68% of health system execs plan deeper AI investments to meet strategic goals

Photo: HIMSS

During the inaugural HIMSS State of Healthcare digital event on Tuesday, Thomas Kiesau, director and digital health leader at The Chartis Group offered some useful benchmarking for hospitals and health systems wondering how their IT investments and digital maturity compare to their peers.

The Chartis Group surveyed 226 health system executives nationwide, at hospitals and health systems of all shapes and sizes. They were asked their perspectives on digital health, artificial intelligence and machine learning – and about the financial health of their respective organizations.

The health system execs had generally positive sentiments about digital health. Most agreed with the sentiments, said Keisau: “It should be integrated into the care model. It will advance care over the long term. It can create a competitive advantage. It can help close care disparities. Across the board what we saw from our respondents was that they believe those relatively strongly.”

As for some of the commonly-held negative perceptions associated with digital health – that it’s “separate from care delivery, it doesn’t have clear ROI, that it’s not suitable for certain populations” – most respondents didn’t agree, he said.

“This is a positive affirmation about the future of digital health and believe that it shows there’s an opportunity that our health system executives see and that they don’t buy into the hype and the negative perceptions associated with it.”

Asked about the greatest barriers to digital health adoption, meanwhile, 59% cited regulatory and reimbursement issues, and 42% mentioned clinician buy-in. Only 18% listed technology issues.

Chartis also asked health system executives to describe their digital health efforts to date, and elicited “a wide range of responses,” said Keisau.

“We asked, are these system-wide coordinated? Are they targeted, high impact? Are there a bunch of pilots but not really clear rollout plans? All the way down to ‘we really haven’t started.’ 

What’s notable is that almost everyone has started, whether because of COVID-19 or something they’d already initiated. But 52% of the programs have not progressed beyond the pilot stages.”

As for the priorities and strategic imperatives these digital health efforts are aimed at, “patient access is the number one priority for over a third of the health systems,” said Keisau, “with 70% of the respondents putting it as a top three priority.” Another major goal, of course, is cost reduction, with about 60 percent of execs putting it in their top three.

“You can only improve what you measure,” said Kaisau. “So we also asked our survey respondents to respond to the initiatives they’re currently operating and are they being tracked and monitored?

“Interestingly, 60% say they are tracking them at a system-wide, real-time progress level, and another 30% are saying they’re tracking them with quarterly updates, but regularly and consistently tracking progress; 40% are in various stages of centrally tracked, but not effectively managed or monitored – or not monitored at all. What this says is if you are not managing and monitoring your digital progress, you’re behind.”

To the question “Do you have a digital health officer today?” Only about a third of respondents said yes. However, “more than a third said they don’t have one today, but they plan to hire one in the next one to two or five plus years,” said Keisau. “They recognize the need.”

AI investments on the rise

To the question of AI and machine learning, the Chartis Group report suggests that it’s still in the early going at most health systems, with 70% of the respondents yet to establish any sort of strategic artificial intelligence programs.

But when asked whether their investments in AI will change in the future to help achieve strategic goals, “we saw a pretty strong endorsement that it’s going to need to increase,” said Keisau.

Thirty-eight percent of execs predict a marginal increase in AI investment in the years ahead, “and a full 30% say it needs to increase significantly,” he said. “Collectively, nearly 70% of the respondents identified artificial intelligence machine learning as an area that would need to receive greater investment to be able to achieve their future enterprise goals.”

Chartis asked execs about an array of AI and ML use cases: insurance, cybersecurity, fraud prevention, dosage error detection, supply chain, voice-assisted charting, registration, remote patient monitoring, referrals, imaging and labs, consumer-facing chatbots, surgical robots, clinical trials, care planning, triage, diagnosis and more.

There was widespread agreement that artificial intelligence applications can help with all of them.

“Every single category, people saw it as a viable candidate,” said Keisau. “But there was a notable difference between those that received the highest scores – operational use cases such as insurance and cybersecurity – while the clinical interventions, things like care planning, care triage and diagnosis, came in much lower. There’s a general belief in AI machine learning and its value, but also a general recognition that it’s not fully understood.

“There’s positive perceptions around AI – that it could increase new job capabilities,” he said. “This could focus workload. This could make our people do better, reduce health care disparities and reduce cost. But at the same time, there are negative perceptions that it could also cause job losses. It could create some level of risk when we’ve got artificial intelligence guiding more decisions.”

But even with some uncertainty about just what AI and ML can accomplish practically – and even with most initiatives still in a relatively nascent state – majorities of executives “realize it’s an area they need to invest in,” said Keisau, “even if it’s just to address the operational use cases or to address the cost pressures they see coming.”

Twitter: @MikeMiliardHITN
Email the writer: [email protected]

Healthcare IT News is a HIMSS publication.

Source: Read Full Article