A professor in the Krannert School of Management teamed up with an Indiana University business professor to help the IU Health network manage resources during the COVID-19 crisis.
Pengyi Shi, an operations management professor in Krannert, and Jonathan Helm, an associate professor at Indiana University’s Kelley School of Business, led teams at their respective schools with the goal of predicting potential scenarios that may result from a COVID-19 surge.
The tool they created is based on a computer program that takes in data from state and national levels and then provides predictions on hospital workload (such as the number of patients in ICU) and resource needs, Shi said.
“The hospital system we’re helping is the IU Health system,” Shi said. “Just recently they are serving more than a million residents in our state.”
Shi said they were conducting this research for free, wanting to help a fellow community partner fight COVID-19.
“Of course all the credit should go to the front-line health care workers, but at least we feel like we’re using our knowledge and expertise to perhaps alleviate some of their workload and the system congestion,” Shi said.
The collaborative effort helped the hospitals plan on a state level, since national numbers can be difficult to interpret. Trends that might show up in New York or Indianapolis could be very different from what happens in West Lafayette, Helm said.
“(We) wanted to tailor things to the different regions of Indiana, rather than the national levels,” Helm said.
The two business professors were already talking to IU Health before COVID-19 became a bigger issue, Helm said, planning to help them keep track of nurse staffing decisions and support. So when things seemed to be worsening, the hospital system decided to reach out for more help.
The two paired up with IU Health early in March, and Helm said the model launched near the end of that month.
“That was when we started to see some of the early COVID-19 cases appearing in Indianapolis, so I would commend them for being vigilant, and taking this proactive measure,” Shi said.
Troy Tinsley, director of workforce strategy and operational excellence at IU Health, said the initiative has been a success.
It took several weeks to build this system and get it right, Tinsley said, but what he finds encouraging is how the model is now able to refresh itself to include the newest data to stay ahead of COVID-19.
“As we start to see more COVID patients showing up, it’s gonna project out the need from that point,” Tinsley said. “So instead of waiting until we see that second surge hitting, we can get out in front of it (by) a couple weeks.”
They developed a deeper analysis to track patients’ movement and best care for them while maintaining safety precautions that manage where patients travel within the hospital.
“My colleagues and I have been working on health care operations for over 10 years now,” Shi said, “so we have a lot of experience on how to predict how actual patients flow around the hospital systems, after they actually show up at their sites.”
Shi said this is useful for hospitals because they can look at the model and see how many patients are in the different units, how many nurses each area needs, how many supplies they will need, and even project how many of those patients will need ventilators or enhanced care.
“Our decision support team, or central account analytics support, use that model to forecast different scenarios for what summer and fall, and ongoing, is going to look like with COVID,” Tinsley said. “They use that model to build a deliverable and then send that deliverable out to all the executives across the state.”
Tinsley said there was already a predictive model in place at IU Health, but with the help of Shi and Helm, they were able to better adjust that model for a pandemic, rather than something like the flu.
Helm said once they are able to test the model to show proof of its value, they plan to expand it even more.
“We are hoping to reach out to see if we could help other hospital systems in Indiana, and if there is a second wave it could be beneficial to all the other hospital systems,” Helm said. “And based on what we are seeing around the country now it looks likely that (a second wave) is going to happen.”
Shi said the team is not slowing its efforts now in case there is a second wave.
“We are closely watching how new cases are sort of evolving with IU Health, they’re still diligently monitoring,” Shi said. “We just need to be prepared.”
COVID-19 has brought together different types of people, Shi said, all trying to help however they can, so there is a lot of academic research happening.
“You see academia and community partners, coming from all different areas, all just trying to solve the problem,” Shi said. “We have many people from different disciplines now coming together working on how to help the hospital systems or how to help the society during COVID-19.”
Tinsley said working with the business schools was a valuable learning experience, one he believes will continue.
“A lot of us don’t have a whole lot of experience in the predictive analytic side, so working with Dr. Shi and Jonathan in that space ... I think our teams and the people that were on those weekly calls learned a ton,” Tinsley said.
Shi said she couldn’t say much about the actual changes in place in the hospital because of nondisclosure agreements within hospital policy. But she did say some issues they discussed included how long to postpone elective surgeries in preparation for higher patient intake, and how much personal protective equipment should be available at all times.
The research teams are examining what may happen in the fall, when classes at Purdue and IU resume.
Shi said they are seeking opportunities to present their research now to their respective universities, and Shi has been presenting some of her work to Purdue’s Regenstrief Center for Healthcare Engineering.
Besides educating students, she said a goal in academia is to create something valuable to give back to society.
“We really feel proud our research has had some real, immediate impact,” Shi said.