Approximately 3% of pregnancies are complicated by preterm premature rupture of membranes (PROM). Administration of antenatal corticosteroids has been somewhat controversial in patients with preterm PROM after 32 weeks of gestation. Reasons cited include increased risk of neonatal sepsis, and increased fetal stress after preterm PROM causing endogenous steroid production and negating the possible benefit.1 Recommendations from the 1994 National Institutes of Health (NIH) Consensus Statement on antenatal corticosteroids echoed these concerns and limited antenatal corticosteroid administration to less than 30 to 32 weeks of gestation in the setting of preterm PROM.1 Administration of multiple courses of steroids was commonplace in the 1990s, both because there was thought to be no harm, and because of a potential further decrease in rates of neonatal respiratory distress syndrome (RDS).2,3 However, after literature suggestive of adverse effects on fetal growth emerged,4,5 the Eunice Kennedy Shriver National Institute of Child Health and Development updated their recommendations in 2000 to specify that the practice of administering more than one antenatal corticosteroid course should continue only in a research protocol.6 Therefore, a single course was recommended for pregnant women at high risk of delivery between 24 and 34 weeks of gestation.6
Recently, a randomized trial by Garite et al7 has shown the benefit of a rescue course of steroids over a single course in decreasing neonatal respiratory morbidity. The rescue course, administered at least 2 weeks after the initial course if the patient remained pregnant at less than 32 weeks of gestation and the threat for preterm delivery recurred, was found to further decrease neonatal morbidity. The authors excluded women with preterm PROM, yet administration of rescue steroids for these patients has become commonplace. Although it is believed that multiple courses of steroids increase neonatal sepsis in the setting of preterm PROM,8 the question of whether a second course of steroids increases both maternal and neonatal infectious morbidity in the setting of preterm PROM has yet to be answered. Therefore, our objective was to evaluate the association between two courses of antenatal corticosteroids and neonatal sepsis.
MATERIALS AND METHODS
This is a secondary analysis of the Beneficial Effects of Antenatal Magnesium randomized, controlled trial conducted through the Eunice Kennedy Shriver National Institute of Child Health and Development’s Maternal–Fetal Medicine Units Network, which enrolled women at 20 centers across the country from December 1997 to May 2004. The purpose of the parent study was to assess whether antenatal magnesium decreased the rate of cerebral palsy or death in children born preterm. Complete details of the study design and methods have been previously reported.9 Data were entered by trained study staff who were specifically credentialed to record the outcomes. This secondary analysis was approved by the Institutional Review Board at Columbia University Medical Center, New York.
In the parent study a total of 2,241 women at risk for preterm delivery were randomized to magnesium sulfate or placebo.9 Women were considered eligible for the parent study if they were carrying singleton or twin gestations at 24–31 weeks of gestation with a high likelihood of preterm delivery, for example, women who had experienced preterm PROM at or beyond 22 weeks of gestation. In fact, more than 80% of women enrolled had preterm PROM. Gestational age was established according to a standardized algorithm,9 which considered last menstrual period and the earliest ultrasonogram before randomization.
This secondary analysis is an observational cohort study with data collected from the Beneficial Effects of Antenatal Magnesium study. The analytic cohort was defined by women with preterm PROM, and preterm PROM was diagnosed if two of the following three criteria were met: 1) pooling of amniotic fluid in the vaginal vault, 2) positive Nitrazine test, or 3) ferning of vaginal fluid. Additionally, preterm PROM could also be diagnosed if one of the following were present: 1) indigo carmine pooling in the vagina after amnioinfusion, or 2) visible leakage of amniotic fluid from the cervix. For this analysis, only women with singleton gestations who presented with preterm PROM were included. Neonates with major congenital anomalies and stillbirths were not included. Women who received more than two courses of antenatal corticosteroids or who had incomplete data related to neonatal sepsis or antenatal corticosteroid exposure also were excluded.
Our primary exposure was antenatal corticosteroids. One course of betamethasone comprises two doses of 12 mg given 24 hours apart. One course of dexamethasone is 6 mg every 12 hours for four doses. The drugs are similar and differ only by a methyl group.10 The primary independent exposure, antenatal corticosteroids, was made a binary variable, defined as either one or two courses. Although data on pregnancies receiving more than two courses were available, these data were not included in the analysis as multiple courses of corticosteroids are no longer clinically relevant.
Our primary outcome was neonatal sepsis, defined in the parent trial as proven sepsis where blood, cerebrospinal fluid, or urine (either catheterized or suprapubic) cultures are positive and infection is suspected on physical examination; or, in the absence of positive cultures, there is evidence of cardiovascular collapse and X-ray findings confirm infection in the setting of a neonate with clinical sepsis. Neonates with only suspected sepsis, who did not meet the criteria listed above for confirmed sepsis, were not considered to have sepsis for the purpose of this analysis. Secondary outcomes included RDS and birth weight.
Patient characteristics and other demographic variables were analyzed by the χ2 test for categorical variables and Student t test for continuous variables. Nonparametric data were analyzed using the Wilcoxon rank-sum test. Predictors of neonatal sepsis were analyzed in a multivariable logistic regression model. Variables included in the regression model included those that were significantly different between groups (sepsis or no sepsis), the study group assignment (because these data are from a randomized controlled trial), and the primary exposure of interest (one compared with two courses of steroids). Different models were assessed by goodness-of-fit testing, including variables that potentially were correlated until the best model to fit the data was established. In cases where the relationship between neonatal sepsis and the predictor was not known to be linear, such as maternal age and hours from membrane rupture to delivery, the variables were divided into clinically relevant groups, and added to the regression model. Significance was set at a P value of less than .05. All analyses were generated using SAS 9.4.
Participants in the parent study fixed our sample size for this observational study. Thus, we estimated a detectable effect size of an odds ratio (OR) of 1.6 based on our sample size to achieve a power of 80% holding our Type I error rate at 5%. We would not consider an effect size lower than this to be clinically relevant.
Of 2,241 women in the parent study, 2,088 had preterm PROM; of those, 326 pregnancies were excluded for being twin gestations, leaving 1,762 eligible for inclusion. Of those, 38 did not receive steroids, 62 had more than two courses, one had missing information regarding steroid exposure, and 20 had missing information about neonatal sepsis. This left 1,641 patients eligible for analysis.
Women in both groups were similar with regard to age, race, gestational age, and mode of delivery (Table 1). However, women who received two courses of steroids had a longer interval from membrane rupture to delivery than those who received one course. The rate of neonatal sepsis was similar whether the child’s mother was exposed to one compared with two courses of steroids (16.2% compared with 17.2%, P=.756, respectively). Although there appears to be no difference in either birth weight or the rate of RDS whether the mother was exposed to one or two courses of steroids, these are unadjusted results and should be taken with caution (Table 1).
In our unadjusted analysis, neonates of women who were older, had an earlier gestational age at delivery, had a shorter interval from rupture to delivery, were not exposed to antibiotics, or had RDS were more likely to have neonatal sepsis (Table 2). Exposure to one compared with two courses of antenatal corticosteroids did not influence the rate of neonatal sepsis. Maternal chorioamnionitis was significantly associated with neonatal sepsis. Those with neonatal sepsis also were noted to have a lower birth weight and were delivered vaginally. There was no difference between groups in rates of fetal or neonatal death. After adjusting for factors thought to influence the rate of neonatal sepsis listed in Table 3, we found that the primary factors associated with neonatal sepsis were a longer length of time from membrane rupture to delivery (OR 1.52, 95% confidence interval [CI] 1.02–2.26], RDS (OR 2.11, CI 1.45–3.07), gestational age at delivery (OR 0.98, CI 0.95–0.99), and birth weight (OR 0.85, CI 0.77–0.94) (Table 3). Maternal chorioamnionitis and vaginal delivery were no longer associated with neonatal sepsis.
We found that neonatal sepsis was not different in neonates of women with preterm PROM exposed to one compared with two courses of steroids, suggesting that two courses can be used in this population without an increase in neonatal infectious morbidity. This study was designed to assess the potential harm of exposure to two courses by assessing neonatal sepsis. However, the potential benefits of two courses in this group need further study.
There are several important limitations to this study that warrant discussion. First, we cannot comment on the exact timing of antenatal corticosteroid administration, nor do we know which steroid was administered. We only know the proportion of patients exposed to steroids and the number of courses given. However, we do know the time from membrane rupture to delivery, and we know that women who had a longer time from membrane rupture to delivery received two courses of steroids. This supports our theory that women who received two courses of steroids simply were pregnant longer, rather than supporting a differential selection of one compared with two courses. Management of preterm PROM during the study period involved administration of steroids and antibiotics shortly after presentation to the hospital. In this scenario, if steroid exposure increased infectious morbidity, women with more time after membrane rupture and exposure to two courses of steroids would certainly demonstrate more neonatal sepsis and bias our estimates away from the null, but this was not the case. Whether betamethasone or dexamethasone was administered should not influence the outcome of neonatal sepsis as the drugs are quite similar and neither is associated with an increased risk of sepsis.
Although we are able to comment on one course compared with two courses of corticosteroids, we cannot comment on “rescue” steroids, where “rescue” is defined as administration of a second course of steroids after the first has been administered when delivery again becomes imminent at some time in the future. This may limit the generalizability of our findings. In one of the only studies to show a benefit to rescue steroids, Garite and colleagues7 gave the rescue course at least 2 weeks after exposure to the first course. In our study, as is discernible from the time of rupture to delivery, most patients received the second course 1 week after the first, a practice that has biological plausibility for increasing the likelihood of neonatal sepsis owing to a higher, more concentrated level of exposure. However, we did not observe such an effect.
There are two studies evaluating multiple courses of steroids in this population. Compared with pregnancies with preterm PROM exposed to a single course of steroids, Yang et al11 found an increased risk of chorioamnionitis, and Vermillion and colleagues8 found an increased risk of neonatal sepsis when repeated courses were administered. Yet, owing to concerns about growth restriction and a possible association with cerebral palsy, administration of multiple courses has fallen out of favor and is no longer considered clinically relevant.12–14
In conclusion we found that ruptured membranes for more than 14 days, a lower gestational age at delivery, lower birth weight, and RDS were associated with neonatal sepsis. These associations are well known,15–18 and before and after adjusting for these in our regression model, two courses of antenatal corticosteroids were not associated with neonatal sepsis. However, neither in this study, nor in a Cochrane review was there benefit to repeat antenatal corticosteroid administration, which argues against repeating a course of antenatal corticosteroids in the setting of preterm PROM.19
1. Gilstrap LC, Christensen R, Clewell WH, D’Alton ME, Davidson EC, Escobedo MB, et al.. Effect of corticosteroids for fetal maturation on perinatal outcomes: NIH Consensus Development Panel on the effect of corticosteroids for fetal maturation on perinatal outcomes. JAMA 1995;273:413–8.
2. Empana JP, Anceschi MM, Szabo I, Cosmi EV, Breart G, Truffert P, et al.. Antenatal corticosteroids policies in 14 European countries: factors associated with multiple courses. The EURAIL survey. Acta Paediatr 2004;93:1318–22.
3. Elimian A, Verma U, Visintainer P, Tejani N. Effectiveness of multidose antenatal steroids. Obstet Gynecol 2000;95:34–6.
4. Newnham JP, Moss TJ. Antenatal glucocorticoids and growth: single versus multiple doses in animal and human studies. Semin Neonatol 2001;6:285–92.
5. Ikegami M, Jobe AH, Newnham J, Polk DH, Willet KE, Sly P. Repetitive prenatal glucocorticoids improve lung function and decrease growth in preterm lambs. Am J Respir Crit Care Med 1997;156:178–84.
6. National Institutes of Health Consensus Development Panel. Antenatal corticosteroids revisited: repeat courses—National Institutes of Health Consensus Development Conference Statement, August 17–18, 2000. Obstet Gynecol 2001;98:144–50.
7. Garite TJ, Kurtzman J, Maurel K, Clark R. Impact of a “rescue course” of antenatal corticosteroids: a multicenter randomized placebo-controlled trial. Am J Obstet Gynecol 2009;200:248.e1–9.
8. Vermillion ST, Soper DE, Chasedunn-Roark J. Neonatal sepsis after betamethasone administration to patients with preterm premature rupture of membranes. Am J Obstet Gynecol 1999;181:320–7.
9. Rouse DJ, Hirtz DG, Thom E, Varner MW, Spong CY, Mercer BM, et al.. A randomized, controlled trial of magnesium sulfate for the prevention of cerebral palsy. N Engl J Med 2008;359:895–905.
10. Jobe AH, Soll RF. Choice and dose of corticosteroid for antenatal treatments. Am J Obstet Gynecol 2004;190:878–81.
11. Yang SH, Choi SJ, Roh CR, Kim JH. Multiple courses of antenatal corticosteroid therapy in patients with preterm premature rupture of membranes. J Perinat Med 2004;32:42–8.
12. Wapner RJ, Sorokin Y, Thom EA, Johnson F, Dudley DJ, Spong CY, et al.. Single versus weekly courses of antenatal corticosteroids: evaluation of safety and efficacy. Am J Obstet Gynecol 2006;195:633–42.
13. Norberg H, Stålnacke J, Diaz Heijtz R, Smedler AC, Nyman M, Forssberg H, et al.. Antenatal corticosteroids for preterm birth: dose-dependent reduction in birthweight, length and head circumference. Acta Paediatr 2011;100:364–9.
14. Walfisch A, Hallak M, Mazor M. Multiple courses of antenatal steroids: risks and benefits. Obstet Gynecol 2001;98:491–7.
15. Vergnano S, Menson E, Kennea N, Embleton N, Russell AB, Watts T, et al.. Neonatal infections in England: the NeonIN surveillance network. Arch Dis Child Fetal Neonatal Ed 2011;96:F9–F14.
16. Stoll BJ, Hansen NI, Sánchez PJ, Faix RG, Poindexter BB, Van Meurs KP, et al.. Early onset neonatal sepsis: the burden of group B Streptococcal and E. coli disease continues. Pediatrics 2011;127:817–26.
17. Herbst A, Källén K. Time between membrane rupture and delivery and septicemia in term neonates. Obstet Gynecol 2007;110:612–8.
18. Bekhof J, Reitsma JB, Kok JH, Van Straaten IH. Clinical signs to identify late-onset sepsis in preterm infants. Eur J Pediatr 2013;172:501–8.
19. Crowther CA, McKinlay CJ, Middleton P, Harding JE. Repeat doses of prenatal corticosteroids for women at risk of preterm birth for improving neonatal health outcomes. The Cochrane Database of Systematic Reviews 2011, Issue 6. Art. No.: CD003935. DOI: 10.1002/14651858.CD003935.