2016 INFORMS Annual Meeting Program

WD26

INFORMS Nashville – 2016

WD26 110B-MCC Information Systems III Contributed Session

2 - A Near-term Mortality Indicator For Terminal Cancer Patients Using High Frequency Medical Data Donald Lee, Yale University, New Haven, CT, United States, donald.lee@yale.edu, Edieal J Pinker Although end of life cancer care accounts for a significant portion of the costs of all cancer care, it often fails to improve quality or quantity of life. Yet, oncologists consistently overestimate the chances of survival, leading to suboptimal delays in the transition to hospice care. To more accurately assess mortality risk, we develop a bedside statistical tool that utilizes high frequency EMR data to predict near term mortality. Projected savings from averting excessive curative interventions suggest significant inefficiencies in current practice. Joint work with the Smilow Cancer Centre. 3 - A Machine Learning Approach For Personalized Health Care Outcome Analysis Guihua Wang, Ross School of Business, University of Michigan, Ann Arbor, MI, United States, guihuaw@umich.edu, Jun Li, Wallace J Hopp Using a patient-level data set across 35 hospitals for cardiovascular surgeries in New York, we first provide empirical evidence that the quality gaps between hospitals are heterogeneous for subgroups of patients. We then use a machine learning approach to identify subgroups of patients that have large or small quality gaps. After that, we estimate the quality gaps between a hospital and the state average for each subgroup of patients to derive patient-centric information. Lastly, we show that providing patient-centric information not only helps patients choose providers but also helps hospitals identify areas for improvement and payers design cost-effective payment programs. 4 - Are Patients Patient? The Effect Of Universal Healthcare On Emergency Department Visits Diwas KC, Emory University, diwas.kc@emory.edu We study a natural policy experiment to examine the impact of universal healthcare on emergency departments. We find that the policy has differential effects on hospitals. Compared to EDs with high pre-policy workloads, we observe a relative increase in volume at EDs with lower levels of pre-policy workload. WD28 201B-MCC Product Strategies and Channel Structure in Supply Chain Management Sponsored: Manufacturing & Service Oper Mgmt Sponsored Session Chair: Laurens G Debo, Dartmouth College, Hanover, NH, United States, laurens.g.debo@tuck.dartmouth.edu Co-Chair: Cuihong Li, University of Connecticut, Storrs, CT, United States, cuihong.li@uconn.edu 1 - Salesforce Incentives For Managing Product Returns Rashmi Sharma, Pennsylvania State University, University Park, PA, 16802, United States, rashmi.sharma@psu.edu, Aydin Alptekinoglu We study a setting where product sales and returns are effort-dependent and the selling activity is conducted by a salesforce. We investigate the effect of salesforce behavior on net sales and compare different incentive schemes to identify optimal incentive strategies. 2 - Returns Policies For Overstock And Consumer Returns In Distribution Channels This paper studies the interactions between an upstream manufacturer and competing downstream retailers on offering returns policies to retailers and end- consumers respectively. The effects of channel structure, demand uncertainty, and retail competition are studied in a unified model covering both overstock returns and consumer returns. 3 - Retail Market Power: A Supplier-side Perspective Meng Li, University of Massachusetts, Amherst, MA, United States, meng.li@umass.edu, Yunchuan Liu

Chair: Benjamin Schooley, Assistant Professor, University of South Carolina, 1301 Gervais Street, Columbia, SC, 29201, United States, schooley.ben@gmail.com 1 - Digital Ecosystem Competition, To Open Or To Close? Chao Ding, Assistant Professor, KK Leung 807, The University of Hong Kong, Pok Fu Lam, Hong Kong, chao.ding@hku.hk IT firms are currently striving to build their own digital empire of ecosystem that profit from sales of both hardware/system and content/service. This study sets under the context of duopoly ecosystems and examines their content/service offering strategies and pricing strategies. 2 - A Quasi Experiment Using Social Network Approach To Effects Of Trust Building In Sharing Economy Shivom Aggarwal, Instututo de Empresa S.L., Instututo de Empresa S.L., IE Business School, Calle de Maria de Molina, 12 Bajo, Madrid, 28006, Spain, dr.shivom@gmail.com, Gautam Ray Trust is inherent in the success of Sharing economy business models, but how building trust affects such success poses a crucial issue for firms competing with sharing economy business models. We analysed a longitudinal data of bike sharing systems across several cities using a social network approach. We found that trust helped in increasing the overall revenues of a given station, but the average revenue per bike decreased significantly. The study discusses the implications and remedial possibilities based on the results. 3 - The Effect Of Ambidextrous Developers On Open Source Project Success Orcun Temizkan, Ozyegin University, Istanbul, Turkey, orcun.temizkan@ozyegin.edu.tr, Ram Kumar Open Source Software (OSS) development is an important, yet poorly understood type of software development with high project failure rates. We analyze a sample of real OSS projects to better understand project success. We argue that there are different types of developers and study their effects on success. We argue that developers can be classified into non-ambidextrous and different types of ambidextrous developers. Our results illustrate direct and interaction effects that ambidextrous developers have on project success and the importance of having multiple types of ambidextrous developers in projects. 4 - Effects Of Health Information Exchange On The Process Of Care, Patient Care Delivery, Administrative Productivity, And Population Health Outcomes Benjamin Schooley, Assistant Professor, University of South Carolina, 1301 Gervais Street, Columbia, SC, 29201, United States, schooley.ben@gmail.com, Sue Feldman, Neset Hikmet Health information exchange (HIE) is expected to transform the U.S. healthcare system through access to patient data from electronic health records across organizational boundaries - for the purpose of supporting care provision to improve care quality and population health. However, relatively few of the more than 100 operational U.S. HIE’s have been the subject of published evaluations. We surveyed Virginia healthcare providers to assess perceived value of HIE, including impacts on the process of care, patient care delivery, information quality, and population health outcomes. Perspectives differ based on organizational, medical specialty, and prior experience characteristics. WD27 201A-MCC Empirical Healthcare Operations Sponsored: Manufacturing & Service Oper Mgmt Sponsored Session Chair: Diwas S KC, Emory University, Atlanta, GA, United States, diwas.kc@emory.edu 1 - The Effects Of Discrete Workshifts On Non-terminating Queues Robert Batt, University of Wisconsin, bob.batt@wisc.edu, Diwas S KC, Bradley R Staats, Brian W Patterson While an emergency department is a non-terminating queue, it is staffed by care providers working discrete workshifts. We examine how productivity changes over the workshift. Further, because care providers work fixed-duration shifts, they sometimes must hand off care of in-process patients to another provider. We examine the impact these handoffs have on treatment time and revisit rate. Using simulation, we show that policies that prohibit starting new patients near the end of the shift can lead to improved system throughput.

Shuya Yin, University of California at Irvine, Irvine, CA, United States, shuya.yin@uci.edu, Yuhong He, Saibal Ray

Does a supplier prefer that its downstream retailers are more balanced in terms of their market power or that one of them is dominant, especially when the market power might be correlated to their bargaining powers? In this paper we address this issue and establish the supplier’s preference for a monopoly value chain as well as when there is value chain competition.

464