Behind every hospital bed, doctor, ventilator, mask and the millions of other components that make up a hospital is the same thing: a prediction. How much will we need, and where, and when? Analytics make those predictions as precise as possible—and that’s never been more essential than during COVID-19.
Analytics software company SAS understood the problem better than almost anyone. And not long after the pandemic started, it partnered with the Cleveland Clinic to create an innovative dashboard that would help hospitals optimize their resources and keep saving lives.
How it started: On March 17, the Cleveland Clinic asked SAS to create models that could predict the spread of COVID-19. They wanted to understand the strain that COVID-19 might put on the hospital, and by extension, its resources—from ventilators to PPE to dialysis machines to their doctors’ time.
Why it’s different: While plenty of organizations around the world were building epidemiology curves to track the course of the virus, SAS and the Cleveland Clinic built a framework that offers more. The collaborative team came up with a range of scenarios based on varying inputs like virus transmissibility and social distancing. With SAS vetting the math behind the models, the Cleveland Clinic identified which curve it was on at a given time and developed action plans in advance.
How it worked: The models helped the Cleveland Clinic identify markers for potential surge scenarios and recognize when the actual severity of the outbreak would fall short of some projections. That means it did not have to cancel planned events like routine surgeries and treatments and was able to continue treating non-COVID-19 patients.
- “One of the challenges of this pandemic is the public health cost of dislodging patients with cancer or chronic disease to make room for COVID-19 patients,” said Dr. Steve Bennett, director of the global government practice at SAS. “These models can tell you that you may not need the surge capacity; you can keep doing the sorts of standard work that you’re doing. That has a valuable public health benefit.”
Sharing the wealth: SAS didn’t want to keep such a potentially valuable tool to themselves—so the team made their code publicly available on software development site GitHub. Other hospitals and public health agencies have adapted it, given feedback and made it their own, thus contributing to innovation and effective response.
- “Cleveland Clinic is very advanced in analytics—but at the same time, they really wanted to help smaller organizations and smaller clinic hospitals that may not have big data science teams,” said Natalia Summerville, senior manager at SAS. “That’s why they allowed us to make everything publicly available, which was amazing.”
What’s next: The technology has applications even beyond the current crisis. “SAS aspires to be the platform of the future,” said Dan Abramson, executive director of U.S. manufacturing at SAS (and an NAM board member). “It’s got capabilities in modeling and AI and data management and visualization. So, the knowledge we gain from projects like these can be a launching point for pretty much any business problem or challenge.”
The last word: “The collaboration worked,” said Andrew Williams, principal analytical solutions architect at SAS. “The analyst community has always spoken very highly of our technology and analytic capabilities in AI, machine learning and optimization—and I think what we’ve shown here is that we can apply them to critical use cases across the board and across industries.”