A PDF version of my research statement can be found here.

My research interests lie at the intersection of healthcare operations, optimization, and decision support. Additionally, I consider the role of IT in the decision making process and how optimization-based models can be incorporated. My work seeks to specifically address these issues within the field of healthcare operations.

My dissertation seeks to contribute to the Management Science literature by taking a holistic view of an important process at a local hospital: cardiac diagnostic testing. Specifically, the hospital was concerned that their cardiac care center was not operating as efficiently as it could be and approached us with the problem. Some of the symptoms they observed to reach this conclusion were: the inability of the cardiac diagnostic center to test inpatients on the day requested thus prohibiting the patient from being discharged that day, long waiting times for patients before stress testing began, and the lack of nuclear doses needed for the imaging portions of nuclear stress tests. In essence, I am working closely with the hospital to help them make better decisions in three main areas: (1) real-time allocation of testing time slots to various categories of patients, (2) scheduling outpatients, and (3) evaluation of strategic options (e.g., expansion).

The close working relationship with the hospital has allowed me to gain a clear understanding of possible decisions and their implications to the cardiac care center and the hospital as a whole. It has also provided me with a real-world historical data set to test my models against before actual implementation of the decision support tools.

Within the existing OR-healthcare literature, I have identified the current healthcare throughput management techniques. I have demonstrated to my dissertation committee that the leading edge existing Markov Decision Process approach to patient service in a diagnostic medical facility cannot be well-adapted to the hospital's operational model. Specifically, when taking a holistic view with all relevant operational variables considered, computational times are prohibitively long. Because the run-time of this approach exceeds the real-time decision epoch, this benchmark model is not useful in practice. To confront this issue, I have developed a multi-commodity, time-space network flow model. The overall goal is to embed these optimization-based models into decision support tools for the cardiac diagnostic center at the hospital which address the three main areas mentioned earlier.

Two phases compose this project. Phase I excogitates the operational setting where I am interested in the real-time allocation of testing time slots and scheduling outpatients. Phase II focuses on the strategic side where I will evaluate more long term hospital decisions. By integrating capacity planning analysis into the short-term operational framework from the first phase, I will be able to see the operational effects of expansion decisions such as new equipment purchases and new cardiologist hires.

Implications for Research. The main theoretical contribution my dissertation makes is to provide new and innovative models for use in the healthcare operations setting. My dissertation also addresses the need for a holistic view of the hospital's processes and how it affects different departments throughout the organization.

Implications for Practice. The methods used and the tools developed will be robust and configurable ensuring their adaptability to other hospitals in different situations. By providing new and innovative models and decision support tools for the local hospital, I am helping the local community as well as the healthcare industry as a whole. The local community benefits by having the hospital become more efficient and save money, thus enabling it to provide better service to its customers. The healthcare sector benefits by having a proof-of-concept decision support system implementation of flexible optimization-based models.

With my research, I seek to enhance the methodologies and tools used in the healthcare setting to make better, more efficient decisions. Specifically, my work attempts to bring optimization-based decision support tools to healthcare management professionals in order to give them better alternatives to make important decisions. While the discussion of my dissertation provides a nice overview of my research interests, I want to highlight two other research projects.

My interest in providing healthcare decision support stems from a successful patient flow project completed with the local hospital. Specifically, in a paper that has been accepted for publication at Operations Research, my co-authors and I provide an innovative decision support system to help manage the allocation and reallocation of admitted patients to beds. The marked improvement in transfer time helped improve overall hospital efficiency and saves the hospital an estimated $600,000 annually.

Additionally, my co-author and I have a forthcoming paper in the Communications of the ACM that addresses the topic of the utility of a national healthcare information network as well as barriers to its development and implementation. In the paper, we argue that although the cost is high, the associated benefits in the long-term far outweigh these costs. In addition, such a network would provide a critical foundation for a number of initiatives important to homeland security and disaster response efforts. We propose that the Federal Government play a much stronger role in advancing the creation of this network, including at least partial financing of development, deployment, and migration.