BACKGROUND: Despite the advances and expansion of HIV testing and treatment programs, children and adolescents in sub-Saharan Africa (SSA) are being left behind. In 2021, more than 46% of children living with HIV were not receiving treatment; among SSA adolescents (aged 10-19) AIDS is a leading cause of death. HIV testing is the first and crucial step necessary in accessing HIV treatment and reducing HIV related mortality, yet current testing strategies have not been sufficient enough to address and overcome the gaps in HIV testing and diagnosis of children and adolescents. Compounded by the fact that HIV case identification will only become more difficult and costly with time, as the remaining undiagnosed population of children and adolescents will require more concerted efforts compared to standard of care, alternative HIV testing strategies are needed. Index-linked HIV testing (ILHIVT) – offering HIV testing to children and adolescents living with an HIV positive parent or guardian – might be an effective method to test and identify HIV positive children and adolescents. As such, the aim of this PhD was to evaluate and generate methods around the economic evaluation (EE) of HIV testing strategies in children and adolescents in Zimbabwe. METHODS: This thesis integrates methods from health economics and mathematical modelling of infectious diseases to both determine and strengthen the ways in which cost-effectiveness analyses of HIV testing strategies focused on the adolescent sub-population in SSA can be conducted. Firstly, a systematic review to identify the modelling methods of EEs of HIV testing strategies in SSA over the past decade was conducted. To assess the quality of model reporting, a multi-component, novel tool based on a gold standard checklist and recommendations was developed. Then, using the findings of the systematic review as a starting point, a narrative review synthesizing, comparing and investigating child and adolescent HIV transmission dynamics representation in popular dynamic mathematical models applicable to the SSA context was conducted. Next, a mixed-methods – both bottom-up and top-down – provider based cost analysis of cost per test and per HIV diagnosis for 2- 18 year olds, through standard of care, along with the incremental cost of ILHIVT via three modalities (facility, home-based and caregiver assisted), in both urban and rural settings was conducted. Finally, using the costs generated via the costing analysis, a proposal for an EE of ILHIVT for children and adolescents in Zimbabwe, through a microsimulation model, (with static, stochastic and individual properties), was generated. RESULTS: The results generated can be classified according to this PhD’s objectives. Objective 1 sought to determine how EEs of HIV testing strategies in SSA were modelled: the majority of model based EEs exhibited dynamic, stochastic and individual properties. Transparency around model-based decisions made were severely lacking, with model reporting across criteria generated via the novel tool, highly mixed. No EE focusing on HIV testing strategies in the sub-population of children and adolescents had been conducted. Objective 2 sought to narratively describe child and adolescent inclusion and representation among high-reviewed and frequently utilized dynamic mathematical model of HIV transmission, pertinent to the Zimbabwean context and SSA as a whole. Half of models meeting the inclusion criteria incorporated child HIV transmission dynamics to some degree within their frameworks, while all models incorporated adolescent HIV transmission dynamics. Each model had room for improvement with regards to child and adolescent integration into the framework. The largest limitations involved omission of children entirely, to poor sexual mixing structuring and partnership formation description amongst adolescents. Objective 3 sought to measure the cost of delivering ILHIVT to children and adolescents in Zimbabwe. In the urban setting, home-based ILHIVT for 2-18 year olds had the lowest incremental cost (US$6.69), while facility-based ILHIVT (US$5.36) was the lowest option in the rural setting. Irrespective of setting, caregiver-assisted testing was always the most expensive option (urban U$17.49, rural US$62.49). Unit costs of ILHIVT was driven by uptake which varied according to both setting and modality. Objective 4 sought to explore practical applications of thesis findings. As this version of ILHIVT had low yield and high incremental cost-per diagnosis, it was unlikely to be a cost-effective option for identifying HIV positive children and adolescents. An EE proposal for ILHIVT, demonstrating the level of transparency needed around 13 modelling features and criteria, was put forth as an example of ‘good modelling practice’ for future researchers. CONCLUSIONS: The strength and reliability of EE findings are dependent on the validity of the underlying model used to answer a specific decision problem. Moving forward, transparency around all model-based decisions is needed to facilitate understanding, generalizability, reusability and reproducibility of models and results. Existing well-described dynamic mathematical models of HIV, used frequently in the context of strategic population-level decision-making, require improvement and expansion within their existing framework to accurately evaluate the impact and cost-effectiveness of interventions targeting this sub-group. Familiarity and acceptability drove uptake and choice of ILHIVT testing modality, highlighting that current messaging and knowledge around pediatric HIV, even amongst HIV positive indexes is insufficient. As a result, the current iteration of ILHIVT is low yielding and has high cost per diagnosis, translating to low cost-effectiveness. Alternative solutions which increase efficiency and reduce unit costs, such as testing the entire household of an index, thereby diffusing personnel costs, identified as the largest driver of unit costs, need to be explored. The findings of this PhD contribute to advancing the field of economic evaluation methods for HIV testing strategies in children and adolescents by addressing: 1.) the lack of practical representation of children and adolescents in decision analytic modelling of HIV testing strategies through a behaviorally driven analysis of model structures both within EEs of HIV testing strategies, and highly regarded and frequently used dynamic mathematical models of HIV; 2.) the lack of EE literature focused on HIV testing strategies targeting children and adolescents through a partial EE of ILHIVT.