OBJECTIVES: Numerous reports indicate differing outcomes regarding chronic O2 use among NICUs for very low birth weight newborns. Significant confounders include differing case mix, lack of adjustment for baseline illness severity, and distortions from losses due to back transfer. We measured these to allow fair comparison among NICUs. METHODS: O2 utilization was followed in all surviving neonates<31 weeks gestation from their initial admission at 6 perinatal centers to discharge home over a 21 month period (10/94-6/96)(n=1288). To adjust for differences in population risk, we controlled for birth weight (BW), gestational age (GA), and day 1 severity of illness (SNAP) as well as race, gender, and antenatal steroid use. We analyzed 2 GA strata separately. RESULTS: See table, *P <.05. SNAP differed significantly among sites. After adjustment, sites A and C were more likely to use O2 at 28 days and 36 weeks PCA. CONCLUSION: This study demonstrates that: 1) day 28 O2 use in babies < 27 weeks does not vary between NICUs. 2) O2 use at 36 wks is common but differs significantly among NICUs. 3) To compare outcomes between NICUs base line severity needs to be considered.SPECULATION: 1)Differences in O2 use in babies >26 wks GA may represent differences in practice style among NICUs. 2)Earlier O2 discontinuation at site F suggests practice strategies that deserve investigation.

Table 1 No caption available.
Full size table