A survey of 500 senior healthcare executives by Optum reveals optimism on the promise of artificial intelligence, offset by the work needed to achieve results.
Participating vice presidents and C-level executives that included CEO, COO, CFO, CTO and CMOs represented hospitals, clinics and delivery systems, life sciences organizations, health insurers and employers. Nearly all the executives reported that they believe AI can make the industry more affordable and accessible.
The survey indicates a tipping point has arrived in adoption of AI in healthcare, with significant investments to be made within the next five years, resulting in positive returns on investment for employers and insurers within three years, and a ROI for hospitals in four or five years.
About 22 percent of surveyed employers already are reporting AI implementations with nearly full deployment.
Further, some surveyed health organizations already are reporting initial AI benefits, with 43 percent of these early adopters automating business processes such as administrative operations and customer service, with 36 percent using AI to detect patterns in healthcare fraud, waste and abuse; and 31 percent using AI to monitor users of Internet of Things devices such as wearable technology, says Steve Griffiths, senior vice president and COO at Optum Enterprise Analytics.
In general, however, now is the time for healthcare organizations to hone artificial intelligence skills, he notes. These include analytics, modeling, machine learning and deep learning, math skills and development of epidemiology skills to identify the causes of disease and outcomes in populations. “We will need people who would help translate strategy into actions to drive business value,” he adds.
Also See: Why organizations need to know full benefits of artificial intelligence
Health organizations of all sizes can begin to hone their skills just as larger ones are doing, Griffiths counsels. “It’s important to have a small data science team, or partner with a larger organization while training existing employees.”
Such in-house training will be necessary because getting professional help could prove difficult as the industry is experiencing a talent crunch to a degree not previously seen. Talent is needed to tackle risk contracting, total cost of care and value-based care, all of which puts a strain on the existing analytics talent.
Organizations with resources should consider getting additional help from provider actuaries that have the math and statistical skills to conduct analytics, data science and computer sciences.
The best type of AI will be automated and operate in the background to perform tasks such as ensuring drugs are delivered in time, prior authorizations are automatically conducted and manual reviews are automated so that staff members, physicians and nurses aren’t spending their time manually reviewing charts and sending data out to insurers to get reimbursed.
The reality, Griffths concludes, is that providers will struggle for a bit as they seek to understand how they can change the tires while still driving down the road toward artificial intelligence. “This will take a corporate transformation strategy,” he believes.