In this 1000-bed primary teaching hospital, the study was conducted in the 8-bed adult general ICU over the course of 13 months. All patients enrolled in this study were mechanically ventilated for more than 48 hrs via endotracheal tubes during the study period. The investigation was approved by the hospital ethics committee, and informed written consent was obtained from the next of kin of every patient. Patients were ventilated in pressure support (PS) during the entire weaning period. The levels of inspiratory PS and positive end-expiratory pressure （PEEP) were progressively reduced depending on the patient's clinical assessment and blood gas values. Patients enrolled in this study had to meet all of the following readiness criteria: show significant improvement or resolution of the underlying cause of acute respiratory failure; be fully awake; require bronchial toilet less than twice in the 4 hours proceeding the assessment; stable hemodynamics without further need of vasoactive agents; arterial oxygen tension (PaO2)/fraction of inspired oxygen (FiO2) ratio \>200 at a PEEP of 4.0 cm H2O with a maximal FiO2 of 0.40; core temperature \<38.0℃; systolic blood pressure \>90 mmHg; respiratory rate to tidal volume ratio (RVR) \<105 breaths/min/L. The RVR was calculated after 1 min of spontaneous breathing.The maximal inspiratory PS of 12 cm H2O and no mandatory machine breaths supplied from the ventilator.
Study protocol The investigators investigated the weaning process with and without SBT. All patients were continuously assessed according to the readiness criteria and were screened for enrollment once a day. As soon as the patient was ready for weaning, he or she was randomly assigned to SBT or no-SBT study groups. Randomization was carried out in a blinded fashion using opaque and sealed envelopes. All patients in the study breathed through the ventilator circuit with flow-triggering (2L/min) and pressure support ventilation mode. In the SBT group, the patients underwent a 1 hr SBT with inspiratory PS of 7 cmH2O with other settings remaining constant (FiO2, PEEP, trigger sensitivity). For patients showing poor tolerance to the SBT, full ventilation support was immediately recommenced. This was defined by the following failure criteria: a decrease in oxygen saturation to \<90%; respiratory rate \>35/min for more than 5 min, in the presence of diaphoresis or thoraco-abdominal paradox; sustained increase in heart rate (\>140/min), or significant change in systolic blood pressure (\>180 or \<90 mmHg). The patients who tolerated the SBT underwent immediate extubation and received (breathed in) supplemental oxygen via a facemask. In no-SBT group, as soon as a patient met readiness criteria, he or she underwent extubation without SBT process and received supplemental oxygen via a facemask. Following extubation, noninvasive ventilatory support was introduced in the following circumstances: hypoxemia (SaO2\<90% for \>15mins) under receiving supplemental oxygen; presence of respiratory acidosis (arterial PH\<7.35 with arterial carbon dioxide tension (PaCO2)\>45mmHg; and respiratory rate \>25 breaths/min for 1 hr. The mode of ventilation was Bi-level Positive Airway Pressure (BiPAP). When such support was deemed inadequate (hypoxemia, hypercapnea, or respiratory distress), the patient was reintubated and mechanically ventilated.
The SBT and extubation were performed by two doctors who are members of the research team. Decisions regarding reintubation were made by doctors who were blinded for the treatment group. Extubation failure is defined as reintubation within 48 hrs. The reasons for reintubation were prospectively recorded.
Statistical Analysis Results are expressed as mean ± SD. Mean values of selected demographic variables and physiologic parameters of patients who underwent SBT were compared by student's t tests to those who underwent extubation directly. The differences in proportions between the two groups were determined using chi square test (X2-test).
