Abstract

Background: Previous studies comparing different models of neonatal intensive care have generally not been population based. The objective of this study was to compare the perinatal services of two total populations.

Methods: Observational study based on two geographically defined populations: the whole of Demark (some centralisation of neonatal intensive care but most delivered locally by small perinatal centres—48 in total) and the Trent Health Region of the UK (no formal centralisation however deliveries almost all focussed on 16 major hospitals with > 90% of the intensive care provided by 13 hospitals). Information was recorded about the course of every liveborn infant < 28 weeks gestation and or < 1000g birth weight and ≥ 21 weeks gestation in 1994 and 1995.

Results: Despite having a smaller population the number of liveborn children meeting the study criteria was significantly higher in Trent (Demark 461 (3.3 per 1000 births, 95% confidence interval (CI) 3.0 to 3.6); Trent 572 (4.9 per 1000 births, 95% CI 4.5 to 5.3)). In Denmark 91.1% of these infants were admitted for intensive care and 85.5% in Trent. Despite significantly more Trent infants being exposed to antenatal steroids their outcome was worse (median Clinical Risk Index for Babies (CRIB) score 7 v 4; proportion receiving ventilation 87.6% v 40.0% ; survival to discharge (uncorrected for disease severity) 42.3% v 35.0%).

Conclusion: The population characteristics of Trent seemed to produce a higher prematurity rate compared to Denmark. These infants as a group appeared sicker and, despite more intensive care delivered by a more specialised service, outcomes were worse.