Chronic lung disease and survival in 4 tertiary neonatal units.

Aim: To compare mortality and respiratory morbidity in preterm infants born at 4 United Kingdom centers during 1994 and 1995.
Method: Collection of CRIB scores, respiratory parameters and mortality rates from unit databases.
Results: Mortality in center A was 27% (actual number of deaths 36/135), in center B was 30% (39/130), in center C was 28% (51/182), in center D was 39% (60/156). The rate of chronic lung disease (36 week definition) in center A was 16%, in center B was 12%, in center C was 13%, in center D was 15%. The predicted number of deaths by CRIB scores in center A was 54 (95% confidence intervals 45-63), in center B was 33 (25-41), in center C was 53 (43-63), in center D was 46 (37-56).
Conclusion: Center A had a lower than predicted mortality. Center D had a higher than predicted mortality. There is an urgent need for a national neonatal database to allow comparison between center and to identify reasons for variation in outcomes.