Abstract
To develop and evaluate a predictive model for intensive care unit (ICU) admission among patients with acute sedative-hypnotic overdose. We conducted a retrospective analysis of patients admitted to the emergency department of West China Hospital, Sichuan University, between October 11, 2009, and December 31, 2023. Patients were divided into ICU and non-ICU groups based on admission criteria including the need for blood purification therapy, organ support therapy (ventilatory support, vasoactive drugs, renal replacement therapy, artificial liver), or post-cardiopulmonary resuscitation. Patients were randomly split into a training set and a validation set in a 7:3 ratio. Least Absolute Shrinkage and Selection Operator (LASSO) regression was used to optimize variables, followed by a multivariate logistic regression analysis to identify independent risk factors for ICU admission. A nomogram model was constructed and assessed using receiver operating characteristic (ROC) curves, calibration curves, Decision Curve Analysis (DCA), and Clinical Impact Curve (CIC). Predictors in the nomogram included barbiturate overdose, Glasgow Coma Scale (GCS) score, and anion gap at admission. The nomogram demonstrated strong predictive performance with an area under the curve (AUC) of 0.858 (95% CI: 0.788-0.927) in the training set and 0.845 (95% CI: 0.757-0.933) in the validation set. Calibration curves showed the model closely matched the ideal curve, and DCA and CIC indicated high clinical applicability and utility. Barbiturate overdose, initial decreased GCS score and decreased anion gap were identified as independent risk factors for ICU admission in acute sedative-hypnotic overdose. The nomogram model based on these indicators demonstrates good predictive accuracy, discrimination, and clinical utility.