Objective
We sought to determine whether the incidence of neonatal respiratory distress syndrome (RDS) is similar with 12- vs 24-hour dosing interval of betamethasone.
Study Design
This was a prospective, randomized, open, noninferiority trial. Mothers (n = 228) with a singleton or multiple pregnancies (fetuses = 260), between gestational age of 23-34 weeks, at risk for preterm delivery, received standard 2 doses of betamethasone either 12 or 24 hours apart in 2:1 ratio, respectively.
Results
Incidence of RDS was similar in the 2 cohorts (36.5% vs 37.3%; P = not significant). Women unable to receive the complete course of corticosteroids with the 24-hour interval can be reduced by half with the 12-hour interval. However, increased incidence of necrotizing enterocolitis was seen with 12-hour dosing (6.2% vs 0%; P = .03).
Conclusion
The 12-hour dosing interval is equivalent to the 24-hour dosing interval for prevention of RDS in neonates of mothers delivering prematurely. A larger multicenter study is needed to confirm our findings.
The administration of glucocorticoids to accelerate lung development in the fetus in patients at risk for preterm delivery is one of the cornerstones of present-day obstetrics and neonatal care. Pioneering studies by Liggins and Howie and subsequently by many others have repeatedly shown that maternal administration of glucocorticoids has a significant beneficial effect in decreasing the incidence of respiratory distress syndrome (RDS) in infants delivered between 24-34 weeks’ gestation. In addition, these studies and metaanalysis of available data provide strong evidence that glucocorticoid administration also significantly reduces the incidence of intraventricular hemorrhage (IVH) and overall neonatal mortality, improves circulatory stability in the premature infant, and reduces requirements for oxygen and ventilatory support. The regimen endorsed by the National Institutes of Health (NIH) involves maternal administration of 2 doses of 12 mg of betamethasone intramuscularly 24 hours apart. The NIH has acknowledged that this dosage interval was selected arbitrarily and has recognized an urgent need for studies on dosage variation.
For Editors’ Commentary, see Table of Contents
The total dosage of 24 mg was based on animal studies and later shown to achieve fetal concentrations comparable to normal physiologic stress levels of cortisol occurring after birth. Administration of total dosage, rather than the interval, is probably more important to attain maximal neonatal benefit. This is supported by data from the National Institute of Health and Human Development Neonatal Research Network that suggest that ≥2 doses of corticosteroids (complete treatment) was more effective than partial treatment (<2 doses). However, approximately 25-45% of women at risk for preterm delivery deliver within 24 hours, prior to receiving the complete course of steroids. Only ∼5% of these women deliver within 12 hours.
Prospective randomized trials have used varying dosage intervals (8, 12, or 24 hours) to administer the total dose of 24 mg of a long-acting corticosteroid (betamethasone or dexamethasone). In fact, the metaanalysis of these studies emphasizes the total dose rather than the interval and, “A policy of administering corticosteroids (24 mg betamethasone, or 24 mg dexamethasone, or 2 g hydrocortisone) to women who are expected to deliver preterm can be expected to achieve substantial reductions in neonatal morbidity and mortality. No adverse consequences of this policy have been identified.” Higher doses did not result in a dose-related response for RDS prevention. The timing of the dosages that was performed in the initial studies was arbitrarily selected, according to the authors. Since preterm delivery is the major cause of infant mortality and morbidity and a quarter to one third of patients deliver within 24 hours of the start of betamethasone therapy, it would be beneficial to ascertain whether a different regimen of administration of betamethasone (12 vs 24 hours apart) would be equally efficacious in preventing the incidence of RDS in preterm infants.
Materials and Methods
This is a prospective, randomized, semiblinded noninferiority study conducted at Cooper University Hospital, Camden, NJ, from July 2006 through May 2009, after obtaining approval from the institutional review board. The trial was also registered with clinicaltrial.gov NCT00453141 . The study population consists of mothers with a singleton or multiple pregnancies, between the gestational ages of 23-34 weeks, who presented with spontaneous preterm labor, preterm premature rupture of membranes, or medically indicated preterm delivery. The only difference in inclusion criteria between trials included in the Cochrane Review and this trial is that we chose to include women starting from gestational age of 23 weeks instead of 24 weeks, as neonatal advancements in the last 20 years has made this a viable gestational age. Also, we chose to restrict the trial participants to the use of betamethasone alone to avoid variability that maybe associated with use of different corticosteroids. No rescue doses of steroids were used during the study period. Trials included in the metaanalyses and consensus guidelines have used corticosteroids in various doses and intervals and still shown improvements in RDS rates. We hypothesized that 12 mg of intramuscular betamethasone administered at a 12-hour interval is equally efficacious in the reduction of incidence of RDS, when compared to the same dose of betamethasone given 24 hours apart. Other secondary outcomes evaluated included maternal fever and postpartum length of stay, neonatal frequency of IVH, chronic lung disease (CLD), necrotizing enterocolitis (NEC), retinopathy of prematurity (ROP), sepsis, need for antibiotics >5 days, admission to neonatal intensive care unit (NICU), and length of NICU stay.
Once mothers were identified at risk for delivering preterm, the first dose of betamethasone was administered and only then women were offered participation in the study. Women <23 or >34 weeks’ gestation, or elapsed time >12 hours since administration of the first dose of betamethasone, known drug allergy to betamethasone, given steroid other than betamethasone for lung maturation, or any contraindication to steroid therapy were excluded. After informed consent was obtained, patients were allocated into 4 groups depending on gestational age: 23.1-26.0 weeks (A); 26.1-29.0 weeks (B); 29.1-32.0 weeks (C); and 32.1-34.0 weeks (D). The patients in each group were randomly assigned to receive 2 doses of 12 mg of betamethasone intramuscularly either 12 or 24 hours apart. Consenting patients were matched to the appropriate group for gestational age and the next opaque envelope was drawn for that group, which became her assigned study number. The obstetrical care and delivery plan was determined by the treating obstetrician in the usual manner. Every attempt was made to contact mothers delivering elsewhere and obtain hospital records for accurate delivery and neonatal data.
The neonatologists were informed if the patient received steroid therapy, and whether she received 1 or 2 doses, but were not informed of the dosing interval. The 12- vs 24-hour allocation was revealed only after outcome data extraction was completed. The diagnosis of RDS was defined as an infant who has a PaO 2 <50 mm Hg on room air, central cyanosis in room air, or a requirement for supplemental oxygen to maintain PaO 2 >50 mm Hg, in conjunction with a chest radiograph with findings consistent with RDS and/or need for surfactant therapy. CLD was defined as oxygen dependency at ≥28 days of life. Neurosonograms were performed on neonates delivering preterm and were read by an experienced neuroradiologist for evidence of IVH and graded as described by Papile et al. NEC was diagnosed clinically (bilious gastric aspiration or emesis, abdominal distention, or occult/gross blood in stool) and confirmed radiographically (pneumatosis intestinalis, hepatobiliary gas, or pneumoperitoneum). Neonatal sepsis was defined by positive blood cultures; but antibiotic use for >5 days in the neonate was also recorded. ROP was diagnosed by an indirect ophthalmologic examination. The secondary outcomes were not further graded due to small numbers. Fetal and neonatal deaths occurring within 24 hours after birth were excluded when neonatal outcomes and length of NICU stay were analyzed. Also, neonates in whom evaluation for chronic condition (IVH, ROP, or CLD) was not deemed clinically relevant (eg, term birth) were excluded from the analysis of these outcomes.
Randomization
The randomization scheme used was generated in Excel (Microsoft, Redmond, WA) using its “random function” in 4 groups of 50 each. Randomization was performed in 2:1 ratio (for 12:24-hour dosing interval) by K.H. and study assignments placed in sealed opaque envelopes with a study number and group number on each envelope by C.H. Participant enrolment was done by E.C., M.K., and the obstetrical residents. Unequal allocation was used to provide a large sample size for the 12-hour dosing group, for which there is little published information on short- or longer-term neonatal outcomes. After 200 patients were enrolled, further randomization was done in blocks of 30 resulting in unequal distribution in groups C and D, as frequency of disease occurrence was much higher in this gestational age group.
Power analysis
A prior power analysis suggested sample sizes of 140 pregnancies for the experimental group (12-hour dosing interval) and 70 for the standard group (24-hour dosing interval) to achieve 80% power to detect a noninferiority margin of 20% using a 1-sided Fisher exact test of proportion differences between groups assuming a 24-hour non-RDS rate of 65% and at minimum a 45% non-RDS rate among 12-hour patients. The margin was set at the expected lower 95% confidence limit of a 24-hour non-RDS rate based on informal historical observation at sites included in the study. To obtain this power the sampled 12-hour rate is assumed to be ∼67%.
Data analysis
Noninferiority was tested by a 1-tail Fisher exact test and by comparison of the lower bound of the percent difference in non-RDS rates observed for the 12- and the 24-hour maternal treatment groups. Means and SD are reported for continuous short- and longer-term neonatal and maternal parameters; medians and interquartile ranges for categorical measures; and percents for binary measures. Independent sample t tests with homogeneity of variance adjustment where indicated by the Levene test were used to compare parametric continuous variables between the 12- and 24-hour group. The Mann-Whitney U test was used to compare nonparametric continuous variables. Pearson χ 2 and Fisher exact tests were used to compare discreet variables. SPSS 15.1 (IBM, Armonk, NY) and Confidence Interval Analysis (BMJ Books, University of Southampton, London, UK) were used for analysis.
Results
The study population included 228 mothers (260 neonates) who were consented for participation and intent-to-treat analysis was conducted. Incomplete information was available for 4 neonates in the 12-hour group and 3 in the 24-hour group. Seven neonates (1 set of triplets) received the incorrect dosing interval due to either failure of order placement or execution, and in 1 case obstetrician preference due to urgency of delivery ( Figure ). By study design, approximately 70% (161/228 mothers; 180/260 fetuses) were assigned to the 12-hour interval dosing; 30% (67/228 mothers; 80/260 fetuses) were assigned the 24-hour interval dosing. Group A included 47 mothers (56 fetuses); group B included 39 mothers (44 fetuses); group C included 72 mothers (84 fetuses); and group D included 70 mothers (76 fetuses).
Women in the 2 treatment arms were similar with respect to gravidity, parity, race, and body mass index ( Table 1 ). The women in the 24-hour arm were about 2 years older than in the 12-hour arm. In the 24-hour cohort, there were more African American women at risk for delivery <26 weeks and more Caucasians at risk in the 32- to 34-week group (but the racial differences were statistically significant only in group D). The groups were enrolled at similar gestational ages with similar frequency of indications for delivery ( Table 2 ), and similar number of growth-restricted neonates ( Table 3 ). Even though the 2 cohorts, overall, delivered at similar gestational ages and intervals from first dose of betamethasone, group B delivered approximately 3 weeks earlier in the 12-hour arm compared to the 24-hour arm ( Table 2 ). Neonatal mode of delivery, birth weights, and sex were also similar in the groups ( Table 3 ).
| Characteristic | 12 hour | 24 hour | 95% CI | ||||
|---|---|---|---|---|---|---|---|
| n | Mean | SD | n | Mean | SD | ||
| Maternal age, y | |||||||
| Overall | 161 | 27.20 | 6.82 | 67 | 29.67 | 6.83 | −4.42 to −0.51 |
| Group A | 31 | 28.87 | 6.24 | 16 | 31.19 | 5.91 | −6.12 to 1.49 |
| Group B | 27 | 28.78 | 7.79 | 12 | 28.25 | 5.14 | −3.77 to 4.83 |
| Group C | 50 | 26.86 | 6.16 | 22 | 29.82 | 8.99 | −7.26 to 1.34 |
| Group D | 53 | 25.75 | 7.01 | 17 | 29.06 | 5.62 | −7.04 to 4.27 |
| Gravidity | 161 | 67 | |||||
| Overall | 3.11 | 2.16 | 3.55 | 2.34 | −1.08 to 0.19 | ||
| Parity (median) | 1 | 0–2 | 1 | 0–3 | −1 to 0 | ||
| Race, n (%) | |||||||
| Overall | 161 | 67 | |||||
| White | 57 (35.4) | 29 (43.3) | |||||
| Black | 64 (39.8) | 19 (28.4) | |||||
| Hispanic | 35 (21.7) | 16 (23.9) | |||||
| Asian | 2 (1.2) | 1 (1.5) | |||||
| Other | 3 (1.9) | 2 (3.0) | −0.23 to 0.26 | ||||
| Group A, n (%) | 31 | 16 | |||||
| White | 11 (35.5) | 3 (18.8) | |||||
| Black | 10 (32.3) | 11 (68.8) | |||||
| Hispanic | 10 (32.3) | 2 (12.5) | −0.7 to 0.48 | ||||
| Group B, n (%) | 27 | 12 | |||||
| White | 11 (40.7) | 6 (50.0) | |||||
| Black | 10 (37.0) | 2 (16.7) | |||||
| Hispanic | 4 (14.8) | 4 (33.3) | |||||
| Asian | 1 (3.7) | 0 (0.0) | |||||
| Other | 1 (3.7) | 0 (0.0) | −0.54 to 0.52 | ||||
| Group C, n (%) | 50 | 22 | |||||
| White | 20 (40.0) | 10 (45.5) | |||||
| Black | 21 (42.0) | 4 (18.2) | |||||
| Hispanic | 7 (14.0) | 6 (27.3) | |||||
| Asian | 0 (0.0) | 1 (4.5) | |||||
| Other | 2 (4.0) | 1 (4.5) | −0.38 to 0.48 | ||||
| Group D, n (%) | 53 | 17 | |||||
| White | 15 (28.3) | 10 (58.8) | |||||
| Black | 23 (43.4) | 2 (11.8) | |||||
| Hispanic | 14 (26.4) | 4 (23.5) | |||||
| Asian | 1 (1.9) | 0 (0.0) | |||||
| Other | 0 (0.0) | 1 (5.9) | −0.61 to −0.003 | ||||
| Maternal BMI, lbs/ft 2 | n | Mean | SD | n | Mean | SD | 95% CI |
|---|---|---|---|---|---|---|---|
| Overall | 135 | 28.94 | 6.25 | 41 | 29.67 | 7.76 | −3.06 to 1.60 |
| Group A | 24 | 32.88 | 6.42 | 8 | 35.72 | 12.70 | |
| Group B | 26 | 28.95 | 5.24 | 8 | 30.71 | 6.53 | |
| Group C | 37 | 26.92 | 5.37 | 16 | 26.99 | 5.27 | |
| Group D | 48 | 28.53 | 6.57 | 9 | 28.13 | 3.98 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
