Prognostic value of lung ultrasonography after on-pump cardiac surgery

Rationale : Along with bedside plain chest radiography, lung ultrasonography is being increasingly used for detection of postoperative respiratory complications. Aim : Our study was aimed at the evaluation of lung ultrasonography efficacy for the diagnosis of postoperative respiratory complications in patients after the on-pump cardiac surgery. Materials and methods : The study included 39 patients who had undergone elective cardiac on-pump surgery. Assessment of hemodynamic parameters and blood gases was done at admittance to the intensive care unit (ICU), as well as at 6 and 24 hours after surgery. Lung ultrasonography was also performed including counting of B-lines in 12 lung quadrants at 6 and 24 hours after surgery, as well as chest radiography at 24 hours. Duration of mechanical ventilation, time in ICU and in-hospital stay were also evaluated. Results : Gas exchange deterioration was associated with increased numbers of B-lines: 9 (5 to 15) at 24 hours after surgery. In the patients with PaO2/FiO2 above 300 mm Hg the number of B-lines at 24 hours after surgery was 4 (2 to 8) (р = 0.04). Plain chest radiography at 24 hours after surgery revealed abnormalities in 69% of the patients. Discoid atelectases were the most common findings (n = 13). The ROC analysis showed that increased numbers of B-lines above 10 at 6 hours after completion of the surgery was predictive of the development of X-ray abnormalities at 24 hours (AUC 0.82, р = 0.02, sensitivity 86%, specificity 76%). At 6 hours after the intervention the patients who subsequently required prolonged mechanical ventilation had increased numbers of B-lines (15 [14–27]) compared to those who could be extubated within the first 24 hours after surgery (10 [3–13], p = 0.02). Conclusion : Lung ultrasonography monitoring accelerates the diagnosis of respiratory problems after cardiac surgery and allows timely identification of the patients requiring prolonged respiratory support and ICU stay.