Development and validation of a model for predicting prolonged weaning from mechanical ventilation in patients with abdominal trauma.

ABSTRACT

BACKGROUND: This study aimed to develop and validate a model to predict the risk of prolonged weaning from mechanical ventilation in patients with abdominal trauma. METHODS: Patients with abdominal trauma were included and were divided into the training cohort and the validation cohort. The model was constructed using predictive factors identified by univariable and multivariable logistic regressions, and was validated by receiver operating characteristic curve, calibration curve, and decision curve analysis. Clinical outcomes were compared between model-stratified risk groups. RESULTS: In total,190 patients were included, with 133 in the training cohort and 57 in the validation cohort. Six predictive factors, the Acute Physiology and Chronic Health Evaluation II score, Injury Severity Score, Glasgow coma scale, total bilirubin, skeletal muscle index, and abdominal fat index, were identified and were included in the model. The model predicting prolonged weaning owned a good discrimination, had an excellent calibration, and exhibited a favorable net benefit within a reasonable range of threshold probabilities. Significant differences were shown in prolonged weaning and clinical outcomes between the high-risk and low-risk groups (P < .05). Multivariable Cox regression analysis showed that patients in the high-risk group had greater risk of 28-day mortality (P < .05). CONCLUSION: This study established a model to predict the risk of prolonged weaning from mechanical ventilation and clinical outcomes in patients with abdominal trauma. Skeletal muscle index was identified as one of independent risk factors of prolonged weaning. The findings offer valuable insights for respiratory management in patients with abdominal trauma.