Using Positive End-Expiratory Pressure to Predict and Quantify Tracheobronchomalacia in Grade 3 Bronchopulmonary...

Tracheobronchomalacia (TBM) is common in Grade 3 Bronchopulmonary Dysplasia (BPD), but there is wide variation in frequency and method of evaluation. Objective quantitative methods are needed for translational studies and future drug trials. This study aimed to determine whether positive end-expiratory pressure (PEEP) and oxygen saturation index (OSI) can predict TBM in Grade 3 BPD.

We retrospectively reviewed flexible bronchoscopy videos from a cohort of infants with Grade 3 BPD to determine the presence of TBM. PEEP and mean OSI in the week before the first bronchoscopy were assessed as predictors of TBM using multiple logistic regression. Subject Demographics: The study included 56 infants with Grade 3 BPD with a median gestational age (GA) of 25.3 weeks (IQR 2.8) and median post-menstrual age (PMA) at first bronchoscopy of 40.7 weeks (IQR 6.1).

Thirty-seven subjects (66%) were found to have TBM, although 13 subjects lacked evidence of TBM on the first evaluation (false negative rate 35%). PEEP and OSI had 86% sensitivity for predicting TBM, but only 30% specificity (p = 0.02). In a post-hoc analysis of the 13 initial bronchoscopies that "missed" TBM, 10 had a higher level of sedation in the first evaluation compared to the later assessment that revealed TBM.

We found that PEEP and OSI are sensitive but not specific for predicting TBM in Grade 3 BPD. In addition, there are frequent false negative evaluations when bronchoscopy is performed under higher levels of sedation. We propose a more precise bronchoscopy method for quantifying TBM.