Indicators of different outcomes after prolonged weaning

Area of Science:

• Critical Care Medicine
• Pulmonology
• Respiratory Therapy

Background:

• Prolonged mechanical ventilation (PMV) weaning outcomes vary widely, from successful liberation to mortality.
• Specialized weaning centers manage complex cases, necessitating outcome prediction.
• Identifying predictors for different weaning trajectories is crucial for patient management.

Purpose of the Study:

• To identify specific indicators associated with distinct prolonged mechanical ventilation weaning outcomes.
• To develop a risk stratification model for patients undergoing prolonged weaning.
• To inform clinical decision-making and resource allocation in weaning centers.

Main Methods:

• Retrospective analysis of 915 patients undergoing prolonged weaning at a specialist center (Dec 2008-Dec 2023).
• Categorization of patients into four outcomes: successful weaning (with/without non-invasive ventilation) and discharge with IMV or death.
• Multivariate logistic regression and Receiver Operating Characteristic (ROC) curve analysis to identify predictive indicators.

Main Results:

• Successful weaning occurred in 73.2% of patients. Outcomes included successful weaning (36.1%), successful weaning with non-invasive ventilation (37.1%), discharge with IMV (20.2%), and death (6.6%).
• Distinct indicators were identified for each outcome, including comorbidities, BMI, cause of IMV (e.g., AECOPD, neuromuscular disease), pre-existing conditions (COPD, malignancy), and complications (nosocomial pneumonia, CIP).
• ROC curves demonstrated acceptable predictive accuracy (AUC 0.72-0.79) for the identified indicators.

Conclusions:

• Specific patient characteristics and clinical factors predict different prolonged mechanical ventilation weaning outcomes.
• Early assessment of individual risk profiles can optimize outpatient care planning.
• Identifying predictors can shorten length of stay in weaning units and improve bed availability for new patients.