Joel Gordon, chief medical information officer at UW Health, Wisconsin, noted that trusts are a major theme in many conversations among healthcare leaders about how to safely and effectively incorporate AI into the field.
He observed this week in an interview at the Reuters Digital Health Conference in Nashville.
Gordon notes that public trust is fragile – a high-profile failure could have been in trouble for years.
“Whether it’s gene therapy or any possible gene therapy, we have to historically reflect on the things we assumed trust – but we didn’t build or gain trust. Then something bad happened and the floor fell down to buy, funding to get the ability of science to keep science going – and then we can’t let us get into growth. We can’t let us declare.
Therefore, he believes that healthcare leaders must prioritize AI governance and transparency.
Gordon said individual health systems and other organizations have established governance frameworks and clear rules on the path to AI use — but these efforts are still lacking at the national level.
While figuring out how best to manage healthcare AI, he said the industry needs more collaborative learning than redundant research.
In his opinion, more learning consortiums are needed. He describes these as collaborative groups involving various health systems, in which they work together to align research methods, goals and data frameworks to accelerate advances in AI and reduce duplication of efforts.
“We have the opportunity to think about the same idea of learning how colleges are studying, the way these universities look at and where we are trying to go collectively as an industry. We are in infancy, and I think it’s important that we recognize that. If we think that our last 18 months are fast – the next two or three years will be very surprisingly fast, then there are things we can learn, we’re going to learn, Goldo.”
As healthcare providers continue to browse the process, it is important to remember that metrics and utilization are more important than flashy headlines.
Gordon noted that he often sees headlines celebrating the speed and scale of AI launches, such as highlighting that 25,000 doctors use tools to go online within 10 months. But for him, this missed the point.
“It’s cool, but we’re not thinking about the quality of outcomes in all these different perspectives – billing, safety, patient education, continuity of records, continuity of documentation routes – and it really matters in the end,” he said.
Gordon added that many hospitals touted their successful AI deployments – but actual data used, such as frequency and distribution among users, is often lost.
Overall, he believes the industry should prioritize trust, collaboration and real-world outcomes to ensure the continued value of AI in healthcare.
Photo: dmitrii_guzhanin, Getty Images
