Materials and Methods
Data from all neonates admitted to the newborn intensive care unit (NICU) at Miller Children's Hospital in association with Long Beach Memorial Women's Hospital were collected prospectively. The study period was June 1, 1992, through December 31, 1995, and the project was approved by the institutional review board.
Neonatal data collection included gestational age at delivery, birth weight, Apgar scores, mode of delivery, gender, use of rescue surfactant or indomethacin, and overall neonatal outcome. Information gathered from maternal histories included prenatal complications, indications for delivery, use of antenatal corticosteroids, use of tocolytic agents, and type of insurance (private versus Medicaid or no insurance). Any missing maternal information was collected from medical records, and all records were available. Only live newborns from singleton pregnancies with birth weights of 500–1200 g and at least 24 weeks' gestation were included in analysis. Gestational age was based on a documented menstrual history if confirmed by ultrasound or Dubowitz evaluation. When menstrual history was uncertain, gestational age was based on ultrasound confirmed by Dubowitz evaluation. Neonates with fatal anomalies were excluded.
Among eligible neonates, those born at the level III center (inborn) were compared with those transferred to the level III nursery after being born at a level I institution (outborn) in relation to incidence of severe intraventricular hemorrhage, grades III and IV. Neonates were transferred in the first few hours of life. Neonatal resuscitation was done for the inborn group by NICU nurses and neonatologists, and for the out-born group by nursery nurses and pediatricians. Intraventricular hemorrhage, grades III and IV, was defined by criteria of Papile et al.18 Cranial ultrasounds were done within 48 hours of birth in all cases and were repeated between 7 and 14 days of life. Respiratory distress syndrome due to hyaline membrane disease (HMD) was defined as need for supplemental oxygen for more than 24 hours with an arterial to alveolar ratio of less than 0.22, and a chest x-ray showing typical ground-glass appearance with air bronchogram formation. Surfactant during the study was rescue treatment based on HMD as defined and was at the discretion of the neonatologist. On completion of the study, all cranial ultrasounds were reread by an ultrasound radiologist and all chest x-ray films were reread by a pediatric radiologist, both of whom were masked to clinical courses of infants.
Normal umbilical cord arterial blood gas pH was defined as 7.20 or greater. Severe acidosis was defined as less than 7.00, moderate acidosis as 7.00–7.10, and mild acidosis as 7.11–7.19. Acidosis was considered metabolic if the base deficit was above 11.19
Levels of hospital care were described previously.1 Level I facilities offer uncomplicated maternal and neonatal care to women with term pregnancies, with emergency management of unexpected complications. Level II hospitals should have 24-hour anesthesia capabilities, with 24-hour clinical laboratories and radiology services, and should be able to manage high-risk pregnancies at 32 weeks' or more gestation. Level III institutions should be able to supply complete maternal and neonatal care to all high-risk pregnancies at all gestational ages.
The 11 level I hospitals involved were part of a perinatal outreach program in which all deliveries were evaluated on a 1- to 3-month basis. Those centers did not keep newborns weighing less than 1200 g beyond a few hours, so all neonatal transfers from the centers were accounted for during the study. Statistical analysis included the Student t test, Mann-Whitney U test, χ2, and Fisher exact test. P <.05 was considered statistically significant.
There were 37 cases (11%) of grade III or IV intraventricular hemorrhage in the 329 newborns who met study criteria: 27 of 285 delivered in the level III center (9%), and ten of 44 outborn neonates (23%), a statistically lower difference (P < .02, 95% confidence interval of 0.15, 0.87). Table 1 compares outborn and inborn groups for the total study population. The only significant difference was incidence of severe intraventricular hemorrhage.
Table 2 compares newborns who developed grade III or IV intraventricular hemorrhages in relation to inborn versus outborn delivery. Gestational age was significantly lower in the inborn group. Incidence of severe intraventricular hemorrhage increases with lighter birth weight and lower gestational age, so those data further emphasize the difference in the rate of hemorrhage between groups. On the basis of 5-minute Apgar scores, there did not appear to be any difference in incidence of depressed newborns between groups. Umbilical cord blood gases were not measured in any of the outborn neonates with grade III or IV intraventricular hemorrhages. Blood gas levels measured in the first hour of life for those ten newborns did not show any evidence of metabolic acidosis. Eighteen of 27 inborn cases had umbilical cord blood gas levels measured, and two of those were consistent with mild acidosis, with pHs of 7.16 and 7.19.
Use of antenatal corticosteroids is shown in Table 3. The protocol at Long Beach Memorial Women's Hospital and the 11 level I facilities was not to use antenatal steroids in women with preterm premature rupture of membranes (PROM). Use of antenatal corticosteroids was significantly higher in the inborn population; however, use of antenatal corticosteroids in cases with grade III or IV intraventricular hemorrhages was not significantly different. Incidence of grade III or IV intraventricular hemorrhage was still significantly elevated in the outborn group when women who were given antenatal corticosteroids were excluded.
Mode of delivery, preterm PROM, and gender of newborn are possible etiologic factors in intraventricular hemorrhage, so those were evaluated and are presented in Table 4. There was significance between the inborn infants and the outborn group in those delivered by cesarean and those with obstetric complications other than preterm PROM. There also was a trend toward a higher rate in outborn female neonates.
The level I hospitals in the investigation were in proximity to Long Beach Memorial Women's Hospital; therefore, maternal transport could be done within 1 hour of initiation. Among the ten outborn cases with grade III or IV intraventricular hemorrhages, there were no maternal or fetal contraindications to transfer. Nine of those women had the time necessary for maternal transport. Seven cases involved preterm labor, and tocolytic therapy was used in two.
Two of ten outborn infants and three of 27 inborn neonates with grade III or IV intraventricular hemorrhages had drug screens positive for cocaine. Those numbers were not significantly different, although not all women in the study were screened. Antenatal indomethacin usage did not occur in the outborn group with grade III or IV intraventricular hemorrhages; it did occur in three of 27 inborn cases and did not reach significance. Use of antenatal magnesium sulfate was higher in the inborn population (38% versus 18%, P < .02), but there was no difference in use of antenatal magnesium sulfate in cases with grade III or IV intraventricular hemorrhages. It was administered to three of 10 in the outborn group compared with 15 of 27 in the inborn population (P = .08).
Thirty-four percent of the 285 inborn neonates had private insurance, no different from the 32% with private insurance in the 44 outborn cases. However, trends over the 4-year period were completely opposite. The percentages of privately insured patients in the inborn group were 52%, 32%, 30%, and 25%, respectively, for each year of the study from 1992–1995. The percentages of privately insured patients in the outborn population were 14%, 35%, 44%, and 50%, respectively, for each year.
This study analyzed the incidence of severe intraventricular hemorrhage in VLBW neonates on the basis of birth location by level of institution. Neonatal care for both groups was in the same unit, so neonatal management was controlled. The main areas of difference were obstetric care (antenatal and intrapartum management), neonatal resuscitation, and effect of newborn transport. Our data found that the rate of grade III or IV intraventricular hemorrhage is lower in complicated pregnancies delivered in a level III facility.
Multiple factors were analyzed for their potential contribution to grade III and IV intraventricular hemorrhage, so the issue of a multivariate analysis was considered with our perinatal statistician. In creating the model, we collected significant and nearly significant univariate comparisons for grade III and IV intraventricular hemorrhage, gestational age at delivery, use of antenatal corticosteroids, incidence in cases without preterm PROM, cesarean delivery, female gender, and birth weight. The risk of significant intraventricular hemorrhage would decrease primarily with greater gestational age, less preterm PROM, more cesareans, female gender, and a heavier birth weight. All of those factors were more favorable in the outborn population; therefore, the data do not lend themselves to multivariate analysis.
Use of antenatal corticosteroids was higher in the inborn population, which probably led to a better outcome. That also is true because of the higher rate of grade III or IV intraventricular hemorrhage in cases with obstetric complications other than preterm PROM in the outborn population because antenatal corticosteroids were not given to women with preterm PROM. Use of antenatal corticosteroids in cases with grade III or IV intraventricular hemorrhage was not different, and the rate was still higher in the outborn group on analysis of populations excluding those given antenatal corticosteroids (Table 3). Therefore, antenatal corticosteroids were not the only factor involved in the lower rate of grade III or IV intraventricular hemorrhage in the inborn neonates.
Tocolytics were used less in the outborn group. One could question tocolytics' overall benefit in women with preterm labor; however, it has been suggested that aggressive management might result in a 48-hour delay that might allow time for administering antenatal corticosteroids.20
The percentage of privately insured women in the outborn group increased from year to year over the study, the exact cause of which is unknown, but might have been influenced by insurance plans that direct where patient care can occur. Another possibility might be hospital or physician desire to keep privately insured patients for better reimbursement. A complete cost analysis was not done because the total cost of care for infants with grade III or IV intraventricular hemorrhage was not obtainable. It is conceivable that one prevented case of significant intraventricular hemorrhage could pay for multiple maternal transports.
An area of potential bias in the results was capturing all of the transferred VLBW neonates. However, the 11 level I facilities and all deliveries and transports at them were analyzed every 1 to 3 months during the study. We were able to verify completely that there were no transfers of any neonates that met study criteria to any institution other than the NICU at Long Beach Memorial Medical Center.
This study only compared incidence of severe intraventricular hemorrhage between neonates delivered at level I hospitals and transported to the tertiary facility with those delivered at the level III center. We cannot comment on the risk of significant intraventricular hemorrhage of neonates delivered at level II centers, as they are not often transported. A study comparing level II centers to level III centers would be difficult because neonatal care and nursing support and obstetric management would vary between populations.
1. Ryan GM. Toward improving the outcome of pregnancy. Recommendations for the regional development of perinatal health services. Obstet Gynecol 1975;46:375–84.
2. McCormick MC, Shapiro S, Starfield BH. The regionalization of perinatal services. Summary of the evaluation of a national demonstration program. JAMA 1985;253:799–804.
3. Bowes WA. A review of perinatal mortality in Colorado, 1971 to 1978, and its relationship to the regionalization of perinatal services. Am J Obstet Gynecol 1981;141:1045–52.
4. Shenai JP, Major CW, Gaylord MS, Blake WW, Simmons A, Oliver S, et al.
A successful decade of regionalized perinatal care in Tennessee: The neonatal experience. J Perinatol 1991;11:137–43.
5. Paneth N, Kiely JL, Wallenstein S, Marcus M, Pakter J, Susser M. Newborn intensive care and neonatal mortality in low-birth-weight infants. N Engl J Med 1982;307:149–55.
6. Cordero L, Backes CR, Zuspan FP. Very low-birth weight infant. Influence of place of birth on survival. Am J Obstet Gynecol 1982;143:533–7.
7. Mayfield JA, Rosenblatt RA, Baldwin LM, Chu J, Logerfo JP. The relation of obstetrical volume and nursery level to perinatal mortality. Am J Public Health 1990;80:819–23.
8. Modanlou HD, Dorchester WL, Thorosian A, Freeman RK. Antenatal versus neonatal transport to a regional perinatal center: A comparison between matched pairs. Obstet Gynecol 1979;53:725–9.
9. Kollee LAA, Verloove-Vanhorick PP, Verwey RA, Brand R, Ruys JH. Maternal and neonatal transport: Results of a national collaborative survey of preterm and very low birth weight infants in the Netherlands. Obstet Gynecol 1988;72:729–32.
10. Tomich PG, Anderson CL. Analysis of a maternal transport service within a perinatal region. Am J Perinatol 1990;7:13–7.
11. Gagnon D, Allison-Cooke S, Schwartz RM. Perinatal care. The threat of deregionalization. Pediatr Ann 1988;17:447–51.
12. Powell SL, Holt VL, Hickok DE, Easterling T, Connell FA. Recent changes in delivery site of low-birth-weight infants in Washington: Impact on birth weight–specific mortality. Am J Obstet Gynecol 1995;173:1585–92.
13. Yeast JD, Poskin M, Stockbauer JW, Shaffer S. Changing patterns in regionalization of perinatal care and the impact on neonatal mortality. Am J Obstet Gynecol 1998;178:131–5.
14. Phibbs CS, Bronstein JM, Buxton E, Phibbs RH. The effects of patient volume and level of care at the hospital of birth on neonatal mortality. JAMA 1996;276:1054–9.
15. Volpe JJ. Intracranial hemorrhage: Germinal matrix–intraventricular hemorrhage of the premature infant. In: Volpe JJ, ed. Neurology of the newborn. 3rd ed. Philadelphia: W.B. Saunders Company, 1995:403–63.
16. Abdel-Rahman AM, Rosenberg AA. Prevention of intraventricular hemorrhage in the premature infant. Fetal Drug Ther 1994;21:505–21.
17. Clark CE, Clyman RI, Roth RS, Sniderman SH, Lane B, Ballard RA. Risk factor analysis of intraventricular hemorrhage in low-birth-weight infants. J Pediatr 1981;99:625–8.
18. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants with birth weights less than 1,500 gm. J Pediatr 1978;92:529–34.
19. Helwig JT, Parer JT, Kilpatrick SJ, Laros RK. Umbilical cord blood acid-base state: What is normal? Am J Obstet Gynecol 1996;174:1807–14.
20. American College of Obstetricians and Gynecologists. Preterm labor. American College of Obstetricians and Gynecologists technical bulletin no. 206. Washington, DC: American College of Obstetricians and Gynecologists, 1995.