Is it time for AI in healthcare to get an office of its own?
That’s one way to sum up a recent commentary by Chris Scarisbrick, sales director for Sectra, a Sweden-based imaging and cybersecurity vendor, in which he discusses with the UK’s NHS is ready for a Chief Clinical AI Officer.
For perspective, he begins by pointing out that chief clinical information officers didn’t even exist in the NHS a decade ago, but “they have become some of the most influential people in the health service, having a very visible and positive impact on strategic thinking within their own trusts, across regions and even nationally.”
While one could argue that CCIO could have responsibility over AI developments as well, Scarisbrick wonders if CCIOs, with so much on their plate already given the rise of digital everything in healthcare, should also be expected to lead on AI adoption?
Other industries have started to recruit chief artificial intelligence officers, Scarisbrick says, and some AI specialists in the NHS are starting to appear from within specialities.
“At a recent diagnostic AI conference, I met a trust’s radiology AI lead who was focused on identifying new applications to improve the effectiveness of his department. But it is notable that though he had a detailed understanding of what could help radiology, he was far less aware of what was happening in other parts of the hospital – including pathology.”
Moreover, he says, while the idea of AI champions in the NHS isn’t new, there’s a good chance it’s not evolving in the most coordinated or efficient way.
“Could a CAIO or CCAIO align AI strategy across patient pathways, and across the enterprise, and start to prioritize how trusts use AI in a more joined-up strategic way that addresses organizational and even national NHS objectives?” he asks.
The bottom line for Scarisbrick is that mature AI applications focused on helping to improve the rapidity and accuracy of diagnoses of specific and serious illnesses are continuing to develop, and while vendors are happy to help, not surprisingly, administrators come up to speed on the possibilities, the can’t make the final decision on which specific applications will best solve the clinical requirements of individual trusts.
In short, he says, “the NHS needs people in place to judge the clinical efficacy of these applications, to assess whether this works in clinical practice, and to do all of this with an enterprise understanding of trust processes, pathways and priorities.”
And if the NHS needs to give AI an official seat at the executive table, chances are healthcare organizations on this side of the pond probably do, too.