Using AI to Lower Readmissions and Save More Than $4M in Just Months -- Testimonials Title
The University of Tennessee Medical Center, a Level I trauma center and the Knoxville region's only academic medical center, is applying Jvion's advanced Artificial Intelligence (AI) and Clinical Patient Pod technology to lower readmission rates and improve the health outcomes of their patient population.
The University of Tennessee Medical Center is reducing avoidable readmissions using a mix of advanced Artificial Intelligence (AI) and the latest in Clinical Patient Pod (CPP) technologies provided through Jvion's clinical predictive solution. The medical center is integrating predictive analytic outputs directly into the clinical workflow. These insights are empowering the more than 30 case managers responsible for reducing readmissions and improving the health of patients post-discharge. Current efforts have led to an average drop of approximately 70 readmissions per month, which, in accordance with the Medicare average for cost of admission, could save health systems millions annually in potential losses.
UT Medical Center is a 609-bed academic medical center, which delivers care to some of the state's most complex and underserved communities. Dr. Trey La Charité, a UT Medical Center hospitalist, explained, "what everyone knows is that 5% of the population uses 50% of the available resources we have in healthcare and that 20% of the population uses 80% of those same healthcare resources. The problem is that the 5 or 20% you identify today is not going to be the same 5 or 20% next month, next quarter, or next year." Over 2014 and 2015, the medical center saw an increase in monthly readmissions. UT Medical Center's Chief Quality Officer, Dr. Inga Himelright, sought to stop the rise in readmissions and reduce overall rates through the application of advanced predictive analytics.
"We really wanted to find a solution that would take us from the static statistical methods that we are all familiar with like LACE to the next level in clinical predictions," said Dr. La Charité. "Jvion's solution combines advanced artificial intelligence and machine learning with a totally innovative approach that uses Clinical Patient Pod technology. Because of the capabilities inherent to Jvion's solution, we are able to predict with ten times the accuracy if someone is going to have an avoidable readmission event. And we can do this using our existing data and leveraging standard integration into our clinical applications. The results after just a few months of integration have proven to be quite effective.
UT Medical Center is integrating the solution's predictive outputs directly into the case manager workflow. These predictions provide patient level insights into the risks driving the likelihood of a readmission and the best possible predicted interventions. Based on this information, case managers develop the custom and targeted plans that are leading to reduced readmission rates for the medical center.
"Going forward, we will expand the adoption of Jvion's solution outputs to more groups within the medical center who are responsible for care," said Dr. La Charité. "The long-term impact to our readmission rates is huge. We could potentially see a drop of 50-70% as we drive further use and application of the solution. And we are looking at other areas for predictive application including hospital acquired conditions and infections where we could really improve patient health outcomes and the allocation of care resources within UT Medical Center."