Basu, AnirbanSaldarriaga, Enrique M2022-07-142022-07-142022Saldarriaga_washington_0250E_24263.pdfhttp://hdl.handle.net/1773/48733Thesis (Ph.D.)--University of Washington, 2022Local estimates can procure means to achieve a more equitable progress and end the HIV epidemic in the United States within the next decade. The aim of this dissertation was to quantify the incremental costs and health benefits of using prevalence data at the zip code level to inform resources allocation within Atlanta, Georgia, compared to the current distribution based on supply-side criteria.This study was structured in three aims. First, I conducted a simulation-based calibration of an HIV-mathematical model to project the epidemic at the zip code level under varying circumstances. Second, the CDC reports diagnosed HIV cases by zip code, but undiagnosed cases are unknown, which I predicted based on social determinants of HIV spreading and prevalence estimates at the county level. Third, I designed a cost-effective analysis (CEA) to quantify the health and economic consequences up to 2040 of allocating resources across zip codes under three alternatives: status quo, reallocation proportional to diagnosed-only- and to total-cases. For each scenario I estimated the incremental cost-effectiveness ratio (ICER), as the cost per quality-adjusted life year (QALY) gained, compared to the status quo. The CEA showed high variability across zip codes depending upon the direction of reallocation. Compared to the status, the reallocation alternative based on diagnoses-only were dominant among increased-coverage zip codes with costs savings of $13.8Million and 1,026 additional QALYs. Conversely, among decreased-coverage zip codes, the alternative was $3Million more expensive and yielded 2,019 less QALYs. The results under total-cases reallocation were the same across coverage-increased and -decreased zip codes and remarkably similar in the incremental effects. This study provides evidence that the health production function is heterogeneous across zip codes, making the reallocation of resources a non-zero-sum game. This implies the existence of a reallocation algorithm that maximizes the generation of QALYs while minimizing additional costs. These results create opportunities for prioritization of resources at the local level. While Atlanta provides an excellent setting to highlight the benefits of resources reallocation, several other cities have high variability of HIV spreading at the zip code level and presence residential segregation. Therefore, my analytical framework, methodology, and findings could be of interest in other cities and states across the country.application/pdfen-USCC BY-SAcost-effectiveness analysiseconomic efficiencyHIV epidemiclocal dataresource allocationUnited StatesEconomicsHealth sciencesEpidemiologyPharmaceutical sciencesThesis