ABSTRACT
PURPOSE: To evaluate and compare the performance of NT-proBNP levels, plasma protein concentration, hematocrit, and fluid balance for the preceding 24 h in predicting the outcome of the two steps of weaning: (1) spontaneous breathing trial (SBT), (2) extubation. METHODS: This was a prospective observational study of 143 patients who were mechanically ventilated for more than 48 h (55% COPD) and were ready to wean. They underwent an SBT and were extubated when they passed the trial. Immediately before the SBT, we measured the evaluated diagnosis tools. RESULTS: Of 143 patients, 80 (56%) passed the SBT and were extubated. Of these, two were reintubated for laryngeal dyspnea, 57 had no respiratory problem during the next 48 h, and 21 developed post-extubation respiratory distress (26%). Rescue noninvasive ventilation (NIV) prevented reintubation in 15 (71%). None of the tested diagnosis tools predicted the outcome of the SBT. Patients who developed post-extubation respiratory distress were older, had lower values of plasma protein concentration and higher values of NT-proBNP than those who did not. Only NT-proBNP was an independent predictor of the occurrence of post-extubation respiratory distress (OR 1.2; 95% CI 1.09-1.4; p = 0.003); the area under the ROC curve for NT-proBNP to predict post-extubation respiratory distress was 0.78 (95% CI 0.67-0.89; p = 0.0001). NT-proBNP was more accurate to rule out (negative likelihood ratio 0.09 for a cutoff of no greater than 1,000 pg/ml) than to rule in the risk of post-extubation respiratory distress (positive likelihood ratio 3.45 for a cutoff of at least 2,000 pg/ml). CONCLUSION: NT-proBNP levels at SBT help in the prediction of post-extubation respiratory distress and could identify the subgroup of extubated patients requiring close observation and/or prophylactic NIV.
