Objective:

To develop and externally validate a machine learning-based prediction model for postoperative pulmonary embolism (PE) and explore key clinical determinants related to PE risk.

Approach:

Key Findings:

• 1.38% of patients developed postoperative PE, highlighting the importance of the predictive model.
• Key predictors included age, BMI, malignancy history, prolonged bed rest, surgery duration, intraoperative tachycardia, CRP, NLR, and postoperative D-dimer.
• XGBoost model showed best performance with an AUC of 0.925 in external validation.

Interpretation:

The model demonstrated strong discrimination, good calibration, and favorable clinical utility, with SHAP analysis revealing influential predictors for PE risk, which can guide clinical decision-making.

Limitations:

• Potential limitations include the retrospective design and the specific patient population studied, which may affect generalizability.

Conclusion:

A robust machine learning model for predicting postoperative PE was developed and validated, providing insights into key risk factors.

Attribution Notice

This content is an AI-generated, fully rewritten summary based on a published scholarly article. It does not reproduce the original text and is not a substitute for the original publication. Readers are encouraged to consult the source for full context, data, and methodology.