Informs Annual Meeting 2017

SC05

INFORMS Houston – 2017

4 - Product Line Design for Greener Products Iva Petrova Rashkova, Washington University in St Louis, Washington University in St Louis, St Louis, MO, 63130, United States, irashkova@wustl.edu, Lingxiu Dong, Weiqing Zhang The purpose of this project is to explain why some food manufacturers are adopting local sourcing while others position themselves as organic; while some abandon their traditional products while others expand their product line with the greener versions. We build a model which incorporates two quality valuations - traditional and green - and two market segments - low and high green valuation. We characterize explicitly how the interplay between green-valuation uncertainty, production cost and the size of the green segment affect the manufacturer’s optimal product line design. 320B HIV-Related Applications Sponsored: Health Applications Sponsored Session Chair: Evin U. Jacobson, CDC, Atlanta, GA, 30332, United States, wqm4@cdc.gov 1 - Allocation of HIV Prevention and Treatment Resources in the United States SC05 To better provide guidance on how HIV prevention resources in the United States could be focused to prevent the most cases of HIV, we explored the optimal allocation of a treatment and prevention budget. The optimal allocation resulted in more than a 40% decrease in new cases over 10 years compared to the estimated current allocation. Coordination of HIV prevention activities could substantially reduce HIV incidence. The optimal allocation is associated with lower overall treatment budget, freeing up additional resources for prevention. 2 - Impact of Enhanced HIV Care and Treatment and Delivering PrEP in the United States Nidhi Khurana, CDC, Atlanta, GA, United States, Emine Yaylali, Paul Farnham, Katherine Hicks, Benjamin T. Allaire, Evin Jacobson, Stephanie L.Sansom We used a dynamic, compartmental model of HIV transmission and disease progression to estimate the number of HIV infections prevented in the United States from 2016 to 2020 when pre-exposure prophylaxis (PrEP) was implemented for populations at high risk of acquiring HIV under either the current continuum of care or under an improved continuum that results in the achievement of national HIV prevention goals. 3 - Examining Businesses that Provide Products and Services to Gay and Transgender Individuals Living with HIV Guruprasad Gadgil, University of North Texas, 1115 Collier Street, #2, Denton, TX, 76201, United States, guruprasad.gadgil@unt.edu, Gayle Prybutok, Victor Prybutok This mixed method study examines the transference of impression management, the theory of planned behavior, and technology enabled anonymity to develop a model for a HIV-positive population in terms of individuals that run small online businesses that serve the gay and transgender population living with HIV. This work examines why individuals make their data available and why online businesses are motivated to provide practices, products, and services to this community. 320C Dynamic Decision Making with Application to Major Health Challenges Sponsored: Health Applications Sponsored Session Chair: Shan Liu, University of Washington, Seattle, WA, 98195-2650, United States, liushan@uw.edu 1 - Cost-effectiveness Analysis for Prognostic-based Depression Monitoring Ying Lin, University of Houston, Houston, TX, United States, ylin58@uh.edu, Shuai Huang, Shan Liu Inadequate monitoring has been identified as a major challenge in depression treatment follow-up. Prognostic-based monitoring which adaptively allocates monitoring resource to high-risk individuals is promising to improve efficiency and reduce cost. We developed a decision support framework to integrate SC06 Evin U. Jacobson, CDC, Atlanta, GA, United States, wqm4@cdc.gov, Katherine A.Hicks, Justin Carrico, Stephanie L.Sansom

individual prognostic, monitoring strategy design and cost-effectiveness analysis. We compared several prediction algorithms to monitor high-risk patients for major depression (rule-based, logistic regression, collaborative model, lazy learning, etc.) using a dataset of depression population and identified five cost- effective monitoring strategies. 2 - Dynamic Treatment Selection and Modification for Personalized Blood Pressure Therapy Margaret L.Brandeau, Stanford University, Management Science and Engineering, 475 Via Ortega, Stanford, CA, 94305-4026, United States, brandeau@stanford.edu Personalized treatment of disease - identifying which treatments work, for whom, and in what context - can significantly improve patient outcomes compared to treatment based on clinical guidelines. We developed and tested a generalizable computational strategy to help personalize treatment when multiple treatment options, benefits, and risks must be considered, using a Markov Decision Process model informed by high-quality meta-analytic data. We applied the strategy to the selection of blood pressure treatments for a simulated US population, and compared overall benefits and risks to those achieved from application of current blood pressure treatment guidelines. 3 - Adaptive Strategies for Personalized Treatment Selection of Depression in a Heterogeneous Population Jue Gong, University of Washington, Industrial and Systems Engineering, MEB B14 Box 352650, Seattle, WA, 98115, United States, gongjue@uw.edu, Shan Liu We established a framework for characterizing personal dynamics in disease progression and making a series of adaptive treatment decisions based on the estimated dynamics. The objective is to select between two treatments of depression to maximize health outcome. Depression progression is assumed to be a POMDP. An extension of the collaborative model and its solution algorithm are developed to estimate the model parameters. The optimal treatment policy is obtained by solving a rolling horizon POMDP. This research has wide applications in chronic disease treatment problems in the era of personalized medicine. 4 - Resource Allocation for Hepatitis C Elimination Qiushi Chen, Massachusetts General Hospital, 101 Merrimac St., 10th FL, Boston, MA, 02114, United States, qchen@mgh-ita.org, Turgay Ayer, Jagpreet Chhatwal The burden of hepatitis C virus (HCV) infection continues to rise globally. The advent of highly effective HCV treatment in the last few years enables effective epidemic control and even elimination of the disease. However, high drug cost and unawareness of infection are challenges for achieving this goal. In this study, we develop a hepatitis C transmission model, and identify the optimal allocation for HCV screening and treatment to achieve the disease control target at the minimum cost. We present the allocation policies in different health care settings and target population profiles. 5 - Dynamics of Drug Resistance: Optimal Control of an Infectious Disease Naveed Chehrazi, McCombs School of Business, Austin, TX, United States, chehrazi@utexas.edu We study the optimal treatment policy for an SIS-type infectious disease with drug resistance. We prove that the optimal policy is bang-bang with a single switching time and we find that the value function is not Lipshitz continuous. Using numerical analysis, we demonstrate that the optimal policy changes form when the disease transmission rate is a function of disease prevalence (e.g., as a result of social distancing). Our solution approach can be generalized to other control problems. 322A Modeling and Optimization in Healthcare Decision Applications Sponsored: Health Applications Sponsored Session Chair: Ruichen Sun, University of Pittsburgh, Pittsburgh, PA, 15261, United States, rus19@pitt.edu 1 - Screening for Breast Cancer: the Role of Supplemental Tests and Breast Density Information Burhaneddin Sandikci, University of Chicago, Booth School of Business, 5807 South Woodlawn Avenue, Chicago, IL, 60637, United States, burhan@chicagobooth.edu, Mucahit Cevik Mammography screening is the gold standard for breast cancer screening in the US, but it is known to be less accurate for women with dense breasts. Supplemental screening methods such as Magnetic Resonance Imaging (MRI) and ultrasound have been recently introduced to improve detection accuracy. We incorporate the breast density information using a partially observable Markov decision process model, and study the impact of supplemental tests in detecting breast cancer. SC07

68