How predictive analytics, telehealth helped one hospital make patients safer
SAN FRANCISCO – El Camino Hospital had a problem: Patient falls were increasing, according to Chief Nursing Officer Cheryl Reinking.
"We take pride in high quality,” Reinking said at the Big Data and Healthcare Analytics Forum on Tuesday. "That was very concerning to us."
So El Camino, which installed a CPOE system back in 1971 and whose pioneering first chief information officer is the namesake for the CHIME-HIMSS John E. Gall Jr. CIO of the Year Award, put advanced analytics to work tackling the serious patient safety issue.
Effective fall prevention comes down to keeping tabs on different variables, of course. A hospital needs to know its patients, and which ones are at higher risk. El Camino performs and documents a fall risk assessment in its Epic electronic health record.
Some strategies are lower-tech: yellow slippers, for instance, to identify the patients at highest risk.
But clearly there was more technology could do. So El Camino partnered with Qventus, based in next-door Los Altos, California, to put machine learning to work improving the problem.
"They came to our multidisciplinary group that was working on this and said, 'We think we might be able to help you,'" said Reinking.
The company is able to analyze data from the EHR to spot patients at highest risk, and can also combine that insight with data from nurse call lights bed alarms.
"Your call light system – who knew that it held all this fantastic data that we never used?" Reinking asked.
By processing all that data, the Qventus technology was "able to not only predict, but more importantly, prescribe what the caregiver should do in the moment when the patient is at highest risk for falling," she said.
"In that moment – when they've put the bed alarm on by moving around in bed, or the bathroom or chair alarm on quite a bit – this algorithm is working. Then the message goes to the nurse's Vocera," Reinking said. "The nurse then can react immediately and enact a protocol. Assess the patient in the moment and decide what the most appropriate means of intervention should be."
That might include moving a patient closer to nurse's station, monitoring them by camera, or putting them under more constant nurse observation.
The result: El Camino realized a 39 percent reduction in falls in just six months.
Reinking conceded that the efforts won't do anything to fix alert fatigue: "It is another beep," she said. "But what the nurses like about this is it's really using intelligence behind the scenes and telling them which patient is most at risk right now. The nurses like that this beep means something, and they need to act on it now."
The hospital has been finding many other different kinds of applications for this machine learning technology, said Reinking. "We've used it in our nursing department for throughput, as well as in our operating room, for throughput and utilization."
She said she's excited about the ways advanced artificial intelligence can continue to improve quality of care and patient safety at hospitals like El Camino: "AI is something we should understand more in healthcare, in how it can help our patients with higher quality outcomes."
In the meantime, the hospital continues to innovate on other fronts. "We're using some really interesting tools – like telemonitoring for patients at home, where we can track a patient's condition and then intervene in a very proactive way. We've had a lot of success with our initial pilot."
From improving transitions of care and case management to reducing readmissions, "analytics will continue to be essential," she added. "Using data to make decisions will be paramount."