Background: The threshold values of rapid shallow breathing index (RSBI) were compared in pressure support ventilation (PSV) and T-piece assessments for spontaneous breathing trials (SBT). The ability of ΔRSBI to also predict successful weaning was evaluated.
Methods: Two hundred eight patients were weaned from mechanical ventilation through oral intubation. They were randomly divided into PSV group (n = 93) and T-piece group (n = 115). The RSBI was calculated as f/VT at SBT of 3 and 30 minutes. Receiver operating characteristic curves of RSBI were also generated, cutoff values were determined and the changes in the RSBI were calculated.
Results: Of the 208 patients, 168 (80.77%) were successfully weaned from mechanical ventilation, 78/93 (83.9%) in the PSV group and 90/115 (78.3%) in the T-piece group (P > 0.05). In the PSV and T-piece groups, the average RSBI at 30-minute SBT was 67.18 ± 11.55 breaths per min/L and 99.11 ± 15.53 breaths per min/L, respectively (P < 0.01), and the average ΔRSBI was 69 ± 33% and 119 ± 35%, respectively (P < 0.01). Additionally, in the 2 groups, an RSBI of 75 breaths per min/L (PSV) and 100 breaths per min/L (T-piece) yielded a diagnostic accuracy of 87% and 82% respectively. However, a ΔRSBI of 90% (PSV) and 130% (T-piece) yielded a diagnostic accuracy of 82% and 77% respectively.
Conclusions: The threshold values of RSBI, thus 75 breaths per min/L (PSV) and 100 breaths per min/L (T-piece), for predicting successful weaning were more accurate than other values. Similarly, the change in the RSBI could also predict such successes.