2016 INFORMS Annual Meeting Program

TB23

INFORMS Nashville – 2016

TB21 107A-MCC

we demonstrate the value of a third decision alternative, akin to a second opinion, that can be used when the gatekeeper lacks confidence to commit. The error reduction is particularly pronounced for more risk-averse gatekeepers, for customers with a high level of diagnostic uncertainty, and when the unit is busy. 2 - Learning From Many: Partner Diversity And Team Familiarity In Fluid Teams Jonas Jonasson, London Business School, jjonasson@london.edu, Zeynep Aksin Karaesmen, Sarang Deo, Kamalini Ramdas We use data from London Ambulance Service to study the impact of partner diversity of new paramedics on their operational performance. We find that the greater diversity in prior partners directly improves performance for an unstandardized process. For a more standardized process, this effect is moderated by a new recruit’s total experience. We explore the implications of our results for team formation strategies by balancing the benefits of partner diversity with those of team familiarity. 3 - Patient Portals In Primary Care: Impacts On Visit Frequency And Patient Health Hessam Bavafa, University of Wisconsin, hessam.bavafa@wisc.edu Interest in innovative healthcare delivery models has increased due to measures such as the Affordable Care Act, which is designed to expand insurance coverage and contain healthcare costs. One innovation that has been forwarded as a low- cost alternative to physician office visits is “e-visits,” or secure messaging between patients and physicians. We evaluate the effect of e-visit adoption on patient health and physician productivity using a panel dataset from a primary care provider in the US. 4 - Discharge Decision In Emergency Departments: Impact Of Operational Measures And Pay-for-Performance Incentives Eric Park, University of Hong Kong, Hong Kong, g, ericpark@hku.hk, Yichuan Ding We study how operational measures in the emergency department such as number of patients waiting to be seen and physician’s patient load affect patient discharge decisions. We also analyze the impact of a provincial hospital level pay- for-performance incentive scheme on discharge decisions. We study several major hospitals in the metro Vancouver, Canada area. TB23 108-MCC Models in Medical Decision Making Sponsored: Health Applications Sponsored Session Chair: Zlatana Dobrilova Nenova, University of Pittsburgh, 282 Mervis Hall, Pittsburgh, PA, 15260, United States, zdn3@pitt.edu 1 - Chronic Kidney Disease: A Simulation Study Zlatana Nenova, University of Pittsburgh, zdn3@pitt.edu, Jerrold H May We developed a case-based reasoning simulation model to predict the one-year disease progression of chronic kidney disease patients. The model bases its projections on an analysis of the patient’s historical lab values (eGFR, albumin, phosphate, and potassium) and vital signs (systolic and diastolic blood pressure, and weight), together with the history of disease comorbidities and complications (diabetes, heart failure, dialysis, PVD/CVD, and cirrhosis). 2 - Decision-making Models In Kidney Transplantation Eric Chow, Johns Hopkins University, Baltimore, MD, 21231, United States, echow8@jhmi.edu Clinical decision-making in kidney transplantation is a constant challenge for patients: will you benefit from a given organ being offered or are you better off waiting on dialysis for a better organ? The field is thus in need of mathematical models designed to help patients make these decisions. Fortunately, there are rich sources of big data in transplantation to support the design of these models, including a national registry of every patient on the waiting list, every organ offer made, and post-transplant outcomes. This presentation will review these data sources and their integration into several existing models. 3 - Challenges In Markov Modeling Of Cancer Treatment Jiaru Bai, University of California, Irvine, Irvine, CA, United States, jiarub@uci.edu, Cristina del Campo, L Robin Keller We present a way to build a Markov decision tree to model cancer progression and cost-effectiveness analysis for two or more cancer treatments. We propose several problems researchers can encounter in this kind of research and provide possible solutions.

Coordinated Care Delivery Sponsored: Health Applications Sponsored Session

Chair: Pooyan Kazemian, Harvard Medical School, 25 Shattuck Street, Boston, MA, 02115, United States, pooyan.kazemian@mgh.harvard.edu Co-Chair: Mark P Van Oyen, University of Michigan, 1221 Beal Ave, Ann Arbor, MI, 48109, United States, vanoyen@umich.edu 1 - The Impact Of Delay Announcements On Hospital Network Coordination Jing Dong, Northwestern University, jing.dong@northwestern.edu, Elad Yom-Tov We investigate the impact of delay announcements on the coordination within hospital networks using a combination of empirical observations and numerical experiments. We provide empirical evidence that patients may take delay information into account when choosing emergency service providers and that such information can help increase coordination in the network. We analyze different factors that may affect the level of coordination that can be achieved. In particular, we show that delay estimators that are based on historical average may cause extra oscillation in the system when patients are sensitive to delay. 2 - Proactive Inpatient Bed Allocation For Emergency Department Patients Using Predictive Analytics Seung Yup Lee, Wayne State University, Detroit, MI, United States, seung.lee@wayne.edu, Ratna Babu Chinnam, Evrim Dalkiran One of the main factors driving Emergency Department (ED) crowding is boarding delay, where admitted patients are held in ED while waiting for an inpatient bed to be identified and prepared. We propose a queueing network model that allows for the development of ‘proactive’ coordination strategies. In particular, under the proposed setting, the inpatient bed allocation process precedes ED patient disposition. Model also accounts for the performance of the predictive analytics model in predicting disposition decisions. We present analytical results and insights through experiments motivated by a large Midwest healthcare facility. 3 - Care Coordination Models Based On Longitudinal Encounter Data Michael Rossi, Univ of Massachusetts - Amherst, Amherst, MA, United States, mrossi09@gmail.com, Hari Balasubramanian We discuss a framework for analyzing data concerning healthcare encounters at the individual level. These encounters can be of various types - outpatient, emergency room, inpatient, pharmaceutical etc., each corresponding to one or more diagnoses. We provide examples where such data could be used and discuss the stochastic methods that are best suited for generating insights. 4 - Coordinating Clinic And Surgery Appointments To Meet Access Service Levels For Elective Surgery Pooyan Kazemian, Harvard Medical School, Boston, MA, United States, pooyan.kazemian@mgh.harvard.edu, Mustafa Y Sir, Mark P Van Oyen, David Larson, Kalyan Pasupathy Providing timely access to surgery is crucial for patients with high acuity diseases like cancer. We present a methodological framework to coordinate clinic and surgery appointments so that patient classes with different acuity levels can see a surgeon in the clinic and obtain surgery (if found to be needed) within a maximum wait time target. We evaluate six heuristic scheduling policies that exploit information on the need for surgery obtained from the clinic visit. Colorectal surgery at Mayo Clinic is discussed as a case study. Numerical results suggest dramatic improvements in access for urgent patients. TB22 107B-MCC Data-Driven Healthcare Operations Sponsored: Health Applications Sponsored Session Chair: Hessam Bavafa, University of Wisconsin, 4284C Grainger Hall, 975 University Ave., Madison, WI, 53706, United States, hessam.bavafa@wisc.edu 1 - Risk Aversion In Gatekeeping Systems: An Empirical Study Of Admission Errors In Emergency Departments Stefan Scholtes, University of Cambridge, Judge Business School, Cambridge, United Kingdom, s.scholtes@jbs.cam.ac.uk, Michael Freeman In a study of over 450,000 emergency department attendances, we explore the impact of gatekeeper risk-aversion and the level of diagnostic uncertainty on referral errors. While gatekeepers normally make binary decisions to refer or not,

271