A prospective, blinded evaluation of indexes proposed to predict weaning from mechanical ventilation

OBJECTIVE: To conduct a blinded evaluation of the predictors of weaning from mechanical ventilation. DESIGN: A prospective clinical study. SETTING: A 23-bed general intensive care unit. PATIENTS: Ninety-three non-selected patients, ventilated for more than 48 h. METHODS: The study had two steps: at first, patients' data were used to select the cut-off value for weaning predictors (the minimal false classification). The cut-off value for each index was prospectively assessed in a group of 52 patients. The predictive performance of these indexes was evaluated by calculating the area under the receiver operating characteristic curve. In the prospective-validation set we used Bayes' theorem to assess the probability of each test in predicting weaning. The physicians making decisions about the weaning process were always unaware of the predictive values. Weaning was considered successful if spontaneous breathing was sustained for more than 48 h after extubation. MEASUREMENTS AND RESULTS: During the first 2 min after discontinuation of mechanical ventilation the following tests were performed: vital capacity, tidal volume, airway occlusion pressure (P(0.1)), minute ventilation, respiratory rate, maximal inspiratory pressure (MIP), respiratory frequency to tidal volume (f/V(T)), P(0.1)/MIP and P(0.1) x f/V(T). The areas under the curve showed that the tests had not the ability to distinguish between successful and unsuccessful weaning. CONCLUSION: Our results show that all the evaluated indexes are poor predictors of weaning outcome in a general intensive care unit population