Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

TD82

6 - A Continuous-time Markov Model for Estimating Readmission Risk for Hospital Inpatients

n TD82 Hyatt, Phoenix West Health Care IV Contributed Session

Xu Zhang, University of Maryland, College Park, MD, United States, Sean Barnes, Bruce Golden, Paul Smith

Research concerning hospital readmissions has mostly focused on regression models that include various factors that influence the likelihood of this unfortunate outcome. These models are useful in certain settings, but their performance in many cases is lacking, and the dynamics of how readmission risk changes over time is often ignored. Our goal is to develop a model for readmission risk over time—using a continuous-time Markov chain—beginning at the point of discharge. We derive point and interval estimators for readmission risk. Finally, we validate our derived estimators using simulation, and apply our methods to estimate readmission risk over time using discharge and readmission data. n TD83 Hyatt, Remington Practice- Health Care IV Contributed Session Chair: Ting Wu, Nanjing University, Department of Mathematics, No 22 Hankou Road, Nanjing, 210093, China 1 - Analyzing Alternative Doctor to Room Ratio in an Outpatient Clinic as a Means of Improving Overall Patient Access QIng Wen, University of Pittsburgh Department of Engineering, Pittsburgh, PA, United States, Louis Luangkesorn, Anna Svirsko, Jay Rajgopal Children’s Hospital of Pittsburgh hosts a number of specialty outpatient pediatrics clinics. Currently, the space available is believed to be a constraint in the overall capacity for outpatient appointments. One way to increase capacity is to reduce the ratio of rooms to doctors. However, that creates a risk that the doctor’s time will not be as productive due to waiting for patients and rooms to be prepped and cleaned. We develop a simulation to compare the effect of different room to doctor assignment patterns in doctor productivity and throughput. We also use this as part of a larger simulation that models the effect of increased capacity on the patient backlog for the clinic. 2 - A Discrete-event Simulation Model to Optimize Emergency Department Wait Times and Inpatient Unit Flow at an Academic Medical Center Gokhan Kirlik, University of Maryland Medical System, 920 Elkridge Landing Rd, Linthicum Heights, MD, 21090, United States, Bill Bame, Kenneth Wood, Warren DÆSouza A discrete-event simulation model is developed to improve the flow of patients from the emergency department (ED) through the inpatient setting at a tertiary/quaternary academic medical center. After comprehensive analysis of electronic medical records (EMR), the main inputs of the model, which include ED arrivals, patient flow in ED, direct admissions, hospital inpatient unit flows, length of stay, discharges and bed/room cleaning are obtained. Validation and verification of the simulation model against EMR confirmed that the model mimic the real-world observations accurately. This model will serve as a reliable test bed for hospital operations improvement initiatives. 3 - Accelerating Kidney Allocation: Simultaneous and Expiring Offers Michal Mankowski, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia, Sommer Gentry After the new Kidney Allocation System (KAS) began, the kidney discard rate increases, especially for marginal quality kidneys. Placing non-ideal organs quickly might increase utilization by decreasing discards. We simulated making simultaneously expiring offers to multiple centers, where every center must accept or decline within the same 1 hour. Simultaneously expiring offers would burden centers to evaluate offers that might otherwise never have come to their center. We also estimated additional workload caused by one new allocation scheme. 4 - Prioritization Between Boarding Patients and Patients Currently Waiting or Under Treatment in the Emergency Department Kim De Boeck, KU Leuven, Leuven, Belgium, Ra sa Carmen, Nico Vandaele The extra workload introduced by boarding patients is a major concern in emergency departments. Not in the least because this confronts the physicians with a challenging task; making a prioritization decision between boarding patients and patients currently under treatment in the emergency department. The main contribution of this paper is the examination of different control policies for the physicians when needy boarding patients are added to the analysis. Using discrete-event simulation, three static priority policies and one dynamic priority policy are evaluated on various performance measures.

Chair: Xu Zhang, University of Maryland, 4176 Campus Drive, William E. Kirwan Hall, College Park, MD, 20742, United States 1 - Multi-appointment Outpatient Scheduling for Cardiology Programs Lida Apergi, University of Maryland, 8223 Paint Branch Drive, 2243 AV Williams, ISR, College Park, MD, 20742, United States, John Baras, Bruce L. Golden, Kenneth Wood This research tackles the problem of multi-appointment scheduling of outpatients going through elective programs in cardiology. The patients have to visit the hospital multiple times in order to complete the required tests and treatments before the final procedure. A model is developed towards optimizing the allocation of the available resources to the patients. The objective is to minimize the number of visits that the patients have to make to the hospital. A stochastic integer programming formulation is developed to optimize the scheduling decisions over time. 2 - Resilience as a Measure of Preparedness for Pandemic Influenza Outbreaks Walter Alejandro Silva-Sotillo, University of South Florida, 4202 E. Fowler Ave, Tampa, FL, 33620, United States Since spring 2013, periodic emergence of avian influenza A(H7N9) virus in China has heightened concerns for a possible pandemic outbreak, though it is believed that the virus is not yet human-to-human transmittable. From a public health preparedness standpoint, it is essential to assess the possible impact of an Influenza pandemic and to measure resilience: the ability to recover after such potential pandemic. The aim of this research is to measure the level of resilience that a given population could face in case of a potential Influenza Outbreak happens. 3 - Partition-based Simulation Optimization Algorithms for Primary Care Panel Management David Desmond Linz, Graduate Student Researcher, University of Washington, Seattle, WA, 98105, United States, Zelda B. Zabinsky, Paul Fishman When designing primary care panel composition, decision-makers need to balance a number of concerns including patient load, travel times for patients, and accommodating the demand for virtual care. To model primary care systems, a discrete event simulation is developed. To best determine paneling policy, we explore the application of several new partition based algorithms that are designed to explore a high-dimensional problem domain to locate good policy solutions. We compare solutions generated by the optimizers in terms of policy impacts and comment on the relative effectiveness of the optimizers in higher dimensions. 4 - A Reliability-based Approach for Performance Optimization of Service Industries: An Application to Healthcare Systems Hossein Badri, Wayne State University, 4815 4th Street, Manufacturing Building, Detroit, MI, 48202, United States, Taha-Hossein Hejazi, Kai Yang This study aims to apply the Multi-Response Optimization (MRO) method for reliability-based performance optimization of healthcare systems. In this study, we consider all types of response variables so that strength features of the system are optimized against external stresses. The proposed approach in this research is based on the stress-strenth model, and stresses are assumed to be normally distributed. Also, we present the implementation results of the proposed approach to a hospital in Iran for reliability-based optimization of the bed capacity planning. The results indicate that the proposed approach is very powerful in tackling the inherent complexity of healthcare systems. 5 - Evaluating Appointment Postponement in Scheduling Patients at a Diagnostic Clinic Mahsa Kiani, Clemson University, Clemson, SC, 29631, United States, Burak Eksioglu, Tugce Isik In today’s healthcare system, the increase in requests for doctors combined with a shortage of physicians has led to challenges for clinics to give timely appointments to patients. In this study, we propose a postponable acceptance model for scheduling different priorities patients to decrease indirect waiting time of higher priority patients (the time between the day the patient requests an appointment and the appointment day). Higher priority patients receive appointment upon arrival. However, the decision regarding accepting and assigning an appointment to lower priority patients is postponed. By postponing these requests and waiting for more urgent patients the profit is increased.

378