ObjectiveTo evaluate the predictive value of the diaphragm ultrasound for weaning from mechanical ventilation.MethodsThe patients who received mechanical ventilation in Fujian Provincial Hospital between February 2016 to December 2017 and met the criteria for a T-tube spontaneous breathing trial were included in the study. Then right diaphragmatic displacement (DD) and diaphragmatic thickening fraction (DTF) were evaluated using M-mode ultrasonography as well as the rapid shallow breathing index (RSBI, the ratio of respiratory rate to tidal volume). A new index was named as the diaphragmatic-RSBI (D-RSBI, the ratio of respiratory rate to DD). The patients were classified into a success group or a failure group according to the weaning outcomes. The receiver operating characteristic (ROC) curves were calculated to evaluate the predictive performance of each index.ResultsFifty-nine patients were weaned successfully and failure of weaning was found in 29 patients. There were no statistically significant differences in pre-weaning parameters including age, sex, systolic blood pressure, diastolic blood pressure, blood lipid index (total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride), or fast blood glucose between the weaning success group and the weaning failure group (P>0.05), but there were statistically significant differences in body mass index and acute physiology and chronic health condition Ⅱ score between two groups (P<0.05). DD [(13.44±3.23)mm vs. (10.28±2.82)mm, DTF [(32.43±12.35)% vs. (27.64±5.77)%, P<0.05] and D-RSBI [(1.49±0.47) breaths·min–1·mm–1 vs. (2.55±0.87) breaths·min–1·mm–1, P<0.05] differed significantly between the weaning success group and the weaning failure group. A cutoff of DTF≥27.9% yielded a sensitivity of 98.3%, a specificity of 62.1%, and an area under the ROC curve (AUC) of 0.873. A cutoff of D-RSBI≤1.73 breaths·min–1·mm–1 yielded a sensitivity of 76.3%, a specificity of 93.1%, and an AUC of 0.887. By comparison, when RSBI was ≤50.9 breaths·min–1·mm–1, there was a sensitivity of 91.5%, a specificity of 86.2%, and an AUC of 0.927. There was no statistically significant difference in AUC between D-RSBI and RSBI (P>0.05).ConclusionsDiaphragm ultrasound is feasible to predict the outcome of weaning. DTF and D-RSBI are as same accurate as the traditional RSBI in predicting the weaning outcome, but more objective and suitable for clinical application.