Predictive accuracy of lung and diaphragmatic ultrasound in weaning from mechanical ventilation: a comparison with the Rapid Shallow Breathing Index

Predicting weaning outcomes from mechanical ventilation remains a clinical challenge. Conventional indices such as the Rapid Shallow Breathing Index (RSBI) have limitations. This study evaluates diaphragmatic ultrasound parameters alongside RSBI and modified lung ultrasound score (mLUS) to improve the prediction of weaning success. A total of 50 adult patients requiring invasive mechanical ventilation for more than 24 hours were prospectively enrolled. All underwent a spontaneous breathing trial (SBT), and parameters including RSBI, mLUS, diaphragmatic excursion (DE), and diaphragmatic thickening index (DTI) were recorded. Weaning outcome was defined as successful extubation without need for reintubation within 48 hours. Associations were analyzed using receiver operating characteristic curves and multivariate logistic regression. Of the 50 patients, 34 (68%) passed SBT, and 27 (54%) had successful weaning. RSBI and mLUS were significantly lower, and DE and DTI significantly higher, in the weaning success group. Among all, DTI showed the highest predictive value (area under the curve: 0.948). On multivariate regression, DTI and DE were independent predictors of weaning success. A combination of RSBI and DTI yielded the highest diagnostic accuracy (94%). Diaphragmatic ultrasound parameters, particularly DTI, serve as strong, non-invasive predictors of weaning success. Integration of DTI with conventional indices like RSBI enhances predictive accuracy and may be valuable in guiding weaning protocols.