Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

TC83

unique dataset from the NHS acute trusts in England to investigate our hypotheses. 3 - A Physicians and Medical Staffs Scheduling Problem in Hospitals with Multi-branches Wenjuan Fan, Hefei University of Technology, No. 193 Tunxi Road, Hefei, 230009, China This paper investigates the scheduling problem of physicians and medical staffs in large hospital with multi-branches. Each branch has its own medical staffs, while the physicians need to serve in all the branches affiliated to the hospital. The paper takes into account the demand and the available resources of the hospital, the workload of physicians and medical staffs, etc. as the constraints, and the objective is to minimize the dissatisfaction of physicians, and the cost of physicians. Then, a hybrid meta-heuristic algorithm SCA-VNS combining a Sine Cosine Algorithm (SCA) and variable neighborhood search (VNS) is proposed to solve this problem. 4 - Control of an Infectious Disease in a Metapopulation Ceyda Yaba, Clemson University, 110A Martin Circle, Central, SC, 29630, United States, Burak Eksioglu, Amin Khademi An infectious disease for which there is no cure can quickly spread in a meta- population with devastating consequences. Spreading of such a disease can be represented via a compartmental model. We modeled this problem as a Markov Decision Process, to control the spread of the infection by quarantining the infectious individuals. However, due to the curse of dimensionality, we solve the problem by an approximate dynamic programming (ADP) approach. We also compare the policy obtained from ADP with other benchmark heuristic policies, such as restless bandit, and one-step look-ahead policy. We then simulate our results for the 2014 Ebola epidemic in Sierra Leone. 5 - House Calls and Office Visits: A Primary Care Model for Aging People Using Multi-objective Approach Jennifer L. Mendoza-Alonzo, University of South Florida, Tampa, FL, 33613, United States, Jos Zayas-Castro In the US, the elderly population will nearly double in the next 20 years. The primary care delivery system needs to be enhanced to face the new challenges of an aging population. Aligned with the health care reforms, this creates a “perfect storm for the development of house call models. This study analyzes from a strategic level, three perspectives: organization, patient, and care worker, in a mixed model -house calls and office visits- for people 65 years and older, using multi-objective optimization and Nash bargaining solution of an integer- programming model. It is expected to generate evidence about the possible benefits of this sub-model under the emerged model: patient-centered medical home. 6 - Application of Mixture Model to Identify Risk Factors Corresponding to Heart Disease This study explores critical features impacting heart disease for males and females. It investigates whether features contributing to heart disease significantly vary for males and females and determines the critical risk factors for both genders. A medical dataset, Cleveland dataset (UCI. 2009) is used. The Mixture model technique such as the kernel mixture model is applied to classify sick males and females. To select the best mixture models, we use information criteria such as ICOMP. The best mixture model identifies different subpopulations of patients. We study whether these subpopulations are associated with gender and determine the critical risk factors corresponding to each gender. n TC83 Hyatt, Remington Practice- Health Care III Contributed Session Chair: Haolin Feng, Sun Yat-sen University, Lingnan College, Guangzhou, 510275, China 1 - Optimal Resource Allocation Policies in Nonprofit Organizations Faisal M.M. Alkhannan Alkaabneh, Cornell University, Ithaca, NY, 14850, United States, Siddhartha Banerjee, H. Oliver Gao We develop a framework for resource allocation, in the context of food banks operations, to help food banks optimize their share allocation policies by considering the nutrition needs of served population. To this end, we propose a convex programming model, a mathematical model whose solution provides a provable utility optimal policies. Through a novel mathematical model and the utilization of decomposition techniques, we arrive at two simpler sub-problems that can be solved efficiently. We prove that both sub-problems can be solved in polynomial time. Nooshin Hamidian, University of Tennessee-Knoxville, 851 Neyland Dr, Knoxville, TN, 37996, United States, Hamparsum Bozdogan, Rapinder Sawhney

n TC81 Hyatt, Phoenix East Optimization III Contributed Session Chair: Mohsen Mohammadi Dehcheshmeh, University of Louisville, 2301 S. 3rd St., Louisville, KY, 40292, United States 1 - Designing and Optimizing an Integrated Platelet Supply Chain Network Considering Transshipment Yuan Xu, PhD Candidate, North Dakota State University, 1045 17th Ave N. 161 Unit, Fargo, ND, 58102, United States, Joseph Szmerekovsky Given the perishable and lifesaving characteristics of blood, unpredictable and unbalanced supply and demand incur a lot of waste due to expiry. Transshipment can help organizations deal with demand variability and stock outs, which will lead to a more balanced supply chain system. In this study, a mixed integer programming model considering blood transshipment for an integrated platelet supply chain is developed to minimize total operational cost under centralized control. The developed model considers multiple sources of supply and stock age information. Effects of demand variation, age composition, and transshipment will be analyzed. 2 - Using Probability Dominance for Experimental Analysis of Algorithms Hesam Shams, University of Tennessee, 851 Neyland Drive, 525 John D. Tickle Engineering Building, Knoxville, TN, 37996, United States, Oleg Shylo We describe a framework for comparing optimization algorithms based on the concept of probability dominance. This approach provides a rigorous assessment of algorithms’ performance, which is intuitive and statistically sound. It can be widely used as a tool for algorithm design and development in the operations research. 3 - The Outcome Interval Problem Mohsen Mohammadi Dehcheshmeh, University of Louisville, 132 Eastern Pkwy, J.B. Speed School of Engineering, Louisville, KY, 40292, United States, Monica Gentili We introduce and study a new problem, namely the outcome interval problem, to quantify externalities in decision making under uncertainty. The outcome interval problem consists of determining the best and worst values of a given linear function (namely, the outcome function) over the optimal solutions of an interval linear optimization problem. We present some theoretical properties of the outcome interval problem and solve it heuristically and exactly. A comprehensive experiment is conducted to evaluate the performance of our methods and a real case study on healthcare access measurement is presented to show the importance of the problem for reliable decision making. n TC82 Hyatt, Phoenix West Health Care III Contributed Session Chair: Nooshin Hamidian, University of Tennessee-Knoxville, Knoxville, TN, 37920, United States 1 - Improve the Quality of Care After Discharge by Selecting a Referral Network of Skilled Nursing Facilities Yunsi Yang, Washington University in St. Louis, St. Louis, MO, 63130, United States, Fuqiang Zhang Since the implementation of the Affordable Care Act (ACA), hospitals become more incentive to improve the care patient receive after discharge. To improve the efficiency of discharge planning, hospitals are seeking to collaborate with high- quality Skilled Nursing Facilities (SNF) and narrow down referral list. Bed reservation program and Preferred Network are two schemes commonly used to create the referral network. Considering hospital facing with uncertainty from patient choice and availability of SNF beds, we provide an algorithm for hospital to select a portfolio of SNFs for these two schemes. 2 - Codified Knowledge Sharing and Operational Failures in Healthcare: Evidence From NHS Hospitals’ Risk Management Documents Mecit Can Emre Simsekler, Khalifa University of Science & Technology, Dept. of Industrial & Systems Engineering, Abu Dhabi, United Arab Emirates Mecit Can Emre Simsekler, UCL School of Management, London, United Kingdom, Bilal Gokpinar Considering two key components of knowledge sharing among healthcare personnel, (i) codified in the form of written documents and (ii) tacit with behaviors and daily practices, we examine how knowledge sharing capabilities in healthcare settings translate into risk management performance. We employ a

337