![Show Menu](styles/mobile-menu.png)
![Page Background](./../common/page-substrates/page0464.png)
INFORMS Nashville – 2016
462
WD21
107A-MCC
Predictive Modeling for Healthcare Applications
Sponsored: Health Applications
Sponsored Session
Chair: Ozgur M Araz, University of Nebraska-Lincoln, CBA 260 1240 R
Street, P.O. Box 880491, Lincoln, NE, 68588-0491, United States,
oaraz2@unl.edu1 - Predicting Hospital Admissions From Emergency Department
Ozgur M Araz, University of Nebraska-Lincoln,
oaraz2@unl.eduEmergency departments (EDs) are critical for healthcare services delivery and
coordination between EDs and inpatient units are essential for higher quality of
service in hospitals. In this study, we are investigating the predictive factors of
hospital admissions from the emergency department (ED) in order to inform
resource capacity planning for the ED boarding process. We used ED visits data
which include demographic and administrative variables from a major hospital in
Omaha Metro area and performed analyses using several predictive models, e.g.,
logistic regression and artificial neural network, to predict admissions from the
ED. The predictive accuracy of these models are discussed.
2 - Seasonal Forecasting For Infectious Disease From
Multiple Data Sources
Zeynep Ertem, University of Texas at Austin,
zeynepsertem@gmail.com,Lauren Meyers, Kai Liu,
Ravi Srinivasan
Epidemics of contagious diseases may cause widespread loss in terms of mortality,
morbidity, and economic burden. There have been several studies about
forecasting focusing on one data source. However, there can be other more data
sources that are correlated with an important data source. I will show a problem
formulation of forecasting a target data source related to an epidemic using
multiple other data sources. I will also present preliminary results for forecasting
when no data is available for the target data source in the current season.
Furthermore, I will relax this assumption and show results for forecasting when
partial data is available for the target data source in the current season.
3 - Psa Screening For Prostate Cancer: A Dynamic Feedback Model
To Understand Long Term Trends In Population Screening
Ozge Karanfil, Massachusetts Institute of Technology, Sloan School
of Management, Cambridge, MA, United States,
karanfil@mit.edu,Hazhir Rahmandad, Jack Homer, John D Sterman
Practice guidelines for routine screening have changed significantly over time and
often not followed, with significant over-screening for some tests and under-
screening for others. In this study we develop a behaviorally realistic simulation
model to explore reasons of this phenomenon. The model is firmly grounded in
empirical evidence through collection of quantitative and qualitative data. Our
formal theory includes a decision theoretic core around costs and benefits,
cognitive and social feedbacks. The model can be used as a guide to understand
future effects of policy scenarios and be tested in a systematic fashion to find ways
to overcome the policy resistance seen in population screening.
4 - A Queueing Model For Nurse Staffing In Critical Care Outreach
Team And Intensive Care Unit
Ali Haji Vahabzadeh, The University of Auckland Business School,
Auckland, New Zealand,
a.vahabzadeh@auckland.ac.nz,Valery Pavlov
We propose a queueing model of CCOT to examine the effects of this team on the
ICU performance and patient outcomes. To gain more insights into the
effectiveness of the role of critical care nurses on the ICU utilisation rate and
patient outcomes, we analyse different nurse allocation policies between ICU and
CCOT. To validate the proposed queueing model, a discrete-event simulation and
accordingly an optimisation study have been performed. Finally, the study
provides recommendations to hospitals on the functionality of the CCOT and the
nurse staffing policy.
WD22
107B-MCC
Modeling Organ Allocation System
Sponsored: Health Applications
Sponsored Session
Chair: Naoru Koizumi, George Mason University, 3351 N. Fairfax
Drive, MS#3B1, Arlington, VA, 22201, United States,
nkoizumi@gmu.edu1 - Mathematical Optimization And Simulation Analyses For Optimal
Liver Allocation Boundaries
Monica Gentili, University of Louisville, Louisville, KY, 40205,
United States,
monica.gentili@louisville.edu, Naoru Koizumi,
Rajesh Ganesan, Chun-Hung Chen
This study combines mathematical programming models and Discrete Event
Simulation to advance existing research on organ allocation system and
geographic equity and efficiency in liver transplantation system. The main
objectives of the study are: (i) to identify key factors determining geographic
disparity in kidney transplantation; (ii) to identify optimal locations for both
existing and new liver transplant centers (iii) to identify new OPO boundaries and
(iv) to test whether the mathematically produced liver allocation system can
perform better than the actual system. We will show the results of our combined
approach when applied to liver transplantation in USA.
2 - Small Representations Of Big Kidney Exchange Graphs
John Dickerson, Carnegie Mellon University, Pittsburgh, PA,
United States,
dickerson@cs.cmu.edu, John Dickerson,
University of Maryland, College Park, MD, United States,
dickerson@cs.cmu.edu, Aleksandr Mark Kazachkov,
Ariel Procaccia, Tuomas W Sandholm
Kidney exchanges are organized markets where patients swap willing but
incompatible donors. We observe that if the kidney exchange compatibility graph
can be encoded by a constant number of patient and donor attributes,
fundamental problems in kidney exchange are solvable in polynomial time. We
give conditions for losslessly shrinking the representation of an arbitrary
compatibility graph. Then, using data from the UNOS nationwide kidney
exchange, we show how many attributes are needed to encode real compatibility
graphs. The experiments show that, indeed, small numbers of attributes suffice.
This has application to optimal pre-transplant immunosuppression policies.
3 - Offer Batching For Organ Placement
Tinglong Dai, Assistant Professor, Johns Hopkins University,
100 International Drive, Baltimore, MD, 21202, United States,
dai@jhu.edu, Sommer Gentry, Sommer Gentry, Sridhar R Tayur,
David Axelrod, Dorry Segev, Dorry Segev
In this study, we consider an organ procurement organization’s problem of
determining the optimal batch size of simultaneous offers made to transplantation
centers. We model the strategic interaction among transplant centers both within
and across batches, leading to structural properties and computational insights.
4 - Redistricting Liver Allocation: Challenges And Extensions
Sommer Gentry, United States Naval Academy,
gentry@usna.eduSommer Gentry, Johns Hopkins University, Baltimore, MD, United
States,
gentry@usna.edu, Josh Pyke, Eric K Chow, Dorry Segev
Livers for transplant in the U.S. are distributed within eleven regions, and are
much more available in some geographic areas, leading to dramatic disparities in
transplant rates. We have used redistricting to design novel sharing districts which
significantly reduce these geographic disparities. The improved districts might be
implemented soon, if concerns about increased transport time for organs and
about the variability of organ supply and demand can be addressed. We will
explore the efficient frontier of the policy space, trading off organ transport for
disparity reduction. We will also discuss a robust formulation of the redistricting
problem.
WD23
108-MCC
Socially-responsible Healthcare Operations
Sponsored: Health Applications
Sponsored Session
Chair: Priyank Arora, Georgia Institute of Technology,
800 West Peachtree, NW, Atlanta, GA, 30308, United States,
priyank.arora@scheller.gatech.edu1 - Ambulance Routing In Resource Constrained Settings
Milind G. Sohoni, Indian School of Business,
milind_sohoni@isb.edu, Lavanya Marla, Achal Bassamboo,
Chandrasekhar Manchiraju
Using real-world data we look at optimal dispatch policies and compare those
with current best practices. We develop insights and guidelines for practicing
managers.
2 - Healthcare Payment Model Impact On Hospital Readmissions
Jon M Stauffer, Texas A&M University, College Station, TX,
United States,
jstauffer@mays.tamu.edu,Jonathan Helm,
Kurt M Bretthauer
We examine the transition from Fee-for-Service (FFS) to pay-for-performance
(P4P) reimbursement plans, such as bundled payments and the Hospital
Readmission Reduction Program (HRRP). We use a game theory approach to
understand how healthcare providers interact to improve their individual
contribution margins. Results show that P4P plans do motivate extra readmission
reduction effort, but that misalignments can occur between the player’s efforts
and the minimum total system cost effort. We find that the smaller post-discharge
player can be over-motivated to reduce readmissions and that HRRP is not
necessary with well-designed bundled payment plans.
WD21