Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WA63

3 - Do Hospital Closures Improve the Efficiency and Quality of Other Hospitals? Lina Song, Harvard University, Soroush Saghafian We study the impact of hospital closures on the surrounding hospitals’ efficiency and the mechanisms through which the changes occur. We also investigate the implications of hospital closures on quality. We do these by examining the efficiency, bed utilization, service duration, patient experience, readmissions, and mortality using a nationally representative panel data of Medicare patients. We find that the closure of a hospital in a market results in improvement in efficiency at the remaining hospitals, but this happens at an expense of reducing the service duration. Furthermore, hospital closures are associated with an increase in 30-day mortality of the surrounding hospitals. 4 - Can Public Reporting Cure Healthcare? The Role of Quality Transparency in Improving Patient-provider Alignment Soroush Saghafian, Harvard Univeristy, Kennedy School of Government, 79 John F. Kennedy Street, Cambridge, MA, 02138, United States, Wallace J. Hopp Public reporting of medical treatment outcomes is being widely adopted by policymakers in an effort to increase quality transparency and improve alignment between patient choices and provider capabilities. We examine the soundness of this approach by studying the effects of quality transparency on patient choices, hospital investments, societal outcomes (e.g., patients’ social welfare and inequality), and the healthcare market structure (e.g., medical or geographical specialization). Our results offer insights into why previous public reporting efforts have been less than fully successful and suggest ways in which future efforts can be more effective. n WA61 West Bldg 102C Supply Chain Management VI Contributed Session Chair: Laharish Guntuka, University of Maryland-College Park, College Park, MD, 20740, United States 1 - A Multi Echelon Supply Chain Model with Multiple Customer Classes Mohammad Najjartabar Bisheh, Kansas State University, 2061 Rathbone Hall, 1701D Platt St., Manhattan, KS, 66506, United States, Behnam Malmir, Aram Bahrini, Gholamreza Nasiri In this paper, an integrated model of location-allocation with inventory control decisions is proposed. This model highlights that customers have different value to the central company. The model is single-product and multi-period with stochastic demands. Also, warehouses were considered to have some capacity constraints. The method was solved on a small scale and in the real size as well. For small instances, GMAS software used and for large instances, two meta- heuristic algorithms of classic genetic and hybrid of genetic-simulated annealing were employed. Finally, the quality of solutions and running time of algorithms was analyzed drawing upon the model parameters. 2 - A Contingency Theory Approach to Mitigating Counterfeiting Risk in the Supply Chain John F. Kros, Vincent K. McMahon Distinguished Professor, East Carolina University, College of Business Dept of M&SCM, 3205 Harold Bate Building, Greenville, NC, 27858-4353, United States, Scott Dellana, Jason Rowe, Mauro Falasca Counterfeiting is an emerging issue supply chain managers confront. A contingency theory perspective is developed to model the impact of counterfeit prevention efforts on SC performance. A structural equation model examines the relationships among five SC constructs. Our findings suggest that firms with greater SC risk management integration have stronger orientation toward counterfeit risk, greater maturity in counterfeit risk mitigation, more consistent metrics, & better performance outcomes. Interestingly, counterfeit orientation alone was not found to significantly improve metric consistency. 3 - Electric Cars Propagation with Subsidized Supply Chain Amir Naderpour, PhD Student & EGTA, University of Texas at Arlington, 701 S. West Street, Box 19437, Arlington, TX, 76019, United States when government’s goal is to propagate electric cars in the society, and has the potential to give the same subsidy to the supplier or the retailer of a product, which one should be subsidized to encourage higher quantity of that product? We answered this question in a couple of ways. The first way was game theory. And the theory predicted this supplier retailer relation. To answer our research question, we also used experiment study to see if the results from experiment is consistent with the theory prediction.

4 - Collaborative Emission Targets Joining Dincer Konur, Texas State University, San Marcos, TX, United States

This study analyzes a two-echelon channel under a Stackelberg setting. Particularly, the leader of the channel determines the quantity flow along the channel. Both agents have environmental targets that should be respected. We investigate collaboration mechanisms between the leader and the follower for joining their environmental targets. Two mechanisms are discussed in detail. A comparison between non-collaborative and collaborative emission targets joining decisions is presented. 5 - The Impact of Supply Chain Disruptions on Competitors: Propagation of Disruption Impacts Through Supply Chain Laharish Guntuka, Doctoral Student, University of Maryland- College Park, 4335 Rowalt Drive, #301, College Park, MD, 20740, United States, Adams Steven We investigate spillover effect of supply chain disruptions on competitor firms who themselves are not involved in the disruption. Measured through abnormal returns and return on assets, the competitors gain when a firm announces a major disruption but the gain is watered down by the number of shared suppliers, vertical relatedness, and supplier exposure, but exacerbated by event exposure. n WA63 West Bldg 103B Joint Session DM/Practice Curated: Data Science for Forecasting and Economic Modeling Sponsored: Data Mining Sponsored Session Chair: Kai Yang, Wayne State University, 4815 Fourth Street, Detroit, MI, 48201, United States 1 - Heteroskedasticity-based Instrumental Variables for Endogeneity Treatment Bernardo F. Quiroga, Assistant Professor, Pontificia Universidad Católica de Chile, Av. Vicuña Mackenna 4860, Comuna de Macul, Santiago, 29601, Chile We present a way to achieve econometric identification in instances when outside instruments are not available. By exploiting the structure of the variance- covariance matrix under heteroskedasticity, it is plausible to build orthogonality conditions in presence of endogeneity in otherwise unidentified systems of linear regression equations. Our results build upon the findings presented by different authors in the internal instruments literature (e.g., Arellano and Bond (1991); Lewbel (1996, 2013)) 2 - Recent Progresses in Continuous-time Contract Theory Dylan Possama, Assistant Professor, Columbia University, 500W 120th Street, Mudd 308, New York, NY, 10027, United States This talk will consists in an overview of recent progresses made in contracting theory, using the so-called dynamic programming approach. The basic situation is that of a principal wanting to hire an Agent to do a task on his behalf, and who has to be properly incenticized. We will show how this general framework allows to treat volatility control problems arising for instance in delegated portfolio management, in electricity pricing, or in central clearing houses. We will also, if time permits, analyze the situation of a Principal hiring a finite number of Agents who can interact with each other, as well as the associated mean-field problem. 3 - How to Project Outpatient Appointments Utilization Fangzheng Yuan, Doctoral Candidate, UGPTI, 1340 Administration Avenue, Fargo, ND, 58108, United States, Joseph Szmerekovsky, Vera Tilson In this paper, a probability model known as “shifted-beta-geometric model is implemented as an alternative to commonly used regression models to project the outpatient appointments utilization. This model is easy to use and can be implemented using a simple Excel spreadsheet and the result shows a great accuracy of forecasts and diagnostics for appointment utilization. 4 - A Comparative Study for Patient Workload Prediction Kai Yang, Wayne State University, 4815 Fourth Street, Detroit, MI, 48201, United States, Mohammad Hessam Olya This paper suggests a framework for patient workload prediction by using patients’ data from VA facilities across the US. To capture the information of patients with similar attributes and make the prediction more accurate, a heuristic cluster-based algorithm for single-task learning is developed in this research. . In this research, we have considered patient-dependent and facility-dependent attributes and the relation between tasks into the model while implementing Multi-Task Learning (MTL) approach and training multiple related tasks simultaneously.

433