To wean respiratory support, preterm infants with severe respiratory failure are often administered systemic corticosteroids. We sought to evaluate if postnatal age or clinical characteristics predicted death or tracheostomy following systemic dexamethasone in evolving bronchopulmonary dysplasia. We performed a retrospective study of infants born at ≤ 30 weeks' gestational age cared for at a Level IV referral center from 2009 to 2019 who received a complete course of systemic dexamethasone beyond 4 weeks of age for the indication of preventing death and/or liberating from positive pressure ventilation. Clinical variables at the initiation of dexamethasone were examined by regression analysis, and receiver operator curves were constructed for their ability to predict death or tracheostomy. Of the 81 infants that received a complete course of dexamethasone, 52% died or required tracheostomy. At onset of the dexamethasone course, infants ultimately requiring tracheostomy were more likely to be small for gestational age (SGA) (p = 0.02), receive dexamethasone at a significantly later postmenstrual age (p < 0.01), require higher respiratory support (p < 0.01), and have concomitant pulmonary hypertension (p < 0.01). The combination of SGA, a history of late bacterial sepsis, the need for multiple dexamethasone courses, a diagnosis of pulmonary hypertension, and higher respiratory support at onset of dexamethasone course predicted tracheostomy or death (AUC 0.87). For infants receiving systemic dexamethasone for severe respiratory failure, a combination of clinical characteristics and magnitude of respiratory support may reliably predict inability to wean from positive pressure ventilation, potentially avoiding ineffective dexamethasone exposure. Further validation is required before clinical use.
House et al. (Mon,) studied this question.