This study was performed between March 2017 and June 2019 in the affiliated Changshu hospital of Xuzhou medical university. All the patients were recruited from the medical intensive care unit. The participants were enrolled into the study when they were ready for weaning from MV. Prior to the study, ethical approval for this study was obtained from the ethics committee of affiliated Changshu hospital of Xuzhou medical university (ID number: 0056-2017). All the patients or their relatives gave written informed consent prior to the investigation.
2.1 Patient selection
Inclusion criteria included age ≥ 18 years, and MV for more than 24 h. The readiness of the patients for weaning largely dependent on their ability to meet all the criteria: PaO2/FIO2> 200 mmHg, using PEEP ≤5 cmH2O and fraction of inspired oxygen (FIO2) ≤0.4, stable hemodynamic (heart rate ≤140 beats/min, absence or low-dose vasopressors required i.e., dopamine or dobutamine ＜5-10 ug/kg.min), good consciousness (Glasgow coma scale ≥13), stable metabolism (i.e.，acceptable electrolyte levels), effectiveness of cough strength (positive white card test result), and in the recovery stage of disease.
2.2 Exclusion Criteria
Exclusion criteria included age less than 18 years, and MV for less than 24 h, pregnancy, end-stage tumor, neuromuscular disease, presence of pneumothorax, flail chest, large pleural effusion, patients with severe deformity, obesity, organ failure before weaning, lack of informed consent.
2.3 Study Design
The patients who met the aforementioned inclusion criteria were injected intravenously with remifentanil 0.2 mg for brief deep sedation. Next, when patients were awakened, they were put on triggering assisted breaths on volume support ventilation of MV with Vt 6 ml/kg of ideal body weight and a PEEP of 0 cmH2O. Respiratory mechanical parameters (Pplat and PEEPtot) were recorded. Pplat and PEEPtot were measured by holding breath for 2s after inhalation and exhalation, respectively. The DP was measured as Pplat minus PEEPtot. The treating physician evaluated each case individually. When the study subjects were deemed ready for an SBT with the intent to weaning, each participant was disconnected from the ventilator to SBT for 1 h using T piece with oxygen supplementation (FIO2 of 0.25–0.5) to achieve pulse oxygen saturation (SpO2) ＞ 90%. Then, attending physicians who were not involved in this study obtained the maximal RR by ECG monitor(PHILIPS IntelliVue MP30) during the 3-min SBT and recorded it in progress note. After that, the DP multiplies by RR was calculated by investigator. Notably, patients who passed the 1 h SBT were extubated, while others were reconnected to the ventilator with the previous ventilatory parameters if failed to the SBT. The decision to continue with the SBT , use noninvasive ventilation or reintubate after extubation was left up to the discretion of attending physicians who were not allowed to know the research indicators.
Intolerance signs：①RR ＞35 or＜8 breaths/min；②Heart rate ＜140 beats/min or a 20% increase or decrease from baseline；③SBP＞ 160 mmHg or ＜ 90 mmHg；④Arterial blood oxygen saturation ＞90%；⑤Sweating, agitation, decreased level of consciousness.
2.4 Weaning failure definition
Weaning failure was defined as either the failure of SBT or the need for reintubation within 48 h following extubation. Patients who failed the SBT, or required reintubation or noninvasive ventilation (NIV) within 48 h of discontinuation of MV were considered as having failed to wean.
2.5 Statistical analysis
Normally distributed continuous variables were expressed as means with standard deviations. The variables were compared with Student’s t test for independent samples. Non-normally distributed continuous variables were compared via Mann–Whitney U test, while categorical variables were compared with the Chi-Square Test or correction of continuity. Receiver operating characteristic (ROC) curve analysis was performed to assess the capability of DP×RR and DP to predict patients who may succeed at weaning or fail. Statistical analysis was performed with Statistical package for social sciences (SPSS) 17.0 software (SPSS, Chicago, IL, USA). A p values less than 0.05 were chosen to indicate the statistical significance.