Predicting development of bronchopulmonary dysplasia in extremely preterm neonates using flow data from a mechanical ventilator

Abstract

Background and aim: Bronchopulmonary dysplasia (BPD) frequently complicates preterm birth and leads to significant long-term morbidity. Predictive tools determining an infant9s risk of developing BPD are needed, enabling tailored treatment for infants at high risk. The purpose of this pilot study was to investigate if simple lung mechanics obtained from flow data from a mechanical ventilator can predict BPD in extremely preterm infants. Methods: 21 preterm neonates with mean gestational age 25.9 weeks (25.3, 26.5) who needed mechanical ventilator support during their first day of life were studied. Raw bi-directional flow data from the ventilator were logged for minimum 20 minutes. By special software, the flow data files were converted to volume as a function of time, and after visual inspection of the traces and selection of minimum 100 consecutive breaths, different breathing parameters were calculated and averaged. 13 of the infants later developed moderate or severe BPD, while the remaining 8 developed no or only mild BPD. Results: Fe50 (flow at 50% exhaled volume as ratio of peak flow) differed significantly in the two groups, being larger in the moderate/severe BPD group: 68.0 (CI 52.0, 84.0) vs. 84.4 (CI 80.1, 88.6), p=0.011. The parameters Fe75, Tptef/Te and FVg did not differ significantly, but they all consistently tended to be lower in the no/mild BPD group. There were no differences in the two groups according to GA, birth weight, gender or ventilator support at the time of data logging. Conclusion: This pilot study suggests that flow data obtained from ventilators during the first hours of life may help to predict later BPD in preterm neonates.