Model Pinpoints Predictive Factors for Long-Term Opioid Use

Newswise — As efforts continue to decrease opioid addiction, clinicians seek to tailor care based on patients’ individual needs. Every person has unique characteristics that may affect their likelihood for long-term opioid dependence. Patients who undergo surgery are population with a well-studied risk for chronic opioid use. In the United States, the number of patients who receive total knee or hip replacement surgery is expected to increase by 174% in 2030. Previous studies have identified a relationship between persistent use of opioids before surgery and chronic opioid use following surgery. In a new study, researchers sough to create a model that can predict which patients have a high risk for long-term opioid use after total joint replacement.

Kathyrn Iwata, Engy Said, and Rodney Gabriel of the University of California, San Diego, received a Resident/Fellow Travel Award from the American Society of Regional Anesthesia and Pain Medicine (ASRA) for its 46th Annual Regional Anesthesiology and Acute Pain Medicine Meeting, being held May 13-15, 2021. The authors will present Abstract #1964, “A Predictive Model for Persistent Opioid Use Following Total Joint Arthroplasty,” on Thursday, May 13.

Iwata et al. looked at various characteristics in 198 patients undergoing hip or knee replacement surgery, of which 87 (43.9%) needed long-term opioid use for at least 3 months after surgery. With these data, they created a statistical model that found a relationship between chronic opioid use and the following characteristics: female sex (odds ratio [OR] 2.56, 95% confidence interval [CI] 1.23 – 5.26, p = 0.01), smoking history (OR 8.16, 95% CI 1.45 – 45.90, p = 0.02), hospital length of stay (OR 1.98, 95% CI 0.92 – 4.27, p = 0.08), in-hospital opioid use (OR 1.67, 95% CI 0.77 – 3.61, p = 0.19), and opioid use before surgery (OR 5.73, 95% CI 2.65 – 12.40, p<0.001). 

The model’s findings have exciting implications for helping to guide clinical decision making in patients who undergo surgery. “In the future, we can apply these types of predictive models in surgical patients to develop unique and personalized care for each patient,” Iwata et al. said.