Change in Diaphragmatic Thickness During the Respiratory Cycle Predicts Extubation Success at Various Levels of Pressure Support Ventilation

Ultrasonographic assessment of diaphragm function with patients on low levels of pressure support (PS) predicts extubation outcomes, but similar information regarding extubation success under other conditions is lacking. The purpose of this study was to determine whether ultrasound (US) measurements of the diaphragm made on various levels of PS can predict time until successful extubation. Fifty-six intubated patients underwent ultrasound of the right hemidiaphragm during a PS wean at varying levels of pressure support (PS 5/5 cm of H2O, 10/5 cm of H2O, and 15/5 cm of H2O). The diaphragm was visualized using a 7.5-10 mHz transducer in the zone of apposition of the diaphragm to the lower rib cage. The percent change in diaphragm thickness between end-expiration and end-inspiration (a dagger tdi%) was calculated at each level of PS. a dagger tdi% > 20 is a robust predictor of extubation success within 48 h of US at PS 5/5 cm of H2O and 10/5 cm of H2O (sensitivity 84.6 and 88.9 % and specificity 79.0 and 75.0 %, respectively). At PS greater than 10/5 cm of H2O, its predictive power was greatly diminished. Of nine patients who were extubated with a dagger tdi% below the cutoff, 66.6 % required emergent reintubation in the next two days. Diaphragm US is a valid predictor of extubation success at some but not all PS settings. Using a a dagger tdi% of 20 % on PS levels up to 10/5 cm of H2O may reduce both unnecessarily prolonged intubations and prevent emergent reintubations.