Ask the Experts: Matt Bailey on health care operations, decision making and future entrepreneurs
November 10, 2011
LEWISBURG, Pa. — Matt Bailey, associate professor of management, discusses health care operations, decision making and future entrepreneurs.|| Related Bucknell news story
Q: Your research focuses on health care operations. What does this research entail, and how can it help hospitals, doctors and patients?
A: Health care operations research is about examining processes in, for example, a hospital, and looking at how efficient a system is in placing patients where they need to be when they need to be there. It is about reducing waste and eliminating steps, such as nurses rechecking orders or patients waiting for test results or being stuck in the emergency department because there are not enough available beds. Operations research is not about dictating how clinicians should practice; it is about setting up a system where resources are used as efficiently as possible. Operations improvements have actually been tied to health improvements. If you have a patient waiting six or seven hours to move from the emergency department to a regular bed, that can result in a longer stay for the patient at the hospital, which is tied to adverse health outcomes.
A lot of people think the easiest solution for making a company or hospital more efficient is to add more resources, but that is usually not the case. You need to look at the system and resources you have and determine why those resources are being used inefficiently. In my research, I create mathematical and computer models to show the benefits of making the system more efficient and determine what needs to change.
Q: You are working with Geisinger Medical Center in Danville, Pa., to build a computer simulation that tracks patient arrival patterns and transfer rates. What are you hoping to learn?
A: Right now, we are in the stage of verification, so we are making sure the data from our model closely matches what is actually occurring in the hospital. What we want to find through our analysis is the percentage of patients the hospital would need to expedite, in terms of reducing the amount of time they are waiting to be discharged, to make the hospital more efficient and improve patient care. These are patients that are medically cleared to be discharged but, due to hospital inefficiencies or poor communication with the patient and their families, the patients stay in the hospital longer than necessary. The delay of a single discharge can have cascading effects on hospital operations such as increasing delays in patient transfers and the wait for available beds throughout the hospital. We intend to analytically determine that impact with our simulation model. Beyond the model, in practice, it takes a significant amount of effort to properly coordinate the resources and expedite the discharge of a patient. If we can illustrate and quantify the operational gains achieved when expediting the discharge of a small population of patients, this will be a significant contribution to the academic field of health care operations research and the practice of health care operations.
At Geisinger, we are taking their data and modeling patient arrival patterns from a variety of sources such as other hospitals, the emergency department or the operating room after surgery. Once they enter the hospital, patients follow different paths through the health care system. Some go to the intensive care unit. Some go to a standard bed. Some go to an isolation bed. We build a model of the hospital with the current bed capacity and bed placement rules to evaluate how it's running and are able to estimate what would happen if we made changes without having to adjust the real system. The model is a sandbox to play with and evaluate the impact of a variety of alternatives.
Q: A few years ago, you were part of a research team studying medical decision making, specifically looking at treatment options for HIV-positive patients. What were you and your collaborators seeking to find in this study?
A: Medical decision making research focuses on questions such as when clinicians should start a therapy for a patient. In the HIV treatment study, we used methodologies for quantitative decision making under uncertainty. I was at the University of Pittsburgh at the time, and with a group there, we examined treatment data on HIV-positive patients. We looked at two major measures in patients: their white blood cell count, which is a measure of how healthy you are, and viral load, which is a blood count of virus cells. In patients with HIV, typically what happens when you start a therapy is that the viral load goes down. Then, the disease mutates, and the therapy becomes less effective, and the viral load creeps back up. So you start with another therapy, and the process goes on until you are out of options.
One of the disadvantages of using this series of therapies is that once you start, you can't go back. There has been debate about whether you should start the therapies right away or wait until the virus is more established. There are risks and tradeoffs to all of this. The risks are if you start too early, you have this mutating virus and you run out of options too soon. If you start too late, HIV becomes full-blown AIDS, you have irreparable damage. My work is all about the tradeoffs.
Q: What were your conclusions about HIV treatment options, and have they resulted in changes?
A: Our biggest conclusion was that "hit hard, hit early" was the best treatment strategy for patients with HIV. So when someone has HIV, you start them on the hard therapy and start them full-on. There was a school of thought that it is better to delay until the patient's viral load gets higher and their white blood cell count goes down, but our analytics showed you get a greater expected number of life years (adjusted for quality of life) by starting patients on therapy earlier. Our study was somewhat exploratory in that few people were applying these methods in 2008, but it was corroborated by clinicians, and is now the recommended standard of care.
Medical decision science helps clinicians make better decisions using data. It is not a randomized control trial, but those are very expensive and are limited in the number of policy options they can evaluate. In addition, there are human concerns with experiments. Our work can be used to guide the types of randomized control trials, allowing the trials to focus on the most promising policies, saving both time and research dollars.
Q: As a professor, why is it important for future business leaders and entrepreneurs to understand business analytics?
A: Obviously I am biased, but one of the things that is and will be important in the future is an understanding of how risk and uncertainty impacts decisions and how it can be accounted for in making decisions. Historically, you had to have a pretty high level of understanding of statistics and probability to analyze risk, and people would largely ignore it. But with the advent of tools like spreadsheet software, which has simulation modeling built in, we are able to account for and analyze risk fairly easily. We can use data to make better decisions. If you are able to do both quantitative and qualitative work, it will be a competitive advantage for you as an individual and any organization with which you are involved.
Interviewed by Julia Ferrante
New editions of "Ask the Experts" will appear on the Bucknell website on most Thursdays during the fall and spring semesters. If you have ideas for future topics or are a faculty or staff member who would like to participate, contact Molly O'Brien-Foelsch.
To learn more about faculty and staff experts who can speak on a variety of news topics, visit Bucknell's searchable Experts Guide.
The following links are virtual breadcrumbs marking the 27 most recent pages you have visited in Bucknell.edu. If you want to remember a specific page forever click the pin in the top right corner and we will be sure not to replace it. Close this message.