Cedars-Sinai Develops Predictive Models for Post-Surgery Pain Management

A team of researchers at Cedars-Sinai Medical Center has developed artificial intelligence (AI) and machine learning (ML) algorithms to predict which patients are most likely to successfully manage their pain post-surgery and which patients might need additional assistance.

The algorithms are currently being used to help spine surgeons evaluate surgery success, which has traditionally been limited to how well a patient can walk, bend or move after spine surgery. However, the use of AI allows clinicians to predict additional outcomes related to surgery success, such as pain management.

“The unique thing we’re doing with this project is really focusing in on the pain medication part of it, because opioid addiction continues to be a challenge, and we are looking for ways to improve pain management after surgery,” said Corey Walker, MD, a neurosurgeon focused on spine treatment at Cedars-Sinai, in the press release.

Walker’s team is collaborating with the Cedars-Sinai Department of Computational Biomedicine to develop and study AI algorithms. The algorithms make predictions based on analyses of millions of data points within a subset of the data and then test these predictions against new subsets. By doing so, the algorithms continually improve and update the methodology as they learn more. While clinical trials usually test one or two variables at a time, algorithms like these can evaluate thousands of variables concurrently.

“The more data you feed it, the better,” Walker said. “We look at everything from a patient’s blood pressure to their age, to what types of medications they were taking before surgery and how long they’ve been on those medications.”

But when the algorithms identify a patient likely to struggle with weaning from pain medications, Walker noted that the process becomes more complicated, as ethical questions must be considered. Patients cannot go without the surgery or treatment once they are being considered for spine surgery, so engaging the pain management team and other specialists early on could be key to navigating possible misuse issues.

Previous research has suggested that other techniques, such as weaning a patient off pain medication or changing the drug before surgery, can impact the patient’s need for pain medications after surgery, Walker added.

One such pain management solution has been shown to significantly reduce opioid prescriptions given following surgery, which has the potential to reduce opioid addiction.

Researchers at the Perelman School of Medicine at the University of Pennsylvania successfully launched a pain management program designed to distribute opioids only to those patients who needed them following robotic urological procedures. Post-operative guidelines suggest sending these patients home with 10 to 15 oxycodone pills.

The researchers wanted to test the program that would allow patients whose pain warranted opioids to get a prescription while ensuring that patients whose pain could be managed with lesser medications were not getting opioids unnecessarily.

Overall, 68 percent of patients were discharged without prescriptions for opioids. Just under a quarter of patients went home with 10 tramadol pills. Only 8 percent of patients were prescribed 10 pills of oxycodone, a 92 percent reduction in opioid prescriptions that would have been given without the program’s guidance.