Kelley and Krannert business schools partner to help IU Health manage surge of COVID-19 patients

IU News  online

2020-04-27

Faculty at two of Indiana's leading business schools -- Indiana and Purdue universities -- are collaborating on a project with IU Health to help the health care provider manage the COVID-19 demand surge in their 16 hospitals across five regions of the state. The interdisciplinary team of professors at IU's Kelley School of Business and Purdue's Krannert School of Management has been working since March 23 to develop a predictive model of the resources required for an adequate response to the pandemic. It integrates disease prediction with a sophisticated patient flow workload model.

view more

An Operational Framework for the Adoption and Integration of New Diagnostic Tests into Emergency Department Workflow

2019 The gap between medical research on diagnostic testing and clinical workflow can lead to rejection of valuable medical research in a busy clinical environment due to increased workloads, or rejection of medical research in the lab that may be valuable in practice due to a misunderstanding of the system-level benefits of the new test. This has implications for research organizations, diagnostic test manufacturers, and hospital managers. To bridge this gap, we develop a Markov Decision Process (MDP) from which we create “adoption regions” that specify the combination of test characteristics medical research must achieve for the test to be feasible for adoption in practice. To address the curse of dimensionality from patient risk stratification, we develop a decomposition heuristic along with structural properties that shed light on which patients and when a new diagnostic test should be used. In a case study of a partner Emergency Department, we show that the conventional myopic medical criterion can lead to poor decision making in both research development and clinical practice. This myopic approach can lead to overvaluing or undervaluing new medical research. This mismatch is accentuated when a simple (current) policy is used to integrate research into the clinical environment compared with our MDP's policy – poor implementation of a new test can also lead to unnecessary rejection. Our framework provides easily interpretable guidelines for medical research development and clinical adoption decisions that can guide medical research as to which test characteristics to focus on to improve chances of adoption.

view more

Timing it Right: Balancing Inpatient Congestion versus Readmission Risk at Discharge

2019 When to discharge a patient plays an important role in hospital patient flow management as well as the quality of care and patient outcomes. In this work, we develop and implement a data-integrated decision support framework to aid hospitals in managing the delicate balance between readmission risk at discharge and ward congestion. We formulate a large-scale Markov Decision Process (MDP) that integrates a personalized readmission prediction model to dynamically prescribe both how many and which patients to discharge on each day. Due to patient heterogeneity and the fact that length-of-stay is not memoryless, the MDP suffers the curse of dimensionality. We leverage structural properties and analytical solutions for a special cost setting to transform the MDP into a univariate optimization; this leads to a novel, efficient dynamic heuristic. Further, for our decision framework to be implementable in practice, we build a unified prediction model that integrates several statistical methods and provides key inputs to the decision framework; existing off-the-shelf readmission prediction models alone could not adequately parametrize our decision support. Through extensive counterfactual analyses, we demonstrate the value of our discharge decision tool over our partner hospital's historical discharge behavior. We also obtain generalizable insights by applying the tool to a broad range of hospital types through a high-fidelity simulation. Lastly, we showcase an implementation of our tool at our partner hospital, to demonstrate broader applicability through our framework's "plug-and-play" design for integration with general hospital data-systems and workflows.

view more

Time for Accountability: Are Readmission Responsibility Windows Too Long?

2018 Hospital readmissions are burdensome and costly to healthcare systems. To incentivize the reduction of unnecessary readmissions, penalty programs and payment models have been created. These models typically define a period of time, a.k.a. an episode of care (EOC), within which hospitals are accountable for readmissions (e.g., 30-day readmission penalty and 90-day bundled payment EOC). This paper studies the policy-levels decisions in the design of EOC-based readmission reduction policies, considering three key policy levers: 1) EOC length, 2) readmission penalty, and 3) readmission avoidance subsidy. Our model provides a simple yet powerful condition for policy design, capturing the interplay between policy levers and hospital’s readmission reduction programs. We argue that EOC length may be the long-missing piece to the puzzle - it has a strong impact on the incentives but it has not received much attention in previous works. Specifically, EOC length has an exponential impact on the cost of readmission reduction programs, whereas the other two policy levers have linear effects. Our findings suggest that the 30-day and 90-day EOCs may be too long to incentivize readmission reduction for any but the low-risk patients, even under additional penalty or subsidy. This may explain the stalled readmission reduction after the implementation of the 30-day Medicare readmission penalty program. Though payers want long EOC to ensure hospitals cover more readmissions, long windows lead to smaller programs as hospitals "give up" on risky patients, whereas a shorter EOCs can encourage hospitals to expand readmission programs to include more and riskier patients.

view more

Surge Capacity deployment in Hospitals: Effectiveness of Response and Mitigation Strategies

2017 Major hospitals frequently lack adequate space to accommodate emergency patients. Managers can take actions to create surge capacity, an immediate additional supply of medical services to accommodate increased demand. We study operational strategies that improve surge capacity, and we identify how they can be deployed most effectively based on the characteristics of individual hospitals.Recent government regulations in the United States have increased pressure on hospitals to improve emergency preparedness. Specifically, hospitals must be able to show that they have taken adequate measures to manage surge capacity. We formulate an optimization model of early disposition actions that can be used to create surge capacity in a hospital. We analyze the model to understand its structural properties and compare two strategies to improve surge capacity: coordinated early discharge, which occurs during the response, and inpatient workload smoothing, which can help mitigate the need for response actions. We show analytically that without coordination, hospitals always act too conservatively in discharging patients to accommodate surge arrivals, and that smoothing the elective inpatient workload reduces the expected cost of surge response. In the numerical study, we find a utilization sweet spot in which smoothing is best at increasing surge capacity, and we show coordination increases the number of surges and number of early discharges, while smoothing mitigates these effects, making surges less frequent and less costly. Coordination is effective at increasing surge capacity for all types of hospitals, but when considering the holistic impact to the hospital, coordination and workload smoothing are often complementary strategies for improving surge response. Moreover, hospitals with sufficiently many electives and moderately high utilization should prioritize mitigation efforts when planning for emergencies.

view more

Dynamic Monitoring and Control of Irreversible Chronic Diseases with Application to Glaucoma

2016 To effectively manage chronic disease patients, clinicians must know (1) how to monitor each patient (i.e., when to schedule the next visit and which tests to take), and (2) how to control the disease (i.e., what levels of controllable risk factors will sufficiently slow progression). Our research addresses these questions simultaneously and provides the optimal solution to a novel linear quadratic Gaussian state space model. For the new objective of minimizing the relative change in state over time (i.e., disease progression), which is necessary for management of irreversible chronic diseases, we show that the classical two-way separation of estimation and control holds, thereby making a previously intractable problem solvable by decomposition into two separate, tractable problems while maintaining optimality. The resulting optimization is applied to the management of glaucoma. Based on data from two large randomized clinical trials, we validate our model and demonstrate how our decision support tool can provide actionable insights to the clinician caring for a patient with glaucoma. This methodology can be applied to a broad range of irreversible chronic diseases to optimally devise patient-specific monitoring and treatment plans.

view more