Risk factors and prediction model of emergence delirium in children after general anesthesia under preoperative sedation with uniform oral midazolam premedication: a prospective observational cohort study.

BACKGROUND: Emergence Delirium (ED) is a frequent complication following general anesthesia in children. It is multifactorial, with no single cause conclusively identified. However, limited research has examined the risk factors for ED in children under such a protocol. Therefore, our goal was to identify independent risk factors for ED in this unified preoperative sedation protocol. METHODS: This prospective observational study collected medical records of children undergoing elective tonsillectomy and adenoidectomy under general anesthesia at Kunming Children’s Hospital between March 2023 and September 2023. Based on the Pediatric Anesthesia Emergence Delirium (PAED) scale, participants were categorized into a non-ED group and an ED group. Univariable logistic regression analysis was performed to compare the groups, and factors showing statistical significance were included in a multivariate analysis. Significant risk factors were identified, and a nomogram prediction model was constructed. The Receiver Operating Characteristic (ROC) curve, model calibration curve, and decision curve analysis were used to assess the model’s performance. RESULTS: Among this single-center cohort of 143 children, 47 children experienced ED, yielding an incidence rate of 32.9%. Logistic regression analysis identified younger age, male sex, low-dose opioids, and absence of remifentanil infusion as significant risk factors for ED in children under preoperative sedation. The area under the ROC curve was 0.818 (95% CI 0.751–0.885), with a sensitivity of 87.2% and specificity of 63.5%. The Youden index was 0.508. The Hosmer-Lemeshow test demonstrated good calibration of the prediction model (P = 0.592), and the C index visualized the nomogram model was 0.818. CONCLUSIONS: Independent risk factors for ED in children were younger age, male sex, low-dose opioids, and absence of remifentanil infusion. The prediction model based on these factors showed reliable performance in predicting ED in children undergoing preoperative sedation within our center.