ObjectiveTo explore the value of diaphragm ultrasound score in predicting the success of mechanical ventilation withdrawal.MethodsSixty-four mechanically ventilated patients underwent bedside ultrasound

the changes of diaphragm thickness and bilateral diaphragmatic motion amplitudes during the spontaneous breathing trial（SBT）

and diaphragmatic-related shallow fast breathing index（D-RSBI）were recorded.According to whether the patients were successfully offline

they were divided into the successful group（39 cases）and the failed group（25 cases）.The ROC curve was drawn to analyze the diaphragmatic thickness rate

diaphragmatic motion amplitude and D-RSBI for predicting offline failure during SBT.According tothebestcriticalvalue

the related continuous variables were transformed into binary variables

the multivariate logistic regression analysis was carried out

the diaphragm ultrasound score table was established and the probability of failed withdrawal was predicted.ResultsThe differences in diaphragmatic thickening rate

diaphragmatic motion amplitude and D-RSBI between the successful and failed group were statistically significant（all P<0.05）. ROC curve showed that the optimal cutoff value of diaphragmatic thickening rate in successful group was 53%

the sensitivity was 89.6%

specificity was 45.4%

and Youden index was 0.283.Optimal cutoff value of diaphragmatic motion amplitude was 11.8 mm

the sensitivity was 62.3 %

specificity was79.4%

Youden index was 0.413.Optimal cutoff value of D-RSBI was 1.7 mm·min/n

the sensitivity was 58.6%

specificity was81.3%

and Youden index was 0.312.Diaphragm thickness rate

diaphragmatic motion amplitude and D-RSBI were independent risk factors for ventilator failure（OR=5.982

2.966

4.356）.According to the diaphragm ultrasound score table

25 patients were divided into 8 cases with low-risk（0～2 points）

11 cases with medium risk（3～6 points）and 6 cases with high-risk group（7～8 points）. The rate of failure to withdrawl were 11.4%

52.6% and 89.7% respectively.ConclusionDiaphragm ultrasound score table can accurately predict the probability of successful offline of mechanical ventilation.