Ten-year trends in neonatal assisted ventilation of very low-birthweight infants.

OBJECTIVES: To examine ventilatory support for the VLBW infant over the past 10 years in a single academic NICU and determine factors that predicted length of ventilation, death, and CLD. STUDY DESIGN: A retrospective cohort review of neonatal blood gases, ventilatory support, and clinical outcomes. RESULTS: From 1992 through October 2002, 6254 infants were admitted, of whom 2388 required intubation for mechanical ventilation. Of these, 837 were <1500 g at birth (VLBW) infants and 453 were less than 1000 g (ELBW). Total duration of ventilation decreased in all weight groups. Noninvasive ventilatory support increased from 20 to 55% of total ventilation from 1997 to 2002. During this same period, CLD decreased from 20 to 11% in ventilated VLBW infants. Duration of total ventilation was best predicted by birth weight, with each 100 g increment decreasing the duration of ventilation by 71 hours. Lower birth weight, male sex, and a longer total duration of ventilatory support were significant factors in predicting the occurrence of CLD. Death alone was best predicted by lower birth weight and maximum oxygen index (OI). Transported infants had significantly increased maximal OIs, durations of ventilation, and incidence of death. A total of 48% of infants with a single OI >10 either died or survived with CLD. CONCLUSIONS: Birth weight is the best predictor of duration of ventilation, and CLD is best predicted by birth weight, duration of ventilation and male sex. The increasing use of noninvasive strategies has not been associated with an observable increase in respiratory morbidity. VLBW infants with a single OI>10 may benefit from inclusion in future interventional rescue studies.