Methods: The Score for Neonatal Acute Physiology (SNAP) measures neonatal illness severity and is highly predictive of mortality risk. We compared inborn mortality rates between two cohorts of infants <1500 gm; a 1989-90 cohort from 2 NICUs and a 1994-95 cohort from 6 NICUs (2 overlapping). Surfactant was available in both eras. Deaths receiving comfort care only were excluded (18/329 in 89-90 [5.5%], 6/583 in 94-95 [1%]).

Results: Overall mortality by cohort was 17% vs 9% (p<.001). Overlapping sites showed significant improvement similar to all 6 sites. Admission illness severity was no different in the two eras (±SE):<750g (16.9 ± 1.1 vs 17.8 ±.7), 750-999g (12.5 ±.7 vs 11.3 ±.6), 1000-1499g (9.3 ±.4 vs 8.7 ±.3).Table

Table 1
Full size table

Conclusions: 1) SNAP remains highly predictive of mortality risk, 2) Average admission severity is the same in 1990 and 1995, 3) There is a highly significant decline in mortality in sicker infants (SNAP>9, p<.001). This is strong evidence of continued rapid improvements in NICU care.