Presentation Number: E08-2; Presentation Title: Red Eye ownz! Author(s): Arielle Lasry; Stephanie L. Sansom; Katherine A. Hicks; Vladislav Uzunangelov; Abstract Content Background: The Division of HIV/AIDS Prevention (DHAP) at the Centers for Disease Control and Prevention has an annual budget of approximately $400 million for funding HIV prevention programs in the US. We demonstrate how resource allocation modeling can inform the optimal use of these funds and how this may benefit HIV prevention efforts. Method: The HIV resource allocation problem consists of choosing the amount to be invested in the interventions considered so the cumulative HIV incidence is minimized over a 5-year horizon, given a fixed budget. We address this problem by defining two models that interact and analyzing their results. First, an epidemic model, defined as a compartmental model, determines HIV epidemic projections given a specified allocation of resources to specific interventions and populations. Second, an optimization model, defined as a non-linear mathematical program, generates different allocation scenarios, supplies them to the epidemic model and determines the optimal scenario when the best outcome is reached. The optimization model is driven by the costs and outcomes of targeting the interventions to the at-risk populations. The at-risk population considered is structured into 15 population subgroups by gender, race/ethnicity and HIV transmission risk group. Risk groups include high-risk heterosexuals, men who have sex with men and injection drug users. The at-risk population is estimated at 21 million, representing 10% of the general population aged 13 to 64 years. Race/ethnicity is defined as black, Hispanic and all others. We consider HIV screening interventions, with or without partner referral services, and programs to reduce risk behaviors. These interventions are targeted to different subsets of at-risk persons by HIV risk group, HIV status, gender and race/ethnicity, and more broadly to the general US adult population. The resource allocation model considers a total of 85 intervention/target group combinations for funding. Results: The output of the model is the optimal funding scenario indicating the amounts to allocate to the 85 intervention/target group combinations considered, as well as the number of new infections in each population subgroup associated with this funding scenario. These results are compared to DHAP's actual allocation scenario enabling an understanding of the number of infections that could be averted by optimizing the allocation of funds and the cost per infection averted. The model supports what-if analysis capabilities, which can be used to help decision-makers understand the impact of trade-offs and deviations from the optimal funding scenario and evaluate the benefits of any additional funds made available to DHAP. Conclusion: This HIV resource allocation model provides valuable guidance to the rational allocation of funds. Incorporating future epidemic trends in the decision-making process for resource allocation enables an optimal selection of which populations and interventions could be targeted. Improving the use of funds by targeting the interventions and population subgroups of greatest return should lead to improved HIV outcomes.