Essential IT: How Providence uses data to care for COVID-19 patients

This is the second in a series profiling IT executives who helped guide their organizations through the COVID-19 pandemic and toward recovery.

B.J. Moore knew he had to stick to his gut.

He had just made a crucial choice as the new chief information officer at Providence, one of the nation’s largest health systems, spanning 51 hospitals and 1,085 clinics in seven Western states. But his choice was getting some pushback.

“I was told,” Moore recalls, “my decision would be politically and tactically difficult.”

That moment came in July 2019 when Moore led Providence to sign a multi-year strategic alliance with Microsoft to develop a suite of digital solutions to improve health outcomes while reducing the cost of care.

More specifically, the two companies agreed to combine Microsoft Azure, Microsoft 365 and advanced compute capabilities like Artificial Intelligence (AI) with Providence’s clinical expertise and care environments. The goal was to accelerate the health care industry’s adoption of cloud-based medicine for the benefit of patient care and caregiver productivity.

The initial game plan was ambitious. Soon, it became downright bold.

In the summer of 2019, Providence began to migrate 120,000 employees – who were using a mix of web conferencing and communication applications – onto a single platform, Microsoft Teams. Their goal was to complete that massive move in 18 months.

Then came COVID-19.

On Monday Jan. 20, 2020, at Providence Regional Medical Center in Everett, Washington, doctors began treating the first confirmed U.S. case of the novel coronavirus. The patient was a 35-year-old man who, five days earlier, had flown back to Seattle after visiting family in Wuhan, China.

A nurse in a face mask and blue scrubs checks a smart tablet.

A nurse checks data on a smart tablet. (Courtesy of Providence)

Inside the hospital’s special pathogens unit, the patient initially grew sicker, developing pneumonia. Doctors treated him with the anti-viral drug remdesivir – once used successfully in Ebola patients. It worked. On Feb. 3, the man was released from the hospital to recuperate at home.

But by then, Moore and other Providence executives recognized that a pandemic might be unfolding.

“We started to see what we could do to accelerate our Teams migration, because we were still on WebEx, Lync, Skype and Teams. We knew, if we went into the shutdowns and we didn’t have all of our caregivers on a single collaborative platform, it was going to be a disaster,” Moore recalls.

Providence doctors and administrators needed to swiftly communicate with each other across the system to share their treatment successes, surge coordination, supply chain challenges and other COVID-19 revelations.

And to contain the infection rate within the hospital system, thousands of caregivers had to switch to do their work remotely, literally overnight.

The IT department worked urgently to shrink the deployment’s remaining 10 months down to mere weeks. By April, all 120,000 employees were on Teams.

“It was make-or-break for us,” Moore says. “We wouldn’t have been able to function as a health system if we didn’t make that transition.”

His one-time skeptics were now appreciative.

“I’ve never received more love email in my entire life,” Moore says with a laugh. “The same people who were doubtful now were emailing to say: ‘You made the right choice.’”

Meanwhile, the move to Azure also began to pay off for Providence, which previously had used only on-premises data warehouses. By spring, Providence executives sought to forecast which hospitals and clinics within the system would need emergency shipments of personal protective equipment (PPE) and ventilators before new COVID waves emerged.

They had two puzzles to solve: Could cloud data help them detect early signs of the virus in the communities they serve – and could that same data help Providence predict where COVID would surge next?

The answers were: yes and yes.

“By then, we had already moved a majority of our data to Azure. And we were able to use the native machine learning and the AI in Azure to move our assets where they were needed,” Moore says. “If we were on-prem, those tools wouldn’t have been available to us.

“We got to the point where we were able to predict – two weeks in advance and with probably 85 to 90% accuracy – where the (COVID-19 case) surges were going to hit.”

A man stands with his back to a plaster pillar, wearing a maroon jacket, white shirt and blue jeans.

B.J. Moore.

The locations of those future spikes resided inside the sea of data generated by patients who recently had visited some of Providence’s 1,085 primary care clinics. They had arrived complaining of fevers, coughs and other symptoms, yet they were not sick enough to require hospitalization.

But about two weeks after those clinic visits peaked in certain areas, a portion of those patients would show up at nearby emergency rooms, needing potentially needing life-saving treatment, Moore says.

“We now knew, for example, we were about to have a surge in Southern California, so we moved ventilators and PPE there in advance,” he says.

And in non-clinical settings, such as medical billing and coding, Providence employees began utilizing Teams and Azure to maintain high performance, productivity and coordination while keeping caregivers safe, Moore says.

Those moves are now transforming how the organization thinks about its real estate footprint as well as the future of the modern medical workforce.

For example, Moore foresees a day when millions of people use, say, smart watches to routinely upload their body temperatures into their cloud-based health records. In that world, a single doctor could then use AI and machine learning to scour that deluge of data to detect an emerging local illness.

“Our doctors are amazing. Our medicines are amazing,” Moore says. “But the best way to improve the health of our communities is to keep people from getting sick or to intervene as soon as possible. Technology can enable us to do that.”

Portraits by Dan DeLong.