Early Prediction of Bronchopulmonary Dysplasia in Very Low Birth Weight Infants with Mechanical Ventilation in the First Week of Life

Choi, Eun Na; Namgung, Ran; Koo, Hoe Kyoung; Park, Min Soo; Park, Kook In; Lee, Chul

Abstract

PURPOSE
Multifactorial in pathogenesis, bronchopulmonary dysplasia is difficult to predict based on any single factor, especially early in life. We evaluated clinical and ventilatory parameters in the first week of life, and their combinations were tested for early prediction of BPD. METHODS: Eighty-one very low birth weight (VLBW) infants born at gestational ages<32 weeks and mechanically ventilated for at least first 7 days were selected and classified into BPD (n=48, radiological findings and oxygen dependency at postconceptional age of 36 weeks) and non-BPD (n=33) groups. Clinical and ventilatory parameters on days 1, 4 and 7 were analyzed. Chi-square and t-test were used to compare individual variables between two groups. Multiple logistic regression analysis was done to identify risk factors for BPD. RESULTS: The mean gestational age and birth weight were lower in BPD group. In multivariate analysis, significant risk factors that predict BPD were gestational age<30 weeks (OR 0.112, 95% CI 0.016-0.767), maximum MAP > or =4.5 on day 7 (OR 3.982, 95% CI 1.046-15.162) and maximum FiO2> or =0.3 on day 7 (OR 7.626, 95% CI 1.570-37.054). The combination of these factors for prediction of BPD had a 79% positive predictive value with an 85% sensitivity. CONCLUSION: A number of clinical and ventilatory parameters in combination (gestation, maximum MAP and FiO2 on day 7) can predict BPD in VLBW infants early in life with a relatively high sensitivity and positive predictiveness. With early identification of infants prone to BPD, the clinicians may resort to more active measures to minimize lung injury and to prevent BPD.