Background: Evaluate the accuracy of IWI predicting weaning in patients with mechanical ventilation greater than 72 hours.
Methods: All patients were divided into two groups, according to FiO2 in the intensive care unit (ICU) with mechanically ventilated for more than 72 hours. Recorded Integrative weaning index (IWI) related indicators in all patients. Evaluate the accuracy of IWI predicting weaning success.
Results: Within these 50 patients, 32 of them were weaning successfully, and 18 of the patients failed. Using IWI >45.70 mL/cmH2O breaths/minute/liter as a threshold of predicting successful weaning, the sensitivity is 0.91, and the specificity is 0.83. The AUC of IWI was 0.91. In the group with FiO2 =0.40, 17 patients were successfully liberated from MV, while 8 patients failed. The sensitivity is 0.8235, and specificity is 0.88 using IWI >50.40 mL/cmH2O breath/minute/liter as a threshold for predicting the outcome of weaning. The AUC of IWI was 0.846±0.117. In the FiO2 =0.35 group, 15 patients were successfully liberated from MV, while 10 patients failed. Using IWI >39.33 mL/cmH2O breaths/minute/liter, as a threshold to predict successful weaning, results in a sensitivity of 0.93 and a specificity of 0.90. The AUC of IWI was 0.953±0.395.
Conclusions: Results showed IWI has a significantly higher AUC value compared with other traditional weaning indexes. Hence, it can be a significant predictor for weaning outcomes.
Keywords: Integrative weaning index (IWI); mechanical ventilation (MV); rapid shallow breathing index; weaning index; weaning screening test.