Summary

Background: A substantial number of premature deliveries occur in hospitals lacking neonatal intensive care facilities. We previously demonstrated a comparable outcome of very low birth weight infants delivered in a level II nursery to that of inborn infants delivered in our tertiary care center, but a similar comparison of extremely low birth weight infants has not been done.

Objectives: To compare the neonatal outcome (mortality, severe intraventricular hemorrhage/periventricular leukomalacia, bronchopulmonary dysplasia and intact survival) of inborn and outborn ELBW[1] infants, accounting for sociodemographic, obstetric and perinatal variables.

Methods: We compared 97 ELBW infants (birth weight ≤ 1000 g.) delivered between the years 2000 and 2004 in a hospital providing neonatal intensive care to 53 ELBW babies delivered in a referring hospital. A univariate model was first applied to examine the associations of the individual independent variables with the outcome variable, followed by a logistic stepwise regression analysis for each of the outcome variables. The odds ratios for each predictor were reported as well as their P values and 95% confidence intervals.

Results: In the stepwise logistic regression analysis, accounting for a possible confounding effect of the independent variables, ‘hospital of birth’ remained a statistically significant predictor in the final step only for mortality, with odds ratio (inborns relative to outborns) of 3.32 (95%CI[2] 1.19–9.28, P = 0.022). No statistically significant associations with the other outcome variables were found (severe IVH[3]/PVL[4] odds ratio = 1.99, 95%CI = 0.77–5.14, P = 0.155; BPD[5] odds ratio = 0.60, 95%CI = 0.19–1.91, P = 0.384; intact survival OR[6] = 0.56, 95%CI = 0.23–1.35, P = 0.195).