How UCHealth is using data analytics to optimize capacity, streamline COVID-19 response

By February 1, 2021 February 3rd, 2021 News

As hospitals and health systems continue to navigate the COVID-19 pandemic, organizations such as UCHealth are increasingly turning to data and predictive analytics tools to better manage supply and demand.

During Becker‘s CEO + CFO Virtual Forum Jan. 20, industry experts discussed the need for data-driven insights supporting operations in healthcare and why predictive analytics are necessary in the age of COVID-19.

The presenters were:

  • Steve Hess, CIO at UCHealth in Aurora, Colo.
  • Mohan Giridharadas, founder & CEO of LeanTaaS

Six key insights from the presentation:

1. The pandemic has increased the need for using data to manage capacity on a daily basis, rather than weekly, monthly or quarterly. At UCHealth, staff members were making decisions around capacity on an hourly basis, and sometimes more often, Mr. Hess said.

“Taking actionable data coming out of our EHR and other systems and serving it up to clinical and operational leaders to make decisions was really critical to our COVID-19 response,” he said. “We use positivity rates in the community to predict hospital census, and we’re looking at multiple models across the country and globe to figure out how they can pertain to Colorado and our health system’s response.”

2. UCHealth also deployed LeanTaaS’ predictive analytics tools to help manage the system’s daily inpatient, emergency department and hospitals’ transfers processes. The company’s tools focus on mapping supply and demand matching to help hospitals better predict outcomes and better manage tasks such as OR block utilization and inpatient bed availability.

3. Optimizing capacity boils down to two main concepts, according to Mr. Giridharadas. Those factors are: Matching, or how to match supply and demand minute for minute every day of every week; and linking, which is stringing together a series of disconnected service deliveries.

“Dashboards and reports aren’t enough. It takes constraint-based optimization methods, machine learning, artificial intelligence and simulation algorithms to solve the problem,” Mr. Giridharadas said.

4. Matching supply and demand patterns is critical in healthcare; on the supply/capacity side, hospitals must ensure that when delivering any health service that the staff, equipment and facilities are all available at the same place and at the same time to accommodate the demand. Conversely, on the demand side, the health system must have an accurate understanding of the number and type of patients they are providing the service, when they will show up and how long each visit will take.

“It’s a very difficult mathematical matter to put together,” Mr. Giridharadas said. “When we talk to health systems, and say ‘you need to bring the sophistication,’ they simply put up their hands and say ‘it’s too hard – you can’t predict the demand, you can’t predict the supply.’ Instead, two people will look at a calendar and decide [scheduling for services] but no one did demand side or supply side math. This is why health systems end up getting backed up and the patient waits at every step of the journey.”

5. UCHealth first deployed LeanTaaS’ technology in 2015 to improve its infusion center workflow, which was seeing long waiting periods during peak visit hours. With LeanTaaS, UCHealth extracted data from its EHR and and was able to calculate how to modify its EHR templates to flatten the peaks, Mr. Hess said.

6. Since 2016, UCHealth has rolled out LeanTaaS’ infusion centers predictive analytics tools across all its infusion centers and has also deployed similar technologies for its operating rooms, ambulatory clinics and inpatient beds.

Mr. Hess described some of the results as “UC Health has seen impressive results across the board from the LeanTaaS implementations. Infusion centers have been able to decrease wait times even as we experienced double-digit growth in volume, excluding COVID period. OR utilization has improved significantly by creating an active ‘marketplace’; Quality of in-patient bed capacity decision making has improved from 60 percent to nearly 100 percent.”