Abstract 1398 Neonatal Epidemiology and Follow-up Poster Symposium, Monday, 5/3

Study Objective: The purpose of this study was to examine the hypothesis that in North Carolina during the 1990's, a reversal of regionalization has occurred with regard to perinatal services for mothers delivering extremely low birth weight (ELBW) infants, thus increasing the mortality for this group. Methods: Data were obtained from the North Carolina Center for Health Statistics on 11,983 neonates with birth weight (BW) 500-1000 gms, born to North Carolina residents in the years 1971 to 1994. Hospitals of birth were categorized as level III if they were capable of caring for premature infants of any gestational age or illness severity. The proportion of births occurring in level III hospitals was used as an index of the degree of regionalization of care. Trends in neonatal survival and in the proportion of births that occurred in level III hospitals were analyzed with the Cochran-Armitage test for trend. Odds ratios (OR) and 95% confidence limits (CL), adjusted for potentially confounding variables, were estimated with logistic regression models. Results: Among neonates with BW 500-1000 gms, neonatal survival increased from 13.7% in 1971 to 58.1% in 1994, and the proportion of births occurring in a level III center increased from 36.1% to 79.4%. Trends were found (increases over time) for both survival as well as the likelihood of birth in a level III center, in interval 1976 to 1991 (p < 0.0001), but not in the years prior to 1975 or after 1991. Adjusting for BW, gestational age, gender, race, and years of maternal education, infants with BW 500-750 gms were less likely to die if born in a level III hospital [OR = 0.6; 95% CL = 0.49-0.73; level I as the referent group]; [OR = 0.51; 95% CL = 0.42-0.61; level II as the referent group]. This was also true among infants with BW 751-1000 gms [OR = 0.56; 95% CL=0.47-0.66; level I as the referent group]; [OR = 0.49; 95% CL=0.41-0.58; level II as the referent group]. The association of birth in a level III hospital and decreased neonatal mortality was more pronounced among infants born since 1990, when surfactant therapy became routine for ELBW infants with respiratory distress syndrome. Among infants with BW 500-750 gms, the OR for infants born in a level III hospital (relative to those born in a level I hospital) was 0.91 (0.31-2.7) in the years 1971-1989, and decreased to 0.13 (0.04-0.39) in the years 1990 to 1994. This was also true among infants with BW 751-1000 gms, [OR= 0.34 (0.14-0.80) for 1971-89 vs. 0.10 (0.03-0.35) for 1990-94]. Conclusions: In North Carolina, regionalization of perinatal care for ELBW neonates began in 1976 and was associated with a steady increase in survival. Since 1976, birth in a level III hospital has been associated with a higher likelihood of survival, and this association has been most pronounced in the years since 1990. Contrary to our expectation, no evidence was found of a reversal of regionalization in period 1990 to 1994.