Background Advances in perinatal care have made it possible to improve survival of infants with low birth weight. The aims of this study were to analyse the clinical impact of morbidities and to identify predictors of in-hospital mortality in preterm infants with low birth weight.

Methods Between January-2011 and November-2012 were included 81 preterm infants at our centre with low birth weight or ≤ 32 weeks gestation. Perinatal variables were included in risk factor analysis. Data are expressed as gestational age (GA) <28 or between 28–32 weeks and birth weight defined as: low <2500 grams, very low 1000–1500 grams and extremely <1000 g. Results: The mean GA was 29 ± 2 (23–32 weeks) and mean birth weight was 1230.8 ± 368 (510–2000 g). The neonatal mortality rate was 17.3%. Preterm infants who died had lower birth weight than were alive, 797 ± 249 vs. 1332 ± 315, p < 0,001. The overall incidence of respiratory distress syndrome was 86.3%, septicaemia 24.7%, neurological damage 18.5% and necrotizing enterocolitis was 7.4%. The SNAP II, SNAPPE II and CRIB II scores showed a high discriminatory power for predicting hospital mortality, ROC area 0.863, 0.925 and 0.925, p < 0.001, respectively. Multivariate analysis of predictors of in-hospital mortality were necrotizing enterocolitis, risk scores, low 5-min Apgar score, inotropic support, and protectors were: the absence of intraventricular haemorrhage, cardiopulmonary resuscitation and increased GA.

Conclusion The survival in preterm infants in addition of GA or birth weight, it depends on the presence of morbidities. The use of risk scores on admission is useful for prediction in-hospital mortality.