Nomogram-based prediction of emergence delirium in elderly patients undergoing laparoscopic surgery

Abstract

BackgroundA nomogram model has been developed to forecast the incidence of emergence delirium (ED) in elderly patients undergoing laparoscopic surgery. MethodsA secondary analysis was performed on the eMODIPOD trial, which involved elective laparoscopic surgery patients at Xiangya Hospital between October 2017 and October 2019. To assess ED, the Richmond Agitation Sedation Scale (RASS) was combined with the Intensive Care Unit (CAM-ICU) scale, and patients were categorized as ED or non-ED based on the results. The two groups were compared in terms of perioperative information and test results. Logistic regression was utilized to investigate factors contributing to ED occurrence in elderly patients following laparoscopic surgery, and a nomogram was constructed to depict this. The diagnostic performance was evaluated through the utilization of the area under the receiver operating characteristic curve (AUROC) and a calibration plot. Simultaneously, to appraise the clinical value, decision curve analysis (DCA) was employed to assess the clinical value. ResultsOf the 565 patients subject to laparoscopic surgery, 94 were subsequently diagnosed with emergence delirium (ED), yielding a positive rate of 16.6 %. The developed Nomogram model was an amalgamation of several variables, which included preoperative cognitive function, educational level, dezocine, penehyclidine, ASA grading, blood loss, fluid infusion, and Post Anesthesia Care Unit-numerical rating scale (PACU-NRS). The model’s performance was commendable in terms of its discrimination power, as signified by an area under the curve of 0.713(95 % CI = 0.655–0.772). Furthermore, it exhibited adequate calibration, as affirmed by the Hosmer-Lemeshow test (χ2 = 2.243, p = 0.326). Internal validation of the model further emphasized its superior discrimination and calibration capabilities. The decision curve analysis pointed out that when the threshold probability exceeded 0.1, the Nomogram model delivered more substantial benefits compared to the treat-all and treat-none approaches. ConclusionsThis scoring system represents the pioneering Nomogram model utilized for predicting emergence delirium (ED) in geriatric patients undergoing laparoscopic surgery. It exhibits robust efficacy in forecasting the risk of ED, thereby furnishing valuable insights for the prevention, management, and treatment of ED.