In the 01 and 02 years of this grant (March 1989-February 1991), we have been developing a physiology-based severity of illness score for newborn intensive care, the Score for Neonatal Acute Physiology (SNAP). This proposal describes an extension of that work already in progress. During the first 1-1/2 years of this 2-year grant, we have completed 60% of data collection and performed preliminary analyses. These analyses provide evidence that such a score is practical and that it appears to correlate well with mortality, and other markers of severity including therapeutic intensity, nursing acuity and physician and nurse judgment scales. Most importantly, measured severity is virtually independent of birthweight and therefore represents a new dimension for investigating outcomes. Extensive refinement and rigorous evaluation of validity and scaling characteristics will be completed by March 1991. As part of the review process, score development was considered a higher priority than correlation with costs. Moreover, further developmental work was needed to obtain cost data. With the preliminary analyses confirming the potential utility of the score, the extensive clinical database of neonatal ICU admissions and the availability of further information on costs, we are now requesting support to analyze a number of important issues relating to severity of illness and NICU resource consumption. The importance of this request lies in the fact that, while there is extensive literature on adult and pediatric ICU utilization and costs, very little is available on the unique characteristics of costs, utilization, and severity of illness in NICU care. We propose to apply a number of the analytic approaches from the adult and pediatric ICU studies to NICUs, specifically: (1) the determinants of NICU admission for other than extreme prematurity, including the whole process of NICU triage which is heretofore undescribed; (2) the determinants of NICU discharge, including the role of severity and availability of step-down facilities; and (3) the determinants of ancillary use according to severity, birthweight, and diagnosis. These analyses are feasible without further primary data collection, and require only the acquisition of hospital financial tapes and one tape from a state agency. The ready availability of the data, high costs of NICU care and paucity of existing literature all suggest the value of these studies.
