News Feature | March 23, 2016

Artificial Intelligence Reduces Hospital Admissions

Source: Elsevier
Christine Kern

By Christine Kern, contributing writer

Hospital

Research reveals benefits of integrating machine learning into remote monitoring.

Home healthcare software provider AlayaCare and home health provider We Care (part of the CBI Health Group)have released a white paper providing insight into how machine learning/artificial intelligence, when integrated into remote patient monitoring, can reduce hospital readmissions and emergency room visits.

According to the study, Better Technology, Better Outcomes: The Effects of Machine Learning Powered Remote Patient Monitoring on Home Health Care, machine learning is a branch of artificial intelligence (AI) based on mathematical algorithms and automation, designed to automate the building of analytical models that use algorithms to learn from data in an iterative fashion. As the machine learns from its mistakes, it can improve its results to produce reliable, repeatable decisions.

Machine learning algorithms have already been successfully applied in a range of industries from finance to retail and even healthcare.

“The transformational potential of machine learning in home healthcare is very promising,” said Jonathan Vallee, Director of AlayaLabs (the research and development arm of AlayaCare software). “With the help of We Care, we have the ability to improve patient outcomes by analyzing incoming patient vitals and referencing against patient data. This means we can predict negative health events — like hospital readmissions and ER visits — with tremendous accuracy.”

The research revealed machine learning can improve predictions by 11 percent while reducing over-diagnoses by 54 percent. When combined with Big Data, AI can also deliver risk scoring to inform and provide insight to clinicians of possible adverse events like falls, episodes, events, and emergency visits, as well as improve patient ability to remain at home longer.

Data for the study was provided by the We Care Remote Access to Care Technology (Re-ACT) program, a remote health monitoring program designed to optimize the management and treatment of seniors living with chronic diseases. The research demonstrated machine learning increased the total cost savings of the program by an additional 6 percent. The machine algorithm correctly predicted 82 percent of events.

The study concluded that machine learning could have the following implications for home health care:

  • Care workers could prevent more than 10 percent of ER visits and hospital readmissions.
  • Patients could spend more time at home.
  • Machine learning could reduce the amount of false positives, in turn reducing patient stress levels and the number of unnecessary procedures.
  • Efficiencies have the potential to be maximized: care workers could prioritize their time based on the urgency of patient needs.
  • Machine learning could significantly reduce the overall cost of care due to the potential decline in unnecessary hospitalization and/or readmissions.

“With increasing demands on the healthcare system and constrained resources, providers like We Care need to find ways to improve the efficiency of home care services,” said Anthony Milonas, Chief Operating Officer of Home Health with CBI Health Group. “Combining the clinical judgement of our healthcare professionals with the information provided by our remote monitoring service Re-Act, we are already improving client outcomes and our ability to serve remote communities. By adding machine learning algorithms with predictive ability, we can reduce negative health events even further. This technology has the potential to dramatically change home healthcare.”