Columbia Business School Professor, DSI Member Carri Chan on How to Improve Hospital Operations in the Midst of a Pandemic
Carri Chan, a hospital operations expert and an associate professor in the division of decision, risk, and operations at Columbia Business School, has spent the last decade developing data-driven models to improve the flow of patients through intensive care units. Since the onset of COVID-19, her expertise has been featured by Bloomberg, Fox News, New York Magazine, Chicago Tribune, New York Daily News, and NBC.com. Here, the DSI member reflects on how hospitals have been contending with the virus and how the pandemic has impacted her research.
You've indicated in past interviews that “hospitals are preparing well and we will emerge from this.” Do you remain optimistic?
We are moving in the right direction. Things are definitely improving. I’m optimistic, but not complacent. The proactive measures the hospitals took put them in a much better position. They are not completely out of the woods yet, but things are moving in the positive direction. Demand is starting to slow down, so hospitals are starting to prepare for “reopening” services beyond COVID-19 patients and emergencies. They are still thinking about how to be prepared if a second wave arrives. Knowing that COVID-19 patients, for instance, often require mechanical ventilation and for long periods of time, clinicians are exploring ways to utilize their resources and alternative therapies to manage demand better.
How have you dealt with the lack of data on the coronavirus?
At the early stages of COVID-19, there was very limited data, especially regarding U.S. patients. The epidemic has been evolving so rapidly that data has been limited, or there has been limited time to verify the reliability of data that is coming out. We had to utilize published data from other countries and try to make inferences about what might happen in the U.S. As more data in the U.S. becomes available, though, we can start calibrating and tailoring our analysis to the U.S. population. Researchers and scientists are doing the best we can with what data is available out there.
Why did you decide to focus on hospitals in your research?
Hospitals comprise the largest portion of health care costs in this country. Improving efficiency of hospital operations has the potential to reduce health care costs and, more importantly, improve quality of care. Additionally, the system is highly complex, so there are many interesting and exciting new areas to study.
Which data science techniques do you use to study hospital operations?
I combine empirical approaches with mathematical modeling to develop evidence-based approaches to improve patient flow through hospitals, and particularly intensive-care units. My primary interests are in data-driven modeling of stochastic systems, dynamic optimization, and queueing with applications in health care operations management. I utilize both econometric and machine-learning approaches to inform decisions. Most of my work focuses on using patient-level data that tracks patients through different units within a hospital. I also use operational data, which captures the number of patients in a unit and the level of congestion in each unit. My work is highly interdisciplinary by nature. It is great that Columbia has structures in place, such as the Data Science Institute, to help researchers connect to each other and to help facilitate their work.
You have spent years studying an area not typically covered by the media. How have you adjusted to being called upon regularly to explain how hospitals operate?
I have always been very passionate about working in this area, and it is incredibly rewarding and fulfilling now to have others see the value in my work. Of course, it goes without saying that I would have preferred that such an epidemic never occurred, and that the dire need to carefully manage hospital resources was not so severe as it has been in the past few weeks and months. Even in non-pandemic times, the effective management of health care resources has important quality implications. It has been made ever more relevant now and I hope that the world will take this opportunity to recognize its importance, so that as things start to “return to normal” more resources will be focused on this area of study.