Development and validation of a predictive model for patients with post-extubation dysphagia

Area of Science:

• Critical Care Medicine
• Gastroenterology
• Neurology

Background:

• Swallowing disorders are common in intensive care unit (ICU) patients, leading to serious complications like aspiration pneumonia and malnutrition.
• Post-extubation dysphagia (PED) is a significant concern for patients requiring endotracheal intubation.

Purpose of the Study:

• To identify risk factors associated with PED in ICU patients.
• To develop and validate a predictive model for PED to assist in prevention and management.

Main Methods:

• Retrospective derivation cohort (n=235) and external validation cohort (n=70) were used.
• Variable screening and LASSO regression identified predictors; multivariate logistic regression and nomogram construction were performed.
• Model performance was assessed using ROC curves, calibration plots, and decision curves.

Main Results:

• Independent predictors for PED included age, sedative pause, level of consciousness, ADL score, nasogastric tube use, sore throat, and voice disorder.
• The nomogram demonstrated strong discriminative ability with an AUC of 0.945 in the derivation cohort and 0.907 in the validation cohort.
• The model showed good calibration and clinical utility as indicated by decision-curve analysis.

Conclusions:

• A predictive model incorporating age, sedative pause, consciousness level, ADL score, nasogastric tube, sore throat, and voice disorder shows potential for predicting PED in ICU patients.
• This model can serve as a valuable tool for early risk assessment and targeted interventions to prevent and manage PED.