Predictive Analytics Models Help Plan for COVID-19 Demands

To inform decision-making during the COVID-19 pandemic, Cleveland Clinic has created predictive analytics models that help forecast patient volume, bed capacity, ventilator availability, and other metrics.

For more coronavirus updates, visit our resource page, updated twice daily by Xtelligent Healthcare Media.

With COVID-19 rapidly spreading among patient populations, hospitals and health systems are facing unprecedented strains on resources and capacity. Organizations are increasingly turning to real-time analytics tools to help track and predict healthcare demands.

The predictive models, developed by Cleveland Clinic and SAS, provide timely, reliable information for hospitals and health systems to optimize care delivery for COVID-19 and other patients.

The analytics models were designed to create worst-case, best-case, and most-likely scenarios, unlike some forecasts that focus on a prediction based on a single set of assumptions. The models can adjust in real time as situations and data change, such as social distancing measures and disease spread.

Using this information, Cleveland Clinic can predict and plan for future demands on the health system, including ICU beds, personal protective equipment, and ventilators. After reviewing possible COVID-19 surge scenarios generated by the models, Cleveland Clinic activated a plan to prepare for its worst-case scenario. The health system built a 1,000-bed surge hospital on its education campus for COVID-19 patients who don’t need ICU care.

Cleveland Clinic also used the models to inform decision-making around organizing and activating new labor pools.

“These predictive models were developed jointly by two organizations that understand patient populations, data and modeling,” said Chris Donovan, executive director of Enterprise Information Management & Analytics at Cleveland Clinic.

“We are sharing the models publicly so health systems and government agencies globally can use them in their own communities. Our hope is that others contribute their ideas and improvements to the models as well.”

The models are freely available via GitHub. The link has been visited over 1,700 times in the past two weeks, resulting in more than 50 downloads.

At the center of the project is an epidemiological Susceptible, Exposed, Infected, and Recovered (SEIR) model, developed by Cleveland Clinic and SAS. The SEIR model is based on a University of Pennsylvania open source model that has been recorded and expanded on the SAS analytics platform and continuously improved with real-time feedback from Cleveland Clinic epidemiologists.

The resulting models include flexible control of model parameters and different model approaches that consider regional health and demographic variations and state-level assumptions.

The models apply advanced analytics to hospital data to help optimize resources, improve situational awareness, and ensure demand-planning stability.

“These models can help hospitals, health care facilities, state departments of health and government agencies forecast the impact of COVID-19 and prepare for the future,” said Steve Bennett, Ph.D., Director of SAS’ Global Government Practice. “The models can also assist more vulnerable, less developed health systems in the fight against COVID-19.”

Organizations have made monitoring and planning for care surges a top priority during COVID-19. In March, Definitive Healthcare and Esri launched a data platform that allowed users to analyze and track US hospital bed capacity, as well as potential geographic areas of risk.

“As the COVID-19 virus spreads in the US and around the world, public health providers and other types of responders have been trying to understand which areas may require the greatest amount of aid. We built this geographic map of US hospital bed utilization and capacity to solve some of these challenges,” said Este Geraghty, ESRI’s Chief Medical Officer.

Researchers from RAND Corporation also recently released an interactive tool that allows decisionmakers to estimate current care capacity and explore strategies for increasing it. Inputs into the spreadsheet tool include baseline number of beds, critical care doctors and nurses, respiratory therapists, and ventilators.

“Because the crisis falls upon a system that already is stretched thin, creating the critical care capacity needed for the surge in COVID-19 patients will require creative thinking about the allocation and use of space, staff and the stuff needed to provide critical care,” said Dr. Mahshid Abir, co-author of the report and senior physician researcher at RAND.

As the COVID-19 pandemic continues, predictive models and other advanced analytics tools will likely play a critical role in measuring and mitigating the impact of the virus.