Patient Falls Drop 39% with Prescriptive Analytics
After incorporating machine learning technology into its fall prevention program, El Camino Hospital significantly improved fall rates.
Identifying patients who are risk for falling is a key component in traditional fall prevention programs. However, successfully reducing falls requires interventions that go beyond simply acknowledging risk factors.
El Camino Hospital, a 446-bed, two-hospital health system located in Silicon Valley, California, has added a high-tech twist to its fall prevention program and significantly improved fall rates.
"We had fall rates that were beyond the benchmark," says the organization's Chief Nursing Officer Cheryl Reinking, RN, MS, NEA-BC. "We're a Magnet hospital, so we like to stay in the top quartile or the top decile. It's important to us not only for that reason, but also for patient safety."
The organization leveraged its Silicon Valley location and worked with Qventus (formerly analyticsMD), a neighboring technology firm, to help it incorporate machine learning and prescriptive analytics into its fall prevention program.
"We explained our problem to them: There's all this data, we screen our patients, we do all these things, and we're still having falls—more than we want to have, and more than the benchmark," Reinking says. "They said to us, 'We think we can help you.'"
And Qventus has. Six-months after the hospital began incorporating the prescriptive analytics technology, fall rates dropped by 39%.
The Data Goldmine
Medicare patients make up about half of El Camino's patient population and often have multiple comorbidities which, in addition to age, put them at risk for multiple adverse events, not just falls.
Keeping track of which issue needs immediate attention can be a challenge.
"Nurses have lots of alarms, lots of data coming at us from all different directions. We do lots of mini risk assessments on our patients. We screen them for their risk of DVT, their risk for falls, their risk for infections, and on and on," Reinking says.