RESUMEN
Postoperative pulmonary edema (PPE) is a well-known postoperative complication. We hypothesized that a machine learning model could predict PPE risk using pre- and intraoperative data, thereby improving postoperative management. This retrospective study analyzed the medical records of patients aged > 18 years who underwent surgery between January 2011 and November 2021 at five South Korean hospitals. Data from four hospitals (n = 221,908) were used as the training dataset, whereas data from the remaining hospital (n = 34,991) were used as the test dataset. The machine learning algorithms used were extreme gradient boosting, light-gradient boosting machine, multilayer perceptron, logistic regression, and balanced random forest (BRF). The prediction abilities of the machine learning models were assessed using the area under the receiver operating characteristic curve, feature importance, and average precisions of precision-recall curve, precision, recall, f1 score, and accuracy. PPE occurred in 3584 (1.6%) and 1896 (5.4%) patients in the training and test sets, respectively. The BRF model exhibited the best performance (area under the receiver operating characteristic curve: 0.91, 95% confidence interval: 0.84-0.98). However, its precision and f1 score metrics were not good. The five major features included arterial line monitoring, American Society of Anesthesiologists physical status, urine output, age, and Foley catheter status. Machine learning models (e.g., BRF) could predict PPE risk and improve clinical decision-making, thereby enhancing postoperative management.
RESUMEN
BACKGROUND: The American Society of Anesthesiologists Physical Status Classification System is commonly used for preoperative assessment. Patient physical status before surgery can play an important role in postoperative nausea and vomiting. However, the relationship between the physical status classification and postoperative nausea and vomiting has not been well defined. METHODS: Adults aged ≥ 18 years who underwent procedures under anesthesia between 2015 and 2020 were included in the study. We analyzed the relationship of postoperative nausea and vomiting with physical status classification score using propensity score matching and Cox hazard regression. Differences in intraoperative use of vasopressor and inotropes and invasive monitoring were investigated according to the classification. RESULTS: A total of 163,500 patients were included in the study. After matching, classification 1 versus 2 included 43,400 patients; 1 versus ≤ 3, 13,287 patients; 2 versus ≤ 3, 23,530 patients (absolute standardized difference, 0-0.06). Patients with physical status classification ≤ 3 had a significantly lower postoperative nausea and vomiting risk than those with classification 1-2 (physical status classification 1 vs. ≤ 3, hazard ratio 0.76 [0.71-0.82], P < 0.001; 2 versus ≤ 3, hazard ratio 0.86 [0.82-0.91], P < 0.001). Intraoperative use of vasopressor or inotrope and invasive monitoring were noted more in the high physical status classification than the low physical status classification (absolute standardized difference [0.19-1.25]). CONCLUSION: There were differences in intraoperative invasive monitoring and use of vasopressor or inotrope among the classifications, and a score of 3 or higher reduced the risk of postoperative nausea and vomiting more than a score of 1-2.