Derivation and validation of a quantitative risk prediction model for weaning and extubation in neurocritical patients.

Abstract

Patients with severe neurological conditions are at high risk during withdrawal and extubation, so it is important to establish a model that can quantitatively predict the risk of this procedure. By analyzing the data of patients with traumatic brain injury and tracheal intubation in the ICU of the affiliated hospital of Hangzhou Normal University, a total of 200 patients were included, of which 140 were in the modeling group and 60 were in the validation group. Through binary logistic regression analysis, 8 independent risk factors closely related to the success of extubation were screened out, including age ≥ 65 years old, APACHE II score ≥ 15 points, combined chronic pulmonary disease, GCS score < 8 points, oxygenation index <300, cough reflex, sputum suction frequency, and swallowing function. Based on these factors, a risk prediction scoring model for extubation was constructed with a critical value of 18 points. The AUC of the model was 0.832, the overall prediction accuracy was 81.5%, the specificity was 81.6%, and the sensitivity was 84.1%. The data of the validation group showed that the AUC of the model was 0.763, the overall prediction accuracy was 79.8%, the specificity was 84.8%, and the sensitivity was 64.0%. These results suggest that the extubation risk prediction model constructed through quantitative scoring has good predictive accuracy and can provide a scientific basis for clinical practice, helping to assess and predict extubation risk, thereby improving the success rate of extubation and improving patient prognosis.