Artificial intelligence and machine learning will play a big role in shaping spine surgery over the next several years. Technology is transforming clinical care and decision-making in multiple ways, but is it always for the better?
Surgeons debated how AI and ML will affect spine surgery over the next several years at the Becker's 21st Annual Spine, Orthopedic & Pain Management-Driven ASC + The Future of Spine Conference in June.
"AI and machine learning are going to take over our assessment of which patients are appropriate for what surgeries, help us design our surgeries and then move into the intraoperative space to help us do surgeries and get better results," said Richard Wohns, MD, founder and medical director of Neospine in Seattle. "On the postop side, it will track our outcomes and be able to present data. The data sets will be enormous of course and AI is best suited for big data sets."
Hospitals and ASCs can use AI to spot trends and further standardize care. The increased personal information about patients accelerates precision medicine to customize treatment plans based on data instead of intuition.
"I think AI and machine learning are here to stay. They're going to be basically taking over what we do as artists, not scientists, where we have subjective knowledge and experience guiding our surgical acumen. AI and ML will take over that aspect of things and objectify it, and get us better results," said Dr. Wohns.
While some surgeons welcome the data-driven future, others resist it. There are still risks with imperfect data and AI model hallucinations.
"One thing that is a double-edged sword with tremendous upside as well as tremendous downside is AI and how we use data," said Carlos Bagley, MD, chair of the department of neurological surgery at Saint Luke's Health System in Kansas City, Mo. "How we analyze data, what it is used for and our understanding of what the data tells us and what it doesn't tell us is going to be very important as our field evolves from within but also from the outside as well."
Dr. Bagley said the upside to AI is more efficient workflows and supporting the clinical staff amid nurse and physician shortages. Surgeons can use AI for risk stratification and take steps to avoid complications with high-risk patients.
"In terms of risk stratification, genetic risk for certain conditions or even complications can help us focus our attention on how to implement avoidance strategies with certain patient populations," said Dr. Bagley. "You see that on the extreme other end as well. If you have genetic markers that make you more at risk for hip osteoarthritis, it is going to cost you more to get health insurance because you're more likely to need a hip arthroplasty. You can see how it can help us, but it could also be used to create barriers for certain patients."
Another downside of AI is how insurers can use it to automate prior authorizations and denials for a variety of procedures. Insurers have a sophisticated data infrastructure to support decision-making, which impacts care surgeons can deliver.