Putting AI to work for veterans’ health
- By Dr. John Showalter
- Jul 06, 2018
Across the United States, millions of veterans are waiting too long for health care. To address this, Veterans Health Administration hospitals are working diligently to speed up and improve the quality of care delivered to veterans through multiple programs. Fortunately, the power of artificial intelligence can be leveraged to not only improve care immediately, but also to predict health care needs and determine the best way to prevent illness and deterioration.
The same AI that is embedded in sophisticated search technology and personal assistants can now be used to help veterans. The most effective AI solutions leverage eigenspace -- a platform for solving complex challenges such as quantum mechanics, facial recognition, search and analysis of consumer behavior.
A simplified analogy is a personalized radio station provided by a service like Pandora. Starting from a small amount of data, the service is constantly improved by whether listeners like or dislike songs provided to them. The system “learns” very quickly, becoming extremely proficient at picking songs the listener will like based on past behavior. The same kind of technology can be put to work for the more vital purpose of saving lives.
An eigen-based approach provides a comprehensive patient view that can predict future health complications and the interventions that will reduce risk, thereby greatly increasing veteran health and fiscal efficiency. Some of the common conditions that can be improved with AI include better prevention of veteran suicide, hospital-onset infections and readmissions caused by ineffective outpatient care.
AI is empowering providers with a solution that extends to all aspects of patient care, across ambulatory and inpatient settings. It provides at scale the capability to “think” about patients the way clinicians do -- as complex, constantly changing individuals -- and translates patient data in a way that directs the right clinical actions care to the right patients at the right time, effectively reducing clinical overload.
AI solutions have proved effective at more than 300 hospitals and institutions, including the Cleveland Clinic, the Mayo Clinic and InterMountain Health Systems. The results have been immediate and impressive:
- Readmissions: A recent hospital trial aimed at reducing readmissions prevented 227 readmissions, improved care and delivered a cumulative ROI of $2.5 million.
- Pressure injuries (bed sores): Leveraging AI resulted in a 45 percent incident reduction and savings of over $770,000.
- Employee health: Integrating AI analysis resulted in a 31 percent reduction in employee ER visits, a 36 percent reduction in inpatient visits and a 50 percent reduction in care costs.
- Avoidable admissions: A health care facility saw a 30 percent reduction in admissions from a targeted, high-risk population, while delivering better care and saving $700,000, a 500 percent return on the program.
The digitized nature of AI means that it easily integrates with current clinical technologies, including Vista, the VA's homegrown health record system. EHR data is processed through the eigenspace, producing risk and recommendation outputs with little to no impact to the current VA workflow.
Improving the care veterans receive is literally a life-or-death issue. The amazing pace of technological development has put powerful AI tools at our disposal, tools that are being used elsewhere in the private sector. This technology must be leveraged as soon as possible by every VA facility.
To not do so to improve veterans’ health care would be a sign of misplaced societal priorities.
Dr. John Showalter is the chief product officer at Jvion.