Very Low Birth Weight Infants in a Portuguese Intensive Care Unit and the Vermont Oxford Network: 15 Years of Registry Data

Abstract

Introduction: Our neonatal service is part of a differentiated perinatal hospital and has contributed to the Vermont Oxford Network for more than 15 years. This data base includes data on the morbidity and mortality of newborns born in the member hospitals with birth weight between 401 and 1500 g and/ or from 22 to 29 weeks and six days of gestation, or those admitted to these hospitals with up to 28 days of age. It thus allows the analysis of clinical practice and its comparison with similar units. The goal of the present paper is to disclose some of our data from the past 15 years and to compare it with the Vermont Oxford Network data trying to identify areas of possible improvement and permitting other neonatal units to compare their data with our in a benchmarking process.
Material and Methods: Observational, retrospective study. It included newborns with birth weight ≤ 1500 g (very low birth weight newborns) born and treated at our hospital from 2001 to 2015. Descriptive data analysis, chi-square test and ANOVA, significance when p < 0.05.
Results: A total of 869 very low birth weight newborns were studied, median weight 1100 g and gestational age 29 weeks. Twinning was found in 37.6%. In the delivery room, 23% did not require any resuscitation, 52.2% of the newborns required invasive intubation, 78.3% had surfactant, and, since 2011, 29.7% have started noninvasive ventilation. Of the total very low birth weight newborns, 12.9% had oxygen therapy at 36 weeks of corrected age, 23% patent ductus arteriosus and late sepsis in 17.1%. There was higher neurological morbidity compared to the Vermont Oxford Network except in the case of retinopathy of prematurity. Overall mortality was 14% (122 newborns). The time of hospitalization was on average 52.7 ± 34.4 days. The 629 newborns that were discharged home had equivalent length of stay and head circumference measure but a lower weight than those in the Vermont Oxford Network, and 14.3% went home with exclusive breastfeeding.
Discussion: This work allowed us to study our very low birth weight newborns data and compare it with one of the largest neonatal world networks. Our population is similar from the point of view of gestational age, somatometric data, pregnancy surveillance rates and cesarean section with the most noticeable difference being the percentage of low birthweight for gestational age babies, twin pregnancies and antenatal corticosteroid treatment, superior in our center. Cardio-pulmonary and gastrointestinal disorders were overlapping. It is urgent to improve our rate of sepsis, neurologic sequelae, post-partum hypothermia control and neuroprotection with magnesium sulphate. The mortality rate and the length of stay at discharge was similar.
Conclusion: This study allowed us to compare our population of very low birth weight newborns with those registered in the network. We have verified that we have been accompanying the evolution of Neonatology over the past years and we have identified areas for improvement.