Prevalence, severity, and predictors of bronchopulmonary dysplasia in a cohort of very preterm infants

• Published in: Archives de pédiatrie : organe officiel de la Société française de pédiatrie Vol. 20; no. 9; pp. 928 - 937
• Main Authors: Richard, M, Ramful, D, Robillard, P-Y, Mussard, C, Loumouamou, Y, Ogier, M, Tasset, C, N'guyen, A-D, Alessandri, J-L, Sampériz, S, Gérardin, P
• Format: Journal Article
• Language: French
• Published: France 01-09-2013
• Subjects: Bronchopulmonary Dysplasia - epidemiology; Bronchopulmonary Dysplasia - therapy; Case-Control Studies; Cohort Studies; Continuous Positive Airway Pressure; Ductus Arteriosus, Patent - epidemiology; Ductus Arteriosus, Patent - therapy; Female; Fetal Growth Retardation - epidemiology; Gestational Age; Glucocorticoids - therapeutic use; Humans; Infant Nutrition Disorders - epidemiology; Infant, Newborn; Infant, Premature; Logistic Models; Male; Prevalence; Pulmonary Surfactants - therapeutic use; Retrospective Studies; Reunion - epidemiology; Risk Factors; Severity of Illness Index; Reunion
• ISSN: 1769-664X
• DOI: 10.1016/j.arcped.2013.05.002

The aim of this retrospective cohort study was to determine the prevalence and severity of BPD and its predictors in a regional cohort of very preterm (VP) infants in Reunion Island.
All autochthonous VP infants, live-born before the 33rd week of gestation (WG) between 1st January 2008 and 31st December 2009, were eligible for the study. Only VP infants surviving at least 28 days, for whom the parameters were known from birth, were included in the case-control study of predictors of moderate to severe BPD (BPDmo/s).
In VP infants less than 33 WG, the rate of overall BPD (3 grades of severity) was 30.7%. Among those who survived 28 days or more, the rate of BPDmo/s was 13.1% (95%CI: 10.2-15.9%). In VP infants less than 32 WG that survived at 36 WG, the prevalence of BPDmo/s was 18.2% (95% CI: 14.2-22.1%). In a fixed-effect logistic model, adjusted for gestational age, postnatal growth, and the mode of ventilation at 24h, 4 key factors were predictive of BDPmo/s: small for gestational age, surfactant, delayed energy intake, and late-onset neonatal infection. In a mixed-effect logistic model adjusted for these same cofactors, the site was associated with BPDmo/s, in line with a center-effect.
The prevalence of BPDmo/s in the mixed-race population of Reunion Island is consistent with those observed in Europe but were site-specific. In our setting, predictors of individual BPDmo/s are similar to those already identified.