AI successfully identifies risk factors linked to more severe pain after knee replacement

A study using artificial intelligence to classify patient pain archetypes and identify risk for severe pain after knee replacement has earned a Best of Meeting award at the 50th Annual Meeting of the American Society of Regional Anesthesia and Pain Medicine (ASRA). The honor, which recognizes excellence in scientific research, is awarded to three of the top 10 highest-scoring abstracts chosen by the ASRA Research Committee.

“It is an honor to have one of the top professional organizations in the field of regional anesthesia and pain medicine highlight the collaborative work of our department’s Pain Prevention Research Center,” said Alexandra Sideris, Ph.D., director of the Pain Prevention Research Center at HSS. “The award reflects our dedication to innovations in patient care and underscores the greater scientific community’s acknowledgment of our efforts.”

More than one million people undergo knee replacement surgery each year in the United States, and those numbers continue to rise, Dr. Sideris notes.

“There is a need to better understand patients’ individual pain trajectories, and one of the most exciting approaches is to leverage artificial intelligence. With our huge patient database at HSS, machine learning can analyze factors such as age, gender, BMI, and presurgical pain levels to predict which patients are at greater risk of severe pain after surgery,” she said.

Armed with this information, the care team can tailor personalized pain management plans to meet patients’ needs.

The HSS researchers had several goals: utilize machine learning to identify pain archetypes following total knee replacement; determine important features for predicting pain outcomes; and classify patients at risk of severe pain in the immediate postoperative period. The retrospective study included 17,200 patients who had total knee replacements at HSS from April 1, 2021, to October 31, 2024.

“Using unsupervised machine learning, we identified two distinct pain archetypes in patients who underwent total knee replacement, which corresponded to those who experienced severe, difficult to control pain after surgery and those whose pain was relatively well controlled,” explained Justin Chew, MD, Ph.D., a clinical fellow at HSS who presented the study at the ASRA meeting on May 1.

“We then utilized supervised machine learning to determine the most significant predictive factors for severe pain. In our study, risk factors included younger age, greater physical/mental impairment, higher BMI, and preoperative opioid or gabapentinoid use.”

Dr. Sideris notes that ongoing and future studies at HSS will continue to leverage AI with the goal of improving patient outcomes. While the study focused on the immediate postoperative period, she said additional studies will follow patients’ pain trajectory and recovery over longer periods of time to determine which strategies doctors can employ before surgery, intraoperatively and in the immediate postoperative period to manage pain in high-risk patients.

A study using artificial intelligence to classify patient pain archetypes and identify risk for severe pain after knee replacement has earned a Best of Meeting award at the 50th Annual Meeting of the American Society of Regional Anesthesia and Pain Medicine (ASRA). The honor, which recognizes excellence in scientific research, is awarded to three of the top 10 highest-scoring abstracts chosen by the ASRA Research Committee.

“It is an honor to have one of the top professional organizations in the field of regional anesthesia and pain medicine highlight the collaborative work of our department’s Pain Prevention Research Center,” said Alexandra Sideris, Ph.D., director of the Pain Prevention Research Center at HSS. “The award reflects our dedication to innovations in patient care and underscores the greater scientific community’s acknowledgment of our efforts.”

More than one million people undergo knee replacement surgery each year in the United States, and those numbers continue to rise, Dr. Sideris notes.

“There is a need to better understand patients’ individual pain trajectories, and one of the most exciting approaches is to leverage artificial intelligence. With our huge patient database at HSS, machine learning can analyze factors such as age, gender, BMI, and presurgical pain levels to predict which patients are at greater risk of severe pain after surgery,” she said.

Armed with this information, the care team can tailor personalized pain management plans to meet patients’ needs.

The HSS researchers had several goals: utilize machine learning to identify pain archetypes following total knee replacement; determine important features for predicting pain outcomes; and classify patients at risk of severe pain in the immediate postoperative period. The retrospective study included 17,200 patients who had total knee replacements at HSS from April 1, 2021, to October 31, 2024.

“Using unsupervised machine learning, we identified two distinct pain archetypes in patients who underwent total knee replacement, which corresponded to those who experienced severe, difficult to control pain after surgery and those whose pain was relatively well controlled,” explained Justin Chew, MD, Ph.D., a clinical fellow at HSS who presented the study at the ASRA meeting on May 1.

“We then utilized supervised machine learning to determine the most significant predictive factors for severe pain. In our study, risk factors included younger age, greater physical/mental impairment, higher BMI, and preoperative opioid or gabapentinoid use.”

Dr. Sideris notes that ongoing and future studies at HSS will continue to leverage AI with the goal of improving patient outcomes. While the study focused on the immediate postoperative period, she said additional studies will follow patients’ pain trajectory and recovery over longer periods of time to determine which strategies doctors can employ before surgery, intraoperatively and in the immediate postoperative period to manage pain in high-risk patients.