Rubin
Geoff Rubin, MD, discussed the idea that radiologists should apply a rubric when thinking about how AI facilitates clinical practice and the value it brings in different areas.
"This is a very important and interesting moment in time for AI and radiology," said Dr. Rubin, professor and chair of the Department of Medical Imaging at the University of Arizona in Tucson. "I would personally like to see us as a profession following a framework that shows the public, payers, and the government that we are prepared to use these tools responsibly for the betterment of patient experience and outcomes."
AI Brings Value to Radiology in Myriad Ways, But Not All Of Them Are Necessary
The value equation, as Dr. Rubin refers to it, looks at factors including the cost of the AI tool, the outcome of the exam including both the quality of the exam itself and the experience from the patient perspective, appropriateness and necessity of the tool for clinical decision making and how much waste it introduces into the process.
"AI solutions can affect any one of those variables, but assessing against all of them can help derive an expectation of whether or not it is reasonable to implement," he said. "The spectrum of clinical applications, value propositions, and integration models for AI and medical imaging are numerous and diverse."
Dr. Rubin's presentation also delved into the issue of reimbursement, though it is still relatively rare to see financial reimbursement for AI tools.
"Reimbursement will depend on who derives primary value and must balance delivering patient and societal value while avoiding overpayment for solutions that bring intrinsic value to hospitals and doctors, risking overutilization, overdiagnosis, and excessive expense." Dr Rubin stated. "It's still early days on this. It's possible many AI solutions may never see reimbursement, so it will be incumbent on radiologists and their management teams to figure out how to derive the value of these tools in a way that still makes them a net positive."
For example, he said, if an AI tool allows radiologists to shorten the amount of time it takes to perform an MRI because it takes raw data and constructs images from them, the hospital or clinic can convert that time savings into increased volume or decreased staffing to find a cost savings that balances the cost of the tool.
"When you think about these things from a holistic view, an AI solution can bring net value without actually having reimbursement," Dr. Rubin said.
Dr. Rubin said he hopes radiologists think about ways to organize their thoughts around new technologies that are coming out all the time and think about the "ifs" and "hows" to justify bringing them into their clinical practice.
Access the presentation, "Artificial Intelligence in Radiology: Managing Professionalism Challenges," (S5-RCP35) on demand at Meeting.RSNA.org.
