Imagine two patients who undergo the same hip replacement procedure, conducted by the same surgeon, at the same hospital, on the same day. One recovers quickly, while the other struggles with severe post-operative pain.
This scenario happens too often. Doctors and researchers have always known that patients' post-operative experiences depend on a complicated mix of variables, from obvious factors like age, sex, preoperative pain, and underlying health conditions to less obvious ones like anxiety, worry, and depressive symptoms.
But now, artificial intelligence tools are making it possible to predict precisely which patients are at higher risk of severe post-operative pain. These groundbreaking tools can enable doctors to tailor patients' treatment regimens before they even enter the operating room to prevent unnecessary suffering and shorten recovery times.
More patients deserve access to them.
Research consistently shows that pain variation among patients is the norm, not the exception. Those with anxiety or other psychosocial risk factors often report higher post-operative pain scores than those without them, even when the medical procedure is identical. Differences in patients' nervous and immune systems and genetic makeup can also influence pain levels. So can biological sex, race, and socioeconomic status.
The same sorts of factors also help predict who's more vulnerable to opioid dependence after surgery.
Opioids remain an effective tool for treating post-surgical pain -- and ultimately, managing and reducing pain is critical to a successful recovery. The key is striking the right balance among pain management methods. That starts with understanding each patient's background and risk factors.
Studies show that post-operative pain and post-operative opioid use don't simply reflect the severity of the surgery. They reflect the whole patient.
One study, for example, found that there is no significant difference in the rate of new persistent opioid use between patients who underwent major procedures and those who had minor ones -- surgeries that should, theoretically, produce very different levels of pain. The researchers suggested that new persistent opioid use was due to "addressable patient-level predictors."
In other words, to treat pain effectively and actually prevent it from becoming severe and chronic, we need a stronger, more holistic assessment of each patient before surgery begins. That's where predictive artificial intelligence (AI) can make a difference.
Consider how Washington University researchers used AI to analyze a series of pre-surgery surveys, along with patient lab results, health history, and other clinical information. They were able to determine patients' probability of developing persistent pain after surgery, along with the certainty of that estimate.
Or consider our Pain Prevention Research Center at the Hospital for Special Surgery (HSS). There, after a comprehensive review of studies that use AI to predict postoperative pain and opioid use, we've developed AI projects that identify risk factors and patients at risk for more problematic pain following surgery.
For example, we used AI to identify important factors for predicting pain and patients at high risk following total knee replacement. That study leveraged HSS's vast patient databases and found that younger age, greater physical or mental impairment, and higher BMI were risk factors for higher post-operative pain.
Identifying these risk factors early can allow teams of doctors -- like HSS's Perioperative Pain Service team -- to work with patients to create tailored pain-management plans that rely less on opioids. After using AI to automatically flag higher-risk patients from standard pre-operative data, they can help patients manage medications before surgery, choose the right medications during and after surgery, and provide access to psychosocial support.
Over time, this level of customization means less pain, lower costs, and fewer lives disrupted by avoidable long-term opioid use.
Ultimately, all surgeons want to reduce their patients' pain and help them heal quickly. With emerging technologies and better use of the information we already collect, health systems across the country can do just that.
Alexandra Sideris, PhD, is the C.V. Starr Director of Pain Research at Hospital for Special Surgery.
