leverage point identification, while the second focused on quantitative model validation and intervention co-design and prioritization. Using the Behavior Change Wheel and implementation science principles, intervention implementation strategies were developed to target priority leverage points. A quantitative system dynamics model was developed in Stella Architect using a susceptible–infected backbone linked to HIV care cascade, viral load monitoring, treatment switching, prevention, quality of care, and health system resource modules. The model was calibrated and historically validated using DUCS data and national estimates and simulated from 2016–2030 to evaluate the potential effects of individual and combined intervention strategies on HIVDR outcomes over time.

Results: Preliminary findings identified three priority leverage points for reducing HIVDR risk: household income, stigmatization, and quality of care. Based on these leverage points, six implementation science-informed intervention strategies were co-designed: public education through places of worship and local government structures; health care worker training on treatment failure management and respectful client care; linkage of PLHIV in need to food support services; linkage of PLHIV in low-concentration facilities to peer support groups; stakeholder involvement in council-level decision-making; and decentralization of regimen switch decisions from national to regional level. These strategies have been translated into intervention scenarios for evaluation within the system dynamics model, which represents feedback mechanisms linking adherence, stigma, quality of care, viral load monitoring, treatment switching, and HIVDR outcomes over time. Scenario analyses will estimate the potential effects of individual and combined intervention strategies on HIVDR outcomes from 2016–2030.

Conclusion: Participatory system dynamics offers an approach for integrating stakeholder perspectives, implementation science-informed intervention design, and complex health system interactions in HIVDR policy analysis. The resulting model provides a platform for prospectively evaluating intervention strategies and informing evidence-based decision-making for HIVDR management in resource-constrained settings.

Beatrice Aiko is a pharmacist, assistant lecturer at Muhimbili University of Health and Allied Sciences (MUHAS), and PhD researcher at the Access-To-Medicines research center, KU Leuven. Her research focuses on design and evaluation of interventions for HIV drug resistance in Tanzania using systems thinking, implementation science, and participatory transdisciplinary approaches. Her broader interests include antimicrobial resistance, health system strengthening, implementation science, and integrating systems approaches into global and One Health-related health challenges.