Predictive Value of Rapid Shallow Breathing Index Measured at Initiation and Termination of a 2-hour Spontaneous Breathing Trial for Weaning Outcome in ICU Patients

Abstract

The rapid shallow breathing index (RSBI) is a weaning parameter usually measured at the start of a spontaneous breathing trial (SBT). This study investigated the value of RSBI measured at the beginning and termination of SBT as a predictor of weaning outcome. RSBI was measured during the initial 1 minute (RSBI1) and at termination (RSBI2) of an SBT in 172 patients recovering from acute respiratory failure. Weaning was successful in 106 patients and failed in 66 patients. Among the 66 patients with weaning failure, 12 required reintubation within 48 hours (extubation failure), and the remaining 54 patients could not be extubated after SBT (trial failure). There were no differences between RSBI1 in the three groups (69.4 +/- 27.5, 81.7 +/- 24.4 and 75.5 +/- 26.5, respectively), but RSBI2 was significantly higher in patients with extubation failure (95.9 +/- 20.6) and trial failure (98.0 +/- 50.0) than in patients with weaning success (64.6 +/- 26.3) (both p < 0.001). Logistic regression revealed that RSBI2 was superior to RSBI1 and various physiologic indices in predicting weaning outcome. For the 118 extubated patients, the mean area under the receiver operating characteristic curve for RSBI2 and RSBI1 was 0.83 and 0.63, respectively. Using a threshold value of 105, the sensitivity, specificity, accuracy and likelihood ratio for weaning outcome were 0.91, 0.25, 0.85 and 1.38 for RSBI2 and 0.89, 0.16, 0.60 and 1.06 for RSBI1, respectively. This study found that RSBI measured at the completion of SBT was superior to that measured at the start in predicting weaning outcome in critically ill patients.