Diaphragm dysfunction predicts weaning outcome after bilateral lung transplant

Background: Diaphragm dysfunction and its effects on outcomes of ventilator weaning have been evaluated in mixed critical care populations using diaphragm thickening fraction (TFdi; the ratio of the difference between ultrasound diaphragm thickness at end-inspiration and end-expiration to diaphragm thickness at end-expiration) or neuro-ventilatory efficiency (NVE; the ratio of tidal volume and peak electrical activity of the diaphragm). Such data are not available in bilateral-lung transplant recipients. We hypothesized that: i) diaphragm dysfunction, as defined by a TFdi<29%, is more likely to occur in difficult weaning; ii) TFdi and NVE predict weaning outcome; iii) and duration of mechanical ventilation before the first spontaneous breathing trial (SBT) is associated with diaphragm dysfunction. Methods: Adult bilateral-lung transplant patients admitted to our ICU were screened at the time of the first SBT (pressure-support of 5cmH2O and zero positive end-expiratory pressure). At the fifth minute, TFdi and NVE were measured over three respiratory cycles. Weaning was classified as simple, difficult or prolonged (successful extubation at the first SBT, within three or after three SBTs, respectively). Results: Forty-four subjects were enrolled. Diaphragm dysfunction occurred in 14 subjects (32%); all of whom had difficult weaning (78% of the subgroup of 18 patients experiencing difficult weaning). Both TFdi (24[20-29] vs 39[35-45]%) and NVE (34[26-45] vs 55[43-62]ml/µV) were lower in difficult weaning (both p<0.001). The areas under the receiver operator curve predicting difficult weaning were 0.88 [95%CI 0.73-0.99] for TFdi, and 0.85 [95%CI 0.71-0.95] for NVE. Duration of ventilation demonstrated a linear inverse correlation with both TFdi and NVE. Conclusions: Diaphragm dysfunction is common following bilateral-lung transplantation and associated with difficult weaning. In such patients, average values for TFdi and NVE were reduced compared to patients with simple weaning. Both parameters showed similar accuracy for predicting success of ventilator weaning, demonstrating an inverse relationship with duration of ventilation.