Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

TA83

2 - Secure Decentralized Decisions in Consolidated Hospital Systems: Intelligent Agents and Blockchain Adrien Badr , Student, University of Oklahoma, Norman, OK, United States, Shima Mohebbi Shared decision making and its security have become very important in consolidated healthcare systems. We propose a universal decentralized Decision- Making architecture utilizing trained artificial intelligence for illness detection and Blockchain technology to address the compelling need of coordination among healthcare providers-patients and secure transmission of medical transcripts. Agent-based simulation and optimization will be used to simulate patients’ referral process in consolidated hospital systems. We demonstrate the feasibility of the proposed approach via extensive simulation analyses. 3 - The Impact of Implementing Full Capacity Protocol on the Operational Performance in an Emergency Department Lu Wang, PhD Candidate, University of Kansas, Lawrence, KS, 66046, United States, Mazhar Arikan, Suman Mallik Full capacity protocol (FCP) is a set of guidelines that coordinates the patients flow when the emergency department (ED) is overcrowded. Utilizing data from a large urban teaching hospital, we show that the operational performance of the study ED is improved when the FCP is triggered. We propose recommendations to further improve the operational outcomes under FCP. We study how the detailed crowding levels impact patients’ length of stay as well as sub-intervals (i.e. waiting time, in-service time and after-service time). 4 - Associations Between Hospital Readmission and Length of Stay with Post Acute Referral Following Coronary Artery Bypass Graft or Valve Replacement Ineen Sultana, Graduate Research Assistant, Texas A&M University, Industrial and System Engineering, College Station, TX, 77843, United States Hospital readmission and length of stay (LOS) are two key indicators of health care quality. Risk factors associated with hospital readmission and LOS were analyzed for Coronary Artery Bypass Graft and Valve Replacement patients. Results indicated non-uniform care across the U.S. census division and an inversely proportional relation between readmission and LOS. Hospital size, teaching facility, and comorbidity influence LOS, while demographic conditions and support infrastructure influence readmission. Although variation in PAC usage across different region indicates that higher LOS patients were subjected to PAC referral, no direct association of PAC type and readmission was observed. 5 - Global Budget Revenue Policy Impact on Emergency Department Efficiency in Maryland Ai Ren, PhD Student, University of Maryland-College Park, Robert H. Smith School of Business, 3330 Van Munching Hall, College We conduct an empirical analysis of the impact of Global Budget Revenue (GBR) implementation on Maryland Emergency Department (ED) efficiency measure ED1 from January 2014 to April 2016. We use public data from the Centers for Medicare & Medicaid Services and the Kaiser Family Foundation database. Our study uses Difference in Difference model with mixed effect. We find GBR has statistically significant negative impact on Maryland ED1 efficiency performance but not practically significant. 6 - Care-coordination: Gain-sharing Agreements in Bundled Payment Models Salar Ghamat, Lazaridis School of Business and Economics, 75 University Ave W, Waterloo, ON, N2L 3C5, Canada, Hubert Pun, Greg Zaric We study gain-sharing agreements, under a bundled payment model, between a hospital and a provider that may cooperate to achieve a desired cost and quality of care collectively. We characterize scenarios that lead to care-coordination between the two parties. We also provide evidence that a target price bundled payment model can be more effective in healthcare settings where the care providers have asymmetric capabilities. We further provide guidance, for the payer, regarding the impact of these models on the health outcome of patients and the total spending. Park, MD, 20740, United States, Bruce L. Golden, Edward Andrew Wasil, Margret V. Bjarnadottir

n TA83 Hyatt, Remington Practice- Health Care I Contributed Session

Chair: David C. Novak, University of Vermont, 310 Kalkin Hall, 55 Colchester Avenue, Burlington, VT, 05405-0157, United States 1 - Determining Effective and Practical Colonoscopy Screening and Surveillance Policies Fatih Safa Erenay, University of Waterloo, Waterloo, ON, Canada, Gizem Sultan Nemutlu, Oguzhan Alagoz Efficient colonoscopy screening and surveillance policies are desired to cost- effectively reduce colorectal cancer incidence and mortality. We developed a partially observable Markov decision process model to dynamically schedule colonoscopy operations based on accumulated risk of having advanced lesions and other risk-factors. However, clinicians may find such optimal policies too complex to adopt. Using a stochastic model, we derive simple yet effective dynamic colonoscopy policies which perform closely to the optimal policies and are easy-to-implement in clinical practice. 2 - A Bayesian Approach to Measuring Healthcare Intervention Impacts Douglas A. Popken, SVP Analytics, NextHealth Technnologies, Inc, 999 18th Street, Suite 2600, Denver, CO, 80202, United States Measuring the impact of healthcare interventions is challenging due to the random, infrequent, and highly variable nature of follow-on cost and utilization data. While a randomized controlled trial approach reduces measurement bias, standard statistical techniques based on t-tests may not adequately separate signal and noise. Bayesian methods have the advantage of utilizing prior knowledge as well as providing richer output metrics to inform decision makers, both automated and human-in-the loop. We describe our implementation and provide comparative statistics from rea-world experience. 3 - Economic Assessment of Policies to Prevent Childhood Lead Poisoning George Miller, Institute Fellow, Altarum, 3520 Green Court, Suite 300, Ann Arbor, MI, 48105, United States, Corwin Rhyan Lowering current levels of lead exposure for young children has the potential to improve health, increase earnings, save government expenditures, and reduce disparities. We describe an analysis that estimated the annual economic burden of childhood lead exposure in the U.S. at $84 billion dollars, including costs of reduced lifetime productivity, increased health care and social assistance spending, and premature mortality. The analysis also identified the potential for residential lead prevention and remediation policies to cut lead exposure levels for young children, resulting in significant long-term economic benefits and in many cases positive net societal returns. 4 - A Group Decision Making Approach for Risk Assessment in Distribution of Pharmaceutical Products Shahrzad Faghih Roohi, Postdoctoral fellow, Eindhoven University of Technology, Eindhoven, Netherlands, Alp Akcay, Yingqian Zhang, Eelco De Jong This paper presents a group decision making approach for risk assessment in distribution of pharmaceutical products. By using a new intuitionistic fuzzy hybrid TOPSIS approach, we prioritize the risks involved in pharmaceutical distribution. The aim is to overcome the current weaknesses of failure modes and effect analysis (FMEA) method by considering unequal weights of risk factors given by experts for different failure modes. A case study on air cargo distribution of pharmaceutical products is presented to demonstrate the potential applications of the proposed approach, and the merits are highlighted by comparing with the traditional FMEA. 5 - Managing Blood Inventory with Random Transfers David C. Novak, Associate Professor, Grossman School of Business, University of Vermont, 310 Kalkin Hall, 55 Colchester Avenue, Burlington, VT, 05405-0157, United States, Marilyn Lucas, Karti Puranam We introduce a method to solve the fixed lifetime perishable inventory problem with a lifetime of m > 2 periods, where there are two independent sources of supply. One source is blood deterministically ordered by the blood bank. The other source is blood that is randomly transferred to the blood bank from smaller, lower-usage hospitals in the health care network. We formulate a DP to solve the multi-period cost minimization problem and test our solution approach theoretically and empirically. We then present a novel heuristic approach for determining a near-optimal ordering policy for any perishable product where redistribution represents a second source of supply.

273