Health economic modeling to optimize the HIV care continuum: methods and applications

Despite over $20B of annual federal funding directed towards domestic HIV efforts, 38,000 new cases were diagnosed in 2017 in the US. The recently announced “Ending the HIV Epidemic: A Plan for America” initiative set ambitious goals to reduce new HIV infections by 90% within 10 years. Achieving these ambitious goals necessitates a resource-intensive response consisting of targeted, context-specific combination implementation strategies. Economic models play a critical role in informing resource allocations for the care and prevention of HIV/AIDS, providing a unified framework to quantify the health and economic value of different strategies. A diversity of modelling designs and approaches exist, each requiring different forms of data. Input data are rarely known with certainty. This in turn might propagate into result uncertainty and lead to suboptimal decisions. However, presently there is a paucity of standardized guidelines on model structural design, evidence selection and methods to address decision uncertainty. The objective of this thesis is to provide methodological advances in decision-analytic modeling in HIV/AIDS, with a focus on model design, the quality of supporting evidence, calibration, validation and analysis of uncertainty. The design of a model and input data are two central factors in ensuring credible inferences. We executed a narrative review of a set of dynamic HIV transmission models to comprehensively synthesize and compare the structural design and the quality of evidence used to support each model parameter (Study 1). Model complexity and uncertainty surrounding its inputs can diminish our confidence in a model. We provided a comprehensive description of the calibration and validation of a dynamic HIV transmission model for six US cities with diverse microepidemics, detailing key methodological innovations and efforts to increase rigorousness in the process. The resulting projections will provide a basis for assessing the incremental value of further investments in HIV combination implementation strategies (Study 2). Value of information analysis quantifies the value of collecting more information to reduce decision uncertainty and helps guide efforts for future data collection. Using the developed HIV model, we performed probabilistic sensitivity analysis on the highest-valued combination strategies and applied metamodels to estimate the value of collecting additional information to eliminate decision uncertainty (Study 3). Findings of this study will make substantial methodological and public health contributions, providing implications for health decision-makers and scientists alike. This methodological approach can serve as a means of optimizing HIV strategies and be applied to diverse settings across North America and internationally.