Development Of A Prediction Model For Emergency Endotracheal Intubation In Patients With Spontaneous Intracerebral Hemorrhage

Purpose1.To screen risk factors of emergency endotracheal intubation in patients with acute spontaneous intracerebral hemorrhage（SICH）,construct a risk prediction model,validate and evaluate it.2.To investigate the incidence and risk time of emergency endotracheal intubation in patients with acute SICH.MethodsA retrospective analysis of medical records of 355 eligible SICH patients admitted to the emergency department of a grade III class A hospital from September 1,2018 to October 31,2021 was performed.The medical data were divided into training set（290cases）and validation set（65 cases）according to the chronological order.Univariate analyses comparing the training set with the validation set were conducted using the Student’s t-test,Mann Whitney U test,Chi-square test and Fisher exact test.The least absolute shrinkage and selection operator（Lasso）binary logistic regression was used for feature selection and model development.The validation of the model was performed using the area under the receiver operating characteristic curve（AUROC）,a calibration curve and decision curve analysis.The model was then visualized using a nomogram.A bar chart was drawn to display the risk time of emergency intubation in patients with SICH.ResultsA total of 355 participants were included in this study,with 65 cases of emergency endotracheal intubation which accounted for 18.31%of all participants.The high-risk period was 4 hours after the onset of disease.Univariate analyses found significant differences in time of onset,heart rate,SPO₂,MEWS,a CCI,vomiting,consciousness,left pupil diameter,reflex to light of bilateral pupils,and whether bilateral pupils were equal between the intubation group and the non-intubated group（P<0.05）.Nine predictors were selected by Lasso including a CCI,vomiting,time of onset,heart rate,SPO₂,MEWS,left pupil diameter,left pupil light reflex,and right pupil light reflex.For the training set and validation set,the AUROC was>0.8,sensitivity>90%and specificity 70%.The calibration curves of the training set and validation set highly overlapped with the reference line and the Hosmer-Lemeshow test P>0.05.The decision curve analysis indicated good clinical efficiency.ConclusionsThe risk prediction model consisted of nine predictors:a CCI,vomiting,time of onset,heart rate,SPO₂,MEWS,left pupil diameter,left pupil light reflex,and right pupil light reflex.This model showed good prediction power,calibration power,and clinical efficiency,which was predictive of the occurrence of emergency endotracheal intubation in patients with SICH within 24 hours of onset.It is helpful for medical staff to identify patients with need for emergency endotracheal intubation beforehand and prepare for airway management.