Predictive Model for Stoma Healing After Tracheostomy Decannulation in Patients With Prolonged Tracheostomy

Overview

Prolonged tracheostomy is prone to induce chronic airway mucosal injury and tracheal cartilage degeneration, which not only elevates the risks of post-decannulation complications such as persistent fistula non-healing, airway stenosis and recurrent infections, but also impairs ventilatory, phonatory and deglutitive functions, thereby exerting a severe adverse impact on patients' prognosis and quality of life.Accordingly, the retrospective observational study aimed to characterize stoma healing time after decannulation, identify key predictive factors, and develop and validate a clinically useful prediction model. These findings will offer evidence-based guidance for future researchers in selecting appropriate modeling methodologies and provide a scientific foundation for the development of targeted clinical intervention and prevention strategies.

Objective To develop and validate a predictive model for post-decannulation stoma healing in patients with prolonged tracheostomy, and screen high-risk predictors of poor wound healing. Methods This retrospective observational study will recruit eligible patients and split subjects into training and validation cohorts at a fixed ratio. Candidate clinical and stoma-related variables will be screened via LASSO regression. Multiple prediction models including logistic regression, decision tree and random forest will be built, and evaluated by AUC, Brier score and DCA. Expected Outcomes Core predictive indicators linked to delayed stoma healing will be identified, and the optimal model with robust predictive performance will be determined to stratify patients into high-risk groups. Expected Conclusions The optimal model can achieve accurate risk stratification for stoma healing disorders. Targeted interventions including personalized nutrition, standardized stoma care and activity management will be recommended for high-risk patients.

Conditions