A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validations between...

Area of Science:

• Anesthesiology
• Clinical Risk Prediction
• Patient Safety

Background:

• Two independent centers developed risk scores for predicting postoperative nausea and vomiting (PONV).
• The current study aimed to validate these scores across different centers and explore simplification without compromising accuracy.

Purpose of the Study:

• To assess the cross-center validity of existing PONV risk scores.
• To simplify logistic regression-based risk scores for PONV prediction.
• To evaluate the discriminating power of original and simplified risk scores.

Main Methods:

• Adult patients undergoing inhalational anesthesia without antiemetic prophylaxis from two centers (n=520 and n=2202) were analyzed.
• Logistic regression models were used to derive risk scores, which were then simplified.
• Cross-validation and analysis of a combined dataset were performed to assess score validity and identify a final risk score.
• Discriminating power was measured using the area under the receiver operating characteristic curves (AUC).

Main Results:

• Risk scores demonstrated validity across centers (AUC = 0.65-0.75).
• Score simplification minimally impacted discriminating power (AUC = 0.63-0.73).
• No significant center effect was detected (OR = 1.06).
• Key predictors identified: female gender, history of motion sickness or PONV, nonsmoking, and postoperative opioid use.

Conclusions:

• Developed PONV risk scores are valid across different centers and can be simplified without losing predictive accuracy.
• The simplified risk score shows broad applicability for predicting PONV in adult patients receiving inhalational anesthesia.
• Prophylactic antiemetic treatment is recommended for patients with at least two identified risk factors.