Abstract:
Objective To construct and validate a risk prediction model for in-hospital death after successful resuscitation in patients with cardiac arrest.
Methods A retrospective study was conducted on 295 patients with cardiac arrest who successfully restored spontaneous circulation after cardiopulmonary resuscitation and were further treated in hospital. The patients were divided into training and validation sets using K-fold cross-validation and then grouped and compared based on whether in-hospital death occurred. A binary Logistic regression analysis was used to screen risk prediction factors, and a nomogram prediction model was constructed. The model performance was evaluated and validated in the training and validation sets, respectively.
Results The results of the multivariate Logistic regression analysis showed that hospitalization duration (OR=1.180; 95%CI, 1.080 to 1.280; P < 0.001), norepinephrine dose (OR=0.980; 95%CI, 0.970 to 0.990; P=0.002), initial respiratory rate after resuscitation (OR=1.090; 95%CI, 1.030 to 1.150; P=0.004), and sinus rhythm recovery after resuscitation (OR=4.280; 95%CI, 1.670 to 10.980; P=0.003) were independent influencing factors for in-hospital death. A nomogram model was constructed based on these independent influencing factors, and it was verified that the model had good discrimination, calibration, applicability, and rationality.
Conclusion The influencing factors for in-hospital death after successful resuscitation in patients with cardiac arrest include hospitalization duration, norepinephrine dose, initial respiratory rate after resuscitation, and sinus rhythm recovery after resuscitation. The nomogram model constructed based on these factors can provide a reference for clinical decision-making.
下载: