Artificial intelligence-informed tools are providing patients and surgeons with highly personalized recommendations for treatment options and their likely outcomes.
The result? Better comfort and capability — along with a greater understanding from patients about their health experience and outcomes.
At UT Health Austin’s Musculoskeletal Institute, the Joint Insights tool is proven to helps make patients informed decision-makers in their own care.
A 59-year-old patient with moderate knee osteoarthritis enters a care delivery room at UT Health Austin’s Musculoskeletal Institute. They’re preparing to discuss a total joint replacement for the affected knee, which is already starting to impact their ability to walk comfortably.
But even before setting foot in the room and shaking hands with the surgeon, the patient’s medical history and personalized predictions based on their “digital twin” have already been examined down to the last detail, using an algorithm that relies on AI. When Karl Koenig, M.D., medical director for the institute and one of its surgeons, walks in the door, he and the patient already know that they have an estimated 68% chance of benefiting from surgery.
Joint Insights, a tool developed by digital health company OM1 and adopted by the institute team for use in Austin and at UT Health San Antonio, takes a holistic view of a patient’s medical history: their past procedures, other health needs or comorbidities, whether and how often they’ve visited the emergency department in the past year, and more. A 2021 randomized controlled trial published in JAMA Network Open from a team led by Prakash Jayakumar, M.D., Ph.D., showed significant improvement in decision quality and patient experience when using the tool, without impacting time spent consulting with the surgeon.
“The factors that determine whether a surgery will be successful or not — or even what ‘success’ means — are quite varied, patient to patient,” Koenig says. “As their surgeon, I offer my recommendations based on my experience and expertise, but having the support of clear-cut data forecasting their likely outcomes levels the playing field when it comes to making important decisions about their care.
“In other words, the patient becomes an informed decision-maker in their own care.”
The instances where a patient chooses not to move ahead with a procedure are just as important as the instances where they do.
Prakash Jayakumar, M.D., Ph.D.
What Researchers (& Patients) Are Learning
While the team continues to study longer-term outcomes via an additional control trial, they’re also going beyond health outcome data: A qualitative study is now pinpointing the ways patients experience the tool itself based on their decision-making preferences and individual backgrounds.
One early takeaway: The tool often helps patients better understand the limitations of surgery from the outset, setting the stage for them to make realistic choices for their health.
“A lot of people think surgery is going to be a complete, surefire fix,” says Haley Ponce, value-based health care fellow at the Musculoskeletal Institute and medical student at Baylor College of Medicine. “In reality, surgery can fix a lot of things, but there’s always a chance of residual mobility issues or lingering pain, especially if the decision to operate is not appropriate and aligned with the patient’s preferences, values and needs.”
As for the 59-year-old with osteoarthritis? The 68% chance of benefit might indicate that surgery was a go. But based on all the factors presented in their report — including favorable predictions of improvements in pain, stiffness and quality of life — compared with cost, recovery time and risks associated with the total knee replacement, they instead chose to begin with a regimen of noninvasive steroid injections.
“Appropriate treatment selection is critical in current health care, especially in orthopaedic surgery where there are a range of treatment choices. The instances where a patient chooses not to move ahead with a procedure are just as important as the instances where they do,” Jayakumar says. “That’s part of what equity-conscious AI and value-based health care is all about.”