Changes In Perinatal Care And Predictors Of In-hospital Mortality For Very Low Birth Weight Premature Infants

Objective: Information of the outcome of very low birth weightpremature infants in developing regions is relatively scarce when comparedwith that in developed regions. To better guide local intervention program,it is essential that current and reliable statistics be collected.Methods: Data concerning perinatal characteristics, morbidity andin-hospital mortality of very low birth weight premature infants admitted toour department of neonatology during2002-2009was retrospectivelycollected. Changes of perinatal care between2002-2005and2006-2009were identified. Factors associated with in-hospital mortality wereidentified by logistic regression and a predictive score model wasestablished.Results:(1) A total of475cases were enrolled. In-hospital mortality decreasedfrom29.8%in2002-2005to28.1%in2006-2009(P＞0.05). More infantsborn very premature (with a gestational age＜28weeks) survived todischarge in the latter epoch (38.1%vs8.3%, P＜0.05). (2) Neonates admitted in2006-2009were more likely (P＜0.01) tohave been born by Caesarean section (47.0%vs36.3%), by a mother withpregnancy induced hypertension syndrome (19.1%vs8.6%), to havepersistent pulmonary hypertension of newborn (5.9%vs0.7%) andintraventricular hemorrhage (44.1%vs17.8%), to receive surfactant(17.6%vs4.6%) and stay longer in hospital (19.3d vs15.0d), but less likelyto receive intravenous immunoglobulin (19.4%vs34.4%).(3) Persistent pulmonary hypertension of newborn, pulmonaryhemorrhage, extremely low birth weight, gestational age＜33weeks, fedbefore3postnatal days and enteral feeding were found predictors ofin-hospital mortality by multivariate logistic regression analysis.(4) The discriminating ability of the predictive model was82.4%. Atthe cutoff point of-0.56, the sensitivity, specificity, positive and negativepredictive value was0.729,0.832,0.634and0.885respectively.Conclusions:(1) Very low birth weight premature neonates admitted in2006-2009did not have significantly reduced in-hospital mortality than thoseadmitted in2002-2005;(2) Infants with the score higher than-0.56were assessed to be at highrisk of in-hospital mortality.