Hayes, Edward J. MD, MSCP1; Paul, David A. MD2; Stahl, Gary E. MD3; Seibel-Seamon, Jolene MD1; Dysart, Kevin MD4; Leiby, Benjamin E. PhD5; Mackley, Amy B. RNC2; Berghella, Vincenzo MD1
From the 1Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, Thomas Jefferson University, Philadelphia, Pennsylvania; 2Division of Neonatology, Department of Pediatrics, Christiana Care Health Services, Newark, Delaware; 3Division of Neonatology, Department of Pediatrics, Cooper University Hospital, Camden, New Jersey; 4Division of Neonatology, Department of Pediatrics, Thomas Jefferson University, Nemours Children’s Clinics, Philadelphia, Pennsylvania; and 5Division of Biostatistics, Department of Pharmacology and Experimental Therapeutics, Thomas Jefferson University, Philadelphia, Pennsylvania.
Presented as a poster at the Society for Maternal–Fetal Medicine 28th Annual Meeting, Dallas, Texas, January 28 to February 2, 2008.
Corresponding author: Edward J. Hayes, MD, MSCP, Ben Franklin House, 834 Chestnut Street, Suite 400, Philadelphia, PA 19107-5127; e-mail: email@example.com.
Financial Disclosure The authors have no potential conflicts of interest to disclose.
OBJECTIVE: To estimate if exposure to antenatal corticosteroids was associated with decreased rate of death in neonates born at 23 weeks of gestation.
METHODS: This is a retrospective cohort study performed at three tertiary centers of neonates born at 23 weeks of gestation between 1998 and 2007. Stillbirths, voluntary terminations, or parental elected nonresuscitations were excluded. Clinical and demographic variables were examined to determine possible confounding variables. A multivariable logistic regression model was used to assess the effect of steroids on the odds of death after adjustment for these confounders.
RESULTS: The sample included 181 neonates. Of the multiple variables examined (institution, race, diagnosis, illicit drug use, antibiotics, assisted reproduction, birth weight, gender, and route of delivery), only multiple gestations were significantly associated (P≤.15) with steroid use and increased odds of death (odds ratio [OR] 3.66, 95% confidence interval [CI] 1.05–12.73) and controlled for in the final model. The multivariable model revealed those exposed to antenatal corticosteroids had decreased odds of death (OR 0.32, 95% CI 0.12–0.84), with no significant differences in the occurrence of necrotizing enterocolitis among survivors (15.4% compared with 28.6%, P=.59) or severe intraventricular hemorrhage (23.1% compared with 57.1%, P=.17). In analyzing the effect of steroid dose, only a complete course of corticosteroids was associated with a decreased odds of death (OR 0.18, 95% CI 0.06–0.54).
CONCLUSION: Neonates at 23 weeks of gestation whose mothers completed a course of antenatal corticosteroids had an associated 82% reduction in odds of death.
LEVEL OF EVIDENCE: II
In 1972, Liggins and Howie1 reported that antenatal corticosteroids significantly reduced the incidence of respiratory distress syndrome and improved neonatal survival in preterm infants. Multiple studies have confirmed that the use of antenatal corticosteroids are associated with a decrease in neonatal mortality, respiratory distress syndrome, necrotizing enterocolitis, and intraventricular hemorrhage in preterm infants.2 These recurrent findings of survival benefit in premature neonates resulted in the 1994 National Institute of Child Health and Human Development (NICHD) and the Office of Medical Applications of Research of the National Institutes of Health consensus panel statement recommendation that all fetuses at risk for delivery between 24 and 34 weeks should be considered candidates for antenatal corticosteroids.3
The consensus panel did not endorse the use of antenatal corticosteroids before 24 weeks of gestation for two reasons. First, at that time neonates born before 24 weeks of gestation were not being routinely resuscitated. Second, the prevailing thinking had been that, before the 24th week of gestation, during the canalicular stage of lung development, the developmental anatomy of the fetal lung lacked the structural units capable of gas exchange, type II pneumocytes, the target cells for the antenatal corticosteroid treatment.4 Since that time, the resuscitation of neonates at 23 weeks of gestation has become more common, with Janvier and Barrington5 in a retrospective review reporting 100% of neonates born between 230/7 and 236/7 weeks at a single tertiary center underwent resuscitation. The practice of resuscitating infants before 24 weeks of gestation has led some to recommend the routine use of antenatal corticosteroids at 23 weeks of gestation,6 and more than 50% of perinatologists surveyed would recommend antenatal corticosteroid administration for fetuses threatening to deliver before 24 weeks of gestation.7 However, the practice of offering antenatal steroids to women threatening to deliver at 23 weeks of gestation has not been universally accepted (Spellacy WN. Letter to the editor. Am J Obstet Gynecol 2005;192:2090) nor proven beneficial.
Because some infants who are delivered between 230/7 and 236/7 weeks are resuscitated, it is important to determine if antenatal corticosteroids provide a measurable survival benefit. Therefore, our objective was to estimate if exposure to antenatal corticosteroids was associated with decreased odds of death in neonates born at 23 weeks of gestation.
MATERIALS AND METHODS
We conducted a retrospective cohort study of neonates born at 23 weeks of gestation at three tertiary care centers: Christiana Care Health Services, Newark, Delaware; Cooper University Hospital, Camden, New Jersey; and Thomas Jefferson University Hospital, Philadelphia, Pennsylvania. These tertiary medical centers allowed for a diverse population to be examined. Cooper University Hospital and Thomas Jefferson University Hospital are inner city tertiary care centers, while Christiana Care Health Services is the sole tertiary care center for the state of Delaware, covering a number of city and rural communities. After obtaining institutional review board approval at all three institutions, mothers who had delivered neonates between 230/7 and 236/7 weeks during the period 1998–2007 were identified. Maternal charts were reviewed, and those patients who delivered a fetus with a major anomaly, had undergone an elective pregnancy termination, or had declined resuscitation of the neonate were excluded. The gestational age at the time of delivery was confirmed from the maternal chart. The gestational age was based on last menstrual period (LMP) confirmed by ultrasonography before delivery. If the patient had an unsure LMP or if there was a significant discrepancy between ultrasound gestational age and LMP-based dating, the ultrasound dating was used for gestational age assignment in concordance within accepted practice guidelines.8
Those maternal demographics that were theorized as biologically plausible to affect neonatal survival were recorded. These include maternal age, primary diagnosis at the time of delivery, race, concomitant illicit drug use, multiple gestations, antibiotics at the time of delivery, assisted reproduction, and route of delivery. Infant death in the delivery room and the neonatal demographics of gender and birth weight were recorded. The type and number of antenatal corticosteroids were documented, and completeness of course determined based on two doses of intramuscular betamethasone 24 hours apart or four doses of intramuscular dexamethasone 12 hours apart.
Those neonates who survived to admission to the neonatal intensive care unit were identified, and outcomes were documented from the neonatal database at each of the individual institutions, with subsequent confirmation from the neonatal medical record. If a discrepancy was determined between the maternal or infant chart and the database, the patient’s chart was used as the source of information for the analysis.
All neonates delivering at 23 weeks of gestation who met the above inclusion criteria were analyzed. The primary outcome of this study was infant death. Death in this population was defined as death before discharge from the hospital. Secondary outcomes were severe intraventricular hemorrhage and necrotizing enterocolitis. Intraventricular hemorrhage was determined via cranial ultrasound examinations routinely obtained on the fourth day of life and then monthly until discharge. The scans were done using a 7.5-MHz transducer and studies were interpreted by a pediatric radiologist. Intraventricular hemorrhage was graded using the classification system of Papile et al.9 Severe intraventricular hemorrhage was considered grade III–IV. Necrotizing enterocolitis was defined as a minimum of stage 2 according to the method of Bell et al.10
Mothers receiving antenatal steroids were compared with those who did not receive antenatal steroids with respect to the distribution of clinical and demographic variables. Pearson χ2 test was used to test for differences in categorical variables, and the Student two-sample t test was used to test for differences in maternal age. Exact χ2 test P values were calculated for low-prevalence variables. Neonates exposed to corticosteroids in utero were compared with those who were not via logistic regression analysis testing for differences in sex and route of delivery. Mixed-effects linear regression was used to compare neonatal birth weights. A multivariable logistic regression model was used to assess the effect of steroids on the odds of death. Variables associated with use of antenatal steroids (P≤.15) were considered for inclusion in the model. A variable was retained in the multivariable model if it changed the log odds ratio for the association of death and steroid use by ±15%. Generalized estimating equation methods were used to adjust standard errors from all logistic regression analyses for correlation of multiple births from the same mother. All statistical calculations were performed with SAS 9.1.3 (SAS Institute Inc., Cary, NC).
Of the 104,614 live births at the three institutions between 1998 and 2007, 181 neonates born between 230/7 and 236/7 weeks of gestation to 149 mothers met inclusion criteria. Sixty-three mothers received antenatal corticosteroids, with 32 having a complete course and 31 having a partial course. The maternal distribution of demographic and clinical factors with the likelihood of steroid use is reported in Table 1, and those neonatal variables associated with steroid use are reported in Table 2. Of the 181 neonates, 66 died in the delivery room, and of the 115 who survived to admission to the neonatal intensive care unit, 20 survived to discharge.
In our study sample, three factors were determined to be associated with completeness of antenatal corticosteroid use on univariable analysis: the primary diagnosis of preterm labor or clinical chorioamnionitis and multiple gestation. Because there was not a significant increase in odds of death with the primary diagnosis of chorioamnionitis (P=.43) or preterm labor (P=.88), these variables were not retained for the multivariable analysis. In contrast, multiple gestation was associated with receiving steroids and increased odds of death (odds ratio [OR] 3.66, 95% confidence interval [CI] 1.05–12.73). The subsequent multivariable model, therefore, adjusted for multiple gestation, revealed that infants exposed to antenatal corticosteroids had decreased odds of death (OR 0.32, 95% CI 0.12–0.84). The effect of completeness of course on survival to discharge is reported in Table 3. Between those infants exposed to antenatal corticosteroids and those who were not, there were no significant differences in the occurrence of necrotizing enterocolitis (15.4% compared with 28.6%, P=.59) and severe intraventricular hemorrhage (23.1% compared with 57.1%, P=.17), respectively.
Although the use of antenatal corticosteroids has proven beneficial between 24 and 34 weeks of gestation,2 the use of corticosteroids at 23 weeks of gestation remains controversial (Spellacy letter, 2005).6,7 In our study population, infants born between 230/7 and 236/7 weeks of gestation whose mothers received a complete course of antenatal corticosteroids had a reduction in odds of death compared with those infants whose mothers received no corticosteroids after controlling for multiple gestations. The association of multiple gestation and increased odds of death at weights less than 750 g had been previously reported.11 Although studies have shown improved outcomes in very low birth weight infants exposed to antenatal corticosteroids,12 to our knowledge, none had specifically addressed use before 24 weeks of gestation and examined its effect on neonatal outcomes.
This study population of 181 neonates born at 23 weeks of gestation serves as a significant resource for practitioners when counseling patients on neonatal outcomes for infants born at this gestational age. Clinicians and patients must consider both mortality and morbidity when dealing with neonates born at this margin of viability. For even with corticosteroids in our study population, a neonate born in the 23rd week has only a 20% chance of survival to discharge. In addition, of those infants in our sample who survived to discharge, 35% had a severe intraventricular hemorrhage and 20% had necrotizing enterocolitis.
The main finding of our study, that a complete course of antenatal corticosteroids is associated with improved survival at 23 weeks of gestation, not only has direct clinical applications, but further, it affects our understanding of the benefits of antenatal corticosteroids in humans. The NICHD concluded that antenatal corticosteroids accelerate fetal lung maturation and improve neonatal outcome.3 Since that time, we have come to appreciate the multisystem effects of corticosteroids in the developing fetus. Corticosteroids have been shown to accelerate the rate of tissue differentiation without altering the sequence of developmental events in 14 other differentiating tissues besides lung.13 Because respiratory distress syndrome was uniform throughout our study population, the survival benefit at this early gestation may primarily be related to the beneficial effects of antenatal corticosteroids in tissues outside of the lung due to the presumed inability of type II alveolar cells to respond at this gestational age.4
Despite finding an association between maternal administration of a complete course of antenatal corticosteroids and infant survival in our study population, the retrospective cohort study design precludes a definitive determination of whether this is a causative relationship. Despite attempting to identify and control for biologically plausible confounding variables, the observed survival advantage may be related to antecedent events or other confounding variables that we were unable to measure and, therefore, control for in the model. Also, although the retrospective design did allow for the primary outcome of death to be determined, it was limited in its ability to determine long-term outcomes of infants who survived.
We cannot rule out the possibility that the gestational age of some infants in our study was misclassified. To minimize the error in gestational age classification, a hierarchy for gestational age determination was used at each study center using published guidelines.8 Dating based on these clinical criteria, in theory, may have allowed for a growth-restricted fetus older than 236/7 weeks to have been inadvertently included in our study population. Despite this limitation, the use of standard practice guidelines strengthens the “real world” findings of our study and may help assist practitioners when faced with a similar clinical scenario. Based on an assumed death rate of 95% with 15% reduction in the intervention group, using an alpha of 0.05, and β of 0.80 to obtain our goal of 82 subjects per group, we analyzed data at each center between 1998 and 2007. During the study time frame, we cannot rule out the possibility that some changes in neonatal and perinatal care occurred that we were unable to control for, such as the timing of surfactant dosing in the resuscitative process, of which 115 of the neonates in this cohort survived to receive. However, the NICHD Neonatal Research Network findings that show no significant improvement in neonatal mortality for infants with birth weights of 501–750 g, when comparing the cohorts from 1995–1996 with those from 1997–2002 (46% compared with 45%, P=.659),14 suggest that recent changes in neonatal care have not led to major improvements in outcomes over this time frame.
In summary, this study examines the short-term outcomes of neonates exposed in utero to antenatal corticosteroids and born between 230/7 and 236/7 weeks of gestation. Neonates at 23 weeks of gestation whose mothers completed a course of antenatal corticosteroids had an associated 82% reduction in odds of death. In light of our findings that a complete course of antenatal steroids is associated with improved survival, it may be reasonable to offer antenatal corticosteroids to a patient who wants “everything done” when at risk to deliver between 230/7 and 236/7 weeks. However, although our data show an associated survival benefit, 80% of those infants exposed to steroids did not survive to discharge, and approximately 50% of those neonates who survived experienced necrotizing enterocolitis or intraventricular hemorrhage or both. Also the question of how to manage patients who are given antenatal corticosteroids at 23 weeks, who do not deliver within 14 days but subsequently are at risk for delivery before 34 weeks, has yet to be determined. It is hoped that our study will lead to subsequent randomized control trials that will not only go on to support our conclusions but will include long-term follow-up and evaluation of the cost-effectiveness of this intervention.
1. Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics 1972;50:515–25.
2. Roberts D, Dalziel S. Antenatal corticosteroids for accelerating fetal lung maturation for women at risk for preterm birth. The Cochrane Database of Systematic Reviews 2006, Issue 3, Art. No.: CD004454.
3. National Institutes of Health. Report of the Consensus Development Conference on the effect of corticosteroids for fetal maturation on perinatal outcomes. Pub No. 95-3784. Bethesda (MD): U.S. Department of Health and Human Services, Public Health Service; 1994.
4. Jobe A. The respiratory system: Part 1. Lung development. In: Fanaroff A, Martin R, editors. Neonatal-perinatal medicine: diseases of the fetus and infant. 7th ed. St. Louis (MO): Mosby; 2002. p. 973–91.
5. Janvier A, Barrington KJ. The ethics of neonatal resuscitation at the margins of viability: informed consent and outcomes. J Pediatr 2005;147:579–85.
6. Wapner R. Antenatal corticosteroids: we continue to learn. Am J Obstet Gynecol 2004;190:875–6.
7. McElrath TF, Norwitz ER, Nour N, Robinson JN. Contemporary trends in the management of delivery at 23 weeks’ gestation. Am J Perinatol 2002;19:9–15.
8. Perinatal care at the threshold of viability. ACOG Practice Bulletin No. 38. American College of Obstetricians and Gynecologists. Obstet Gynecol 2002;100:617–24.
9. Papile LA, Munsick-Bruno G, Schaefer A. Relationship of cerebral intraventricular hemorrhage and early childhood handicaps. J Pediatr 1983;103:273–7.
10. Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, et al. Neonatal necrotizing enterocolitis: therapeutic decisions based upon clinical staging. Ann Surg 1978;187:1–7.
11. Hayes EJ, Paul D, Ness A, Mackley A, Berghella V. Very-low-birthweight neonates: do outcomes differ in multiple compared with singleton gestations? Am J Perinatol 2007;24:373–6.
12. Doyle LW, Kitchen WH, Ford GW, Rickards AL, Lissenden JV, Ryan MM. Effects of antenatal steroid therapy on mortality and morbidity in very low birth weight infants. J Pediatr 1986;108:287–92.
13. Ballard PL, Ballard RA. Scientific basis and therapeutic regimens for use of antenatal glucocorticoids. Am J Obstet Gynecol 1995;173:254–62.
14. Fanaroff AA, Stoll BJ, Wright LL, Carlo WA, Ehrenkranz RA, Stark AR, et al. NICHD Neonatal Research Network. Trends in neonatal morbidity and mortality for very low birthweight infants. Am J Obstet Gynecol 2007;196:147.e1–8.
Figure. No caption available.
© 2008 by The American College of Obstetricians and Gynecologists.