A Prospective Study of Indexes Predicting the Outcome of Trials of Weaning from Mechanical Ventilation

Abstract

BACKGROUND: The traditional predictors of the outcome of weaning from mechanical ventilation--minute ventilation (VE) and maximal inspiratory pressure (Plmax)--are frequently inaccurate. We developed two new indexes: the first quantitates rapid shallow breathing as the ratio of respiratory frequency to tidal volume (f/VT), and the second is termed CROP, because it integrates thoracic compliance, respiratory rate, arterial oxygenation, and Plmax. METHODS: The threshold values for each index that discriminated best between a successful and an unsuccessful outcome of weaning were determined in 36 patients, and the predictive accuracy of these values was then tested prospectively in an additional 64 patients. Sensitivity and specificity were calculated, and the data were also analyzed with receiver-operating-characteristic (ROC) curves, in which the proportions of true positive results and false positive results are plotted against each other for a number of threshold values of an index; the area under the curve reflects the accuracy of the index. RESULTS: Sensitivity was highest for Plmax (1.00), followed closely by the f/VT ratio (0.97). Specificity was highest for the f/VT ratio (0.64) and lowest for Plmax (0.11). The f/VT ratio was the best predictor of successful weaning, and Plmax and the f/VT ratio were the best predictors of failure. The area under the ROC curve for the f/VT ratio (0.89) was larger than that under the curves for the CROP index (0.78, P less than 0.05), Plmax (0.61, P less than 0.001), and VE (0.40, P less than 0.001). CONCLUSIONS: Rapid shallow breathing, as reflected by the f/VT ratio, was the most accurate predictor of failure, and its absence the most accurate predictor of success, in weaning patients from mechanical ventilation.