Predictive value of pre-extubation lung ultrasound score for respiratory support escalation after delayed extubation in adult cardiac surgery patients

Post-extubation respiratory deterioration remains a concern in adult cardiac surgery patients requiring prolonged postoperative mechanical ventilation, even after a successful spontaneous breathing trial (SBT). We evaluated the association between pre-extubation lung ultrasound score (LUS) and escalation to non-invasive ventilation (NIV) within 48 h after extubation. This single-center prospective observational study included adult cardiac surgery patients who required invasive mechanical ventilation for > 24 h and successfully completed an SBT. LUS was assessed after the SBT and before extubation using a standardized 12-zone protocol, with a total score ranging from 0 to 36. The primary outcome was NIV escalation within 48 h after extubation. LUS was analyzed primarily as a continuous variable using logistic regression, and its discriminative performance was assessed using receiver operating characteristic (ROC) analysis. Prespecified LUS strata (< 11, 11–14, and > 14) were used for descriptive analyses. Among 120 patients, 32 (26.7%) required NIV escalation within 48 h after extubation. NIV escalation occurred in 7.9%, 30.6%, and 42.4% of patients with LUS < 11, 11–14, and > 14, respectively ( P = 0.003). Interobserver agreement was high (Kendall’s coefficient of concordance, 0.882). Each 1-point increase in LUS was independently associated with NIV escalation (adjusted odds ratio, 1.346; 95% confidence interval, 1.165–1.555; P < 0.001). The area under the receiver operating characteristic curve was 0.777. A cohort-specific cutoff of 12.5 yielded 87.5% sensitivity and 60.2% specificity. In adult cardiac surgery patients requiring prolonged postoperative mechanical ventilation, a higher pre-extubation LUS was associated with an increased likelihood of NIV escalation within 48 h after extubation. LUS provided moderate discrimination and may complement clinical assessment during post-extubation risk stratification. These exploratory findings, including the ROC-derived cutoff, require external validation before clinical implementation.