New Integrated Weaning Indexes

The study was divided into two parts: First, the threshold values of the indexes that discriminated best between successful weaning and weaning failure were determined in training set (n = 208). Second, the accuracy of each index was scrutinized prospectively in an additional group of patients (prospective-validation data set, n = 967). One thousand one hundred and seventy-five patients who were on mechanical ventilation for more than 24 hours were evaluated.Informed consent was obtained from each patient, whenever possible, or from the patient's next of kin. The samples were selected using a simple random sampling method in which the entry number of eligible patients was considered to be the sampling ID number. Using Random Allocation Software (RAS), random numbers were generated to select the study participant.

The ventilators used were Evita XL and Evita 4 edition (Draeger, Lubeck, Germany).

According to a protocol that have introduced in two studies, the protocol established. All of the mechanically ventilated patients undergoing assessment for extubation undergo a daily SBT if predetermined eligibility criteria for weaning are met . According to this protocol, sedation would be discontinued before evaluation of weaning. Patients who meet these criteria are initially placed on SBT (continuous positive airway pressure of 5 mmHg, FiO2≤0.4) for 3 minutes to obtain weaning parameters. If the Oxygen saturation is ≥ 92% on pulse oximetry with FiO2≤0.4 and RSBI <104 breaths/minute/L, patients are continued on the SBT for 30 to 120 minutes. All clinical and ventilator parameters are monitored closely for signs of respiratory distress (respiratory rate > 30 breaths/minute, SaO2< 90%, heart rate > 140 breaths/minute, or a sustained increase or decrease of heart rate of >20%, blood pressure>200 mm Hg or <80 mm Hg, and agitation, diaphoresis, or anxiety). At the end of the SBT, the RSBI is measured again, ABG is obtained, and the predetermined values are calculated and measured. The decision to return to mechanical ventilation made by the two subspecialists in pulmonary and lung diseases, who were chief supervisor of both ICU in two cities (who were completely blind to the study and the results of the indexes evaluated), based on airway competence (cough, sputum production, neurologic status, level of consciousness, and maximal inspiratory pressure). Patients who remain extubated at 48 hours are classified as having had a successful extubation. Baseline demographics, initial diagnosis, and pre-extubation clinical, ventilator, laboratory, and radiographic data are collected for each patient.

To assess the prognostic value of new indices, receiver operator characteristic (ROC) curves was drawn for each formula in addition for other conventional criteria used in this area. For each ROC curve analysis, sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value, diagnostic accuracy, likelihood ratio of a positive test (LR+) and the likelihood ratio of a negative test (LR-), probability for weaning success when test is positive and probability for weaning success when test is negative of the indexes were used to predict the weaning outcome in the prospective-validation data set.

Hanley and McNeil method was used to calculate the area under curve (AUC) for each formula and then the methods developed by the same authors were used to compare these AUCs. Bayes' theorem was used to compute the weaning outcome the probability for weaning success when test is positive and negative (post-test probability) in the prospective-validation data set to assess the perform