Diaphragmatic ultrasonography as an adjunct predictor tool of weaning success in patients with difficult and prolonged weaning

BACKGROUND: Diaphragmatic ultrasonography has been recently considered as a new weaning predictor method. Previous studies checked diaphragmatic excursion and thickness during quiet or deep breathing in unselected populations of critically ill patients. Our study aimed to investigate diaphragmatic excursion during quiet and unassisted breathing, in comparison to standard predictor tools, such as Rapid Shallow Breathing Index (RSBI) and Maximal Inspiratory Pressure (Pimax), in patients with difficult and/or prolonged weaning. METHODS: Patients with difficult and/or prolonged weaning, who met the criteria for spontaneous breathing trial, were assessed. The excursion of each hemidiaphragm (DEx) was evaluated by B-mode and M-mode ultrasonography while patient was on quiet breathing and at supine position. RSBI and Pimax were simultaneously recorded and weaning outcome was recorded. RESULTS: Thirteen male and fourteen female patients were included. DEx [median and interquartile range, mm] was 14 (8.5-22) for the right hemidiaphragm (RDEx) and 12 (7-23) for the left (LDEx). We found no difference in DEx between sexes. Among the four weaning predictor tools, LDEx at a cut-off 10 mm was the best index to predict weaning success (sensitivity 86%, specificity 85%, Negative Predictive Value 94%). The optimal cut-off values, as determined by the area under the receiver operating characteristic curve, were 10 mm for RDEx, 7 mm for LDEx, 57 breaths/min/L for RSBI and -20 cmH(2)O for Pimax. CONCLUSIONS: Our results suggest that DEx threshold of 10 mm and 7 mm for right and left hemidiaphragms respectively could be used as adjunct tool in the predictive algorithm of weaning in difficult to wean patients.