Machine learning predicts which patients will continue using opioids after hand surgery

A machine learning algorithm performs well in predicting the risk of persistent opioid use after hand surgery, reports a study in the August issue of Plastic and reconstructive surgery.

“We found that a machine learning model works well in identifying patients who have undergone hand surgery and who are likely to become persistent opioid users,” said Kevin C. Chung, an ASPS member and surgeon at the University of Michigan in Ann Arbor.

“This could provide a more efficient strategy to identify high-risk patients and implement measures to prevent opioid addiction. Similarly, the use of artificial intelligence could enable a more personalized approach to prescribing the right painkillers in the optimal amount for a specific patient undergoing a particular surgery.”

Two machine learning models tested to predict persistent opioid use

The study evaluated two previously described machine learning models: one using patient-reported data from the Michigan Genomics Initiative (MGI) and one based on insurance claims data. The models were first evaluated in a large sample of general surgery patients, and then in patients undergoing hand surgery, such as surgery for carpal tunnel syndrome or a wrist fracture.

The study focused on whether the machine learning models could predict which patients would develop persistent opioid use, based on prescriptions filled up to six months after surgery. The MGI model included 889 patients, about half of whom had previously used opioids. The claims model was limited to 439 “opioid-naïve” patients, with no recent opioid use.

In the MGI model, which included prior opioid users, 21% of patients developed persistent opioid use. In the insurance claims model, which excluded prior opioid users, 10% of patients had persistent opioid use.

In the area under the curve analysis, the MGI model performed very well in identifying patients with persistent opioid use: 84% in the model trained on hand surgery data and 85% in the general surgery population. In contrast, in the claims model, the predictive ability was 69% based on hand surgery data and only 52% in the full dataset.

Machine learning could streamline postoperative opioid risk assessment

In the MGI model, having a prescription for opioids before surgery was the strongest predictor of postoperative opioid use. Other predictors were general body pain and the prescription of hydrocodone, a relatively potent opioid often prescribed for postoperative pain.

As with other types of surgery, persistent opioid use is a risk for patients undergoing hand surgery. While some risk factors have been identified, assessing postoperative opioid risk is a challenging and time-consuming process given the diversity of the patient population and the varying complexity of procedures. The new study suggests that machine learning could provide a more integrated, straightforward approach to identifying high-risk patients.

Models that include patient-reported data on factors such as pain and mental health, as collected in the MGI, appear to provide the highest predictive value.

“With access to rich datasets, machine learning can streamline the identification and analysis of detailed factors that influence patients’ pain experiences after surgery,” the researchers write.

The authors note some limitations of their study, which may not reflect changes in prescribing patterns in response to the opioid epidemic. Dr. Chung and co-authors conclude, “In practice, these models can be implemented as decision support tools to help clinicians efficiently identify patients who are most vulnerable to addiction and in need of tailored pain management or counseling.”