Rationale: CAs contribute substantially to neonatal mortality, although inadequate classifications (descriptive or etiologic) have precluded a systematic analysis of their impact. This is essential when comparing NICUs on workload and performance. We developed a simple scoring system for mortality risk due to CAs and tested it on a multisite cohort of VLBW(<1500g) infants.

Methods: All VLBW admissions to 7 NICUs were prospectively abstracted from Oct. 94 to June 96, including birth weight(BW), demographics, illness severity (Score for Neonatal Acute Physiology, SNAP) and outcomes: mortality, discharge gestational age for survivors (GA@DC) and operations(OP#). Major CAs were grouped by organ system and classified as Non-, Possibly- and Acute- life threatening or as lethal. Multivariate mortality odds ratio (OR-M) was adjusted for BW and SNAP.

Results: 151/1750 infants had CAs (8.2%). Cardiovascular(27%), urogenital(21%) and respiratory (14%) accounted for most of CAs. Illness severity and mortality were significantly higher for multiple organ systems, acute- and lethal CAs, with adjusted OR-M not explained by BW and SNAP.Table

Table 1
Full size table

Conclusions: 1) Simple CA classification can identify excess risk for mortality not explained by BW and SNAP. 2) Multiple organ system anomalies increase mortality and illness severity. 3) Scoring CAs is necessary to permit fair inter-NICU comparisons.