Original Article

A rapid shallow breathing index threshold of 85 best predicts extubation success in chronic obstructive pulmonary disease patients with hypercapnic respiratory failure

Reza Goharani, Amir Vahedian-Azimi, Iman H. Galal, Leonardo Cordeiro de Souza, Behrooz Farzanegan, Farshid R. Bashar, Michele Vitacca, Seyedpouzhia Shojaei, Seyed M. M. Mosavinasab, Shunsuke Takaki, Andrew C. Miller


Background: The rapid shallow breathing index (RSBI) is used clinically to help predict a patient’s likelihood of successful liberation from mechanical ventilation (MV). However, the traditional threshold (<105 breaths/min/L) may underperform in patients with chronic obstructive pulmonary disease (COPD). We sought to determine the optimal RSBI threshold for COPD patients to improve the diagnostic accuracy for predicting successful ventilator liberation.
Methods: This was a prospective observational multicenter study of COPD patients [according to Global initiative for Chronic Obstructive Lung Disease (GOLD) criteria] admitted to the Medical ICUs of eight academic medical centers. All patients were intubated for hypercapnic respiratory failure and met the American Thoracic Society/European Respiratory Society guidelines to participate in a weaning trial. Ventilator weaning was conducted according to a defined protocol. RSBI was measured through the ventilator after 120 minutes of spontaneous breathing trial (SBT).
Results: Ninety patients were included (39 males and 51 females). Forty-three patients (48%) were successfully extubated whereas 47 patients (52%) failed extubation. Significant differences were observed between groups for duration-of-intubation [duration of intubation (DoI); P<0.0001], spontaneous tidal volume (VT) (P=0.03), and the ratio of dynamic-to-static compliance (P=0.005). The RSBI threshold of ≤85 breaths/min/L performed best: area under curve (AUC) receiver operating characteristic (ROC) curves 0.91, sensitivity 95.6%, specificity 90.4%, positive predictive value (PPV) 95.5%, and negative predictive value (NPV) 90.6%., positive likelihood ratio (LR+) 5.48, negative likelihood ratio (LR−) 0.25, and the diagnostic accuracy 91.7%. In post-ROC analyses, DoI and hospital length-of-stay (LOS) did not impact performance.
Conclusions: In COPD patients intubated with hypercapnia, RSBI ≤85 breaths/min/L outperformed the widely used threshold <105 breaths/min/L, yielding a 95.5% probability of extubation success, independent of ventilation duration or hospital LOS.

Download Citation