West Tennessee Healthcare is piloting a new technology that uses artificial intelligence to help case managers prevent patients from being kept in the hospital longer than necessary.
The nonprofit health care system will test Dragonfly Navigate, a product designed by Franklin-based AI technology company Xsolis to streamline administrative decisions between a patient’s hospital admission and discharge.
West Tennessee Healthcare serves around half a million West Tennesseans, including residents in rural areas, and began construction of a new Bolivar Hospital in July. WTH has used other Xsolis products since 2022.
The new workflow technology uses artificial intelligence (AI) and real-time data from the hospital to model whether a patient can be discharged and where they should go after leaving the hospital (a skilled nursing facility, home, or to long-term care, for example).
“When a patient is in the hospital longer than what the model is predicting, then it alerts the (case manager) to go back into the case to look to see what’s stopping the patient from being discharged,” Xsolis CEO and founder Joan Butters said.
Barriers can include uncompleted testing, X-rays, treatments, doctor signatures and more. The workflow technology aims to help case managers quickly identify these needs without manually combing through clinical data and documentation.
“Our models are not meant to replace the clinical determination or the clinical expertise of the people that are using our solutions,” Butters said. “We are simply ensuring that they can get to that answer more efficiently.”
Debbie Ashworth, executive director of care management for West Tennessee Healthcare, said in a statement that the workflow will complement clinical expertise and “supports more proactive discharge planning, which will give my teams more time to focus on addressing patient needs.”
Starting discharge planning early helps cut down on extra time spent in the hospital due to logistics. If a patient is being discharged to a skilled nursing facility, for example, pre-planning can ensure there is a bed ready for that patient when they are cleared to leave the hospital, Butters said.
This can save the hospital, health insurance companies, and the patient money, she said.
“Every day a patient is in the hospital more than they need — an extra day — that costs the hospital, on average, about $3,000 a day,” Butters said. “If a hospital has 10,000 discharges and 20% of their patients remain in the hospital one day more than they need to, that’s millions of dollars … that they’ve lost … and that bed’s not available for new patients to come in as well.”
The workflow also helps cut out “a lot of the friction and the back-and-forth” between healthcare providers and insurance companies, she added.
Patients may also see benefits.
“The longer that a patient stays in the hospital inappropriately, there’s higher risk of complications, increased risk of hospital-acquired infections, and higher risk of readmissions,” Butters said.
Xsolis’ AI models are retrained every quarter to adjust them to new conditions and patient populations. The company also goes through a series of annual audits covering data, system and physical security.
This article was originally published by the Tennessee Lookout.
