Toward Opioid-Free Ambulatory Surgery: A Prospective Study Using Machine Learning to Predict Postoperative Opioid Use

Area of Science:

• Anesthesiology
• Pain Management
• Machine Learning in Healthcare

Background:

• Postoperative opioid use presents risks of dependence and diversion.
• Developing effective opioid-sparing strategies is crucial for patient safety.
• Identifying patients at high risk for opioid use is essential for tailored pain management.

Purpose of the Study:

• To develop and evaluate an opioid-sparing regimen for ambulatory surgery.
• To create a machine learning (ML) model to predict postoperative opioid use.
• To identify key factors associated with postoperative opioid consumption.

Main Methods:

• Implemented the Toward Opioid-Free Ambulatory Surgery (TOFAS) program, alternating ibuprofen and acetaminophen with limited oxycodone rescue doses.
• Developed an ML model to predict postoperative opioid use in adult patients undergoing ambulatory surgery.
• Validated the ML model using area under the receiver operating characteristic curve (AUC) with an 80/20 train-test split and 10 random seeds.

Main Results:

• 42% of 223 enrolled patients filled their opioid prescription, with a median of 4 doses used.
• The ML model achieved a mean test AUC of 0.674, with sensitivity of 0.70 and specificity of 0.68.
• Key predictors of opioid use included active cancer, age, anesthesia type, race/ethnicity, COPD history, intraoperative complications, preoperative acetaminophen use, and pain intensity.

Conclusions:

• The developed ML model reliably identifies patients at high risk for postoperative opioid use.
• This predictive capability supports personalized, opioid-sparing pain management strategies in outpatient settings.
• The model facilitates tailored pain management planning, potentially reducing opioid dependence and diversion.