A randomized, placebo-controlled trial of betamethasone for the prevention of respiratory distress syndrome at 24 to 28 weeks' gestation ThomasJ. Garite, MD, PamelaJ. Rumney, RNC, Gerald G. Briggs, BPharm, James A. Harding, MD, Michael P. Nageotte, MD, Craig V. Towers, MD, and Roger K. Freeman, MD Long Beach and Orange, California Previous randomized controlled studies of corticosteroids for the reduction of respiratory distress syndrome have failed to demonstrate benefit in very early premature gestational age groups. A randomized, double-blind, placebo-controlled clinical trial of betamethasone given to mothers with intact membranes and threatened premature delivery between 24 and 28 weeks of pregnancy was conducted. Thirty-six patients were randomized to receive betamethasone, two doses of 12 mg, 24 hours apart, and 41 received placebo. No difference was found in the overall incidence of respiratory distress syndrome between the two groups (betamethasone vs placebo 0.55 vs 0.66) or in the incidence of respiratory distress syndrome in babies delivered between 1 and 7 days after the first dose of drug (betamethasone vs placebo 0.78 vs 0.88). Nor were there any differences observed in any measure of severity of respiratory distress syndrome between the groups. The neonatal death rates were also similar (betamethasone vs placebo 0.25 vs 0.24).The only difference seen was an unexpected reduction in the betamethasone group in the incidence of grades 3 and 4 intraventricular hemorrhage (betamethasone vs placebo 1/31 vs 9/36, P = 0.Q1). Therefore this study was unable to demonstrate any beneficial effect of corticosteroids in reducing respiratory distress syndrome at 4cm No cervical examination Parity
o I
Maternal transfer Contractions at admission Regular Irregular None Unknown Admitting diagnosis* Premature labor Pregnancy-induced hypertension Vaginal bleeding IUGR, antepartum fetal distress Other Tocolytics (intravenous magnesium sulfate) Delivery interval (days) 0-1 2-7 ~8
Route of delivery Spontaneous vaginal Cesarean section Sex of live-born infants Male Female Birth weight (gm) Babies weighing 500-1000 gm Babies with low Apgar score ($3 at I min, $6 at 5 min)
Betamethasone (n = 33) (33 neonates)
Placebo (n = 38) (40 neonates)
26.3 ± 5.4
27.1 ± 6.0
25.5 ± 1.2
25.8 ± 1.3
p Value NS NS
4 I
7
(64%) (12%) (3%) (21%)
23 11 1 3
(61%) (29%) (3%) (8%)
NS NS NS NS
14 19 25/33
(42%) (58%) (75%)
22 16 29/38
(58%) (42%) (76%)
NS NS NS
20 8 4
(61%) (24%) (12%) (3%)
17 8 13
(45%) (21%) (34%)
NS NS NS
16 I
(48%) (3%)
22 6
(58%) (16%)
NS NS
7 2
(21%) (6%)
7
(18%) (5%)
NS NS
7
(21%) (73%)
(3%) (89%)
NS NS
21
I
24/33 6 13 14
17 16 18 15 1242 19/33 9/33
2
1
34/38
NS NS NS
II
17 12 (52%) (48%)
18 20
(58%)
14 26 1071 22/40
(27%)
16/40
± 678
(47%) (53%)
NS NS NS
± 597
(55%)
NS NS
(40%)
NS
*Some patients with more than one admitting diagnosis.
confirmation. When the estimated gestational age by sonogram differed by > 14 days from the menstrual dates, the dates were revised to match the ultrasonography-estimated gestational age. Patients were randomized by members of the pharmacy section using a random number-generated, consecutive, sealed envelope system. Patients randomized to the study group received two doses of intramuscular Celestone (6 mg betamethasone acetate and 6 mg betamethasone phosphate) 24 hours apart. Patients randomized to the control (placebo) group received two doses of normal saline solution in the same quantity at 24-hour intervals. Both drug and placebo were prepared in the hospital pharmacy. Management was otherwise unaffected by inclusion in this study. Patients who received tocolytic ther-
apy were given magnesium sulfate in 4 or 6 gm loading doses and maintained initially at 2 gm/hr adjusted to maintain contraction frequency at six or fewer per hour with serum levels not to exceed 7 mg/dl. Patients with ruptured membranes and clinical and/or laboratory evidence of infection, who had previously received any corticosteroids or had medical conditions contraindicating steroids, including diabetes, were excluded from this study. Patients who had not been delivered 1 week after receiving the first dose, and who were judged still likely to be delivered in the next week, were given the same drug or placebo if the gestational age was still within the study inclusion range. Patients who had not been delivered after 28 weeks and were 1 week from their
648
Garite et al.
last dose were allowed treatment with corticosteroids at the discretion of their attending physicians. Newborns were managed by attending neonatologists, and care was not altered by study participation. No neonate received surfactant therapy during the study period. Outcome end points for the study included incidence and severity of RDS and neonatal survival. The diagnosis of RDS required clinical signs and tachypnea and retractions, supplemental oxygen therapy for at least 24 hours, and a chest x-ray film consistent with hyaline membrane disease. Secondary end points considered included maternal and neonatal infectious complications and other neonatal morbidity associated with prematurity, as well as length of neonatal hospitalization. Statistical comparisons were made with t tests, X2 analysis, and Fisher's exact tests where appropriate. Results
The study was conducted between December 1984 and May 1990. During this time 76 patients were successfully randomized, 37 to the betamethasone group and 39 to the placebo group. Of the 37 randomized to betamethasone, four were delivered elsewhere and were lost to follow-up; similarly, one patient was lost from the placebo group. Thus 33 received betamethasone and 38 placebo. All patients lost to follow-up had progressed beyond 29 weeks and would not have affected the analysis of outcome within 1 week of the last drug administered. Each group included three sets of twins. There were three stillbirths in the betamethasone group and one stillbirth in the placebo group. When twins were included and stillbirths were excluded, there were 33 neonates in the betamethasone group and 40 in the placebo group for the final neonatal analysis. Demographic and matching comparisons for the two groups are shown in Table I. As shown, there were no statistically significant differences between the two groups at the time of randomization in the interval to delivery, route of delivery, birth weight, or Apgar scores. Of the 33 mothers in the betamethasone group, six were delivered
o I
Maternal transfer Contractions at admission Regular Irregular None Unknown Admitting diagnosis* Premature labor Pregnancy-induced hypertension Vaginal bleeding IUGR, antepartum fetal distress Other Tocolytics (intravenous magnesium sulfate) Delivery interval (days) 0-1 2-7 ~8
Route of delivery Spontaneous vaginal Cesarean section Sex of live-born infants Male Female Birth weight (gm) Babies weighing 500-1000 gm Babies with low Apgar score ($3 at I min, $6 at 5 min)
Betamethasone (n = 33) (33 neonates)
Placebo (n = 38) (40 neonates)
26.3 ± 5.4
27.1 ± 6.0
25.5 ± 1.2
25.8 ± 1.3
p Value NS NS
4 I
7
(64%) (12%) (3%) (21%)
23 11 1 3
(61%) (29%) (3%) (8%)
NS NS NS NS
14 19 25/33
(42%) (58%) (75%)
22 16 29/38
(58%) (42%) (76%)
NS NS NS
20 8 4
(61%) (24%) (12%) (3%)
17 8 13
(45%) (21%) (34%)
NS NS NS
16 I
(48%) (3%)
22 6
(58%) (16%)
NS NS
7 2
(21%) (6%)
7
(18%) (5%)
NS NS
7
(21%) (73%)
(3%) (89%)
NS NS
21
I
24/33 6 13 14
17 16 18 15 1242 19/33 9/33
2
1
34/38
NS NS NS
II
17 12 (52%) (48%)
18 20
(58%)
14 26 1071 22/40
(27%)
16/40
± 678
(47%) (53%)
NS NS NS
± 597
(55%)
NS NS
(40%)
NS
*Some patients with more than one admitting diagnosis.
confirmation. When the estimated gestational age by sonogram differed by > 14 days from the menstrual dates, the dates were revised to match the ultrasonography-estimated gestational age. Patients were randomized by members of the pharmacy section using a random number-generated, consecutive, sealed envelope system. Patients randomized to the study group received two doses of intramuscular Celestone (6 mg betamethasone acetate and 6 mg betamethasone phosphate) 24 hours apart. Patients randomized to the control (placebo) group received two doses of normal saline solution in the same quantity at 24-hour intervals. Both drug and placebo were prepared in the hospital pharmacy. Management was otherwise unaffected by inclusion in this study. Patients who received tocolytic ther-
apy were given magnesium sulfate in 4 or 6 gm loading doses and maintained initially at 2 gm/hr adjusted to maintain contraction frequency at six or fewer per hour with serum levels not to exceed 7 mg/dl. Patients with ruptured membranes and clinical and/or laboratory evidence of infection, who had previously received any corticosteroids or had medical conditions contraindicating steroids, including diabetes, were excluded from this study. Patients who had not been delivered 1 week after receiving the first dose, and who were judged still likely to be delivered in the next week, were given the same drug or placebo if the gestational age was still within the study inclusion range. Patients who had not been delivered after 28 weeks and were 1 week from their
648
Garite et al.
last dose were allowed treatment with corticosteroids at the discretion of their attending physicians. Newborns were managed by attending neonatologists, and care was not altered by study participation. No neonate received surfactant therapy during the study period. Outcome end points for the study included incidence and severity of RDS and neonatal survival. The diagnosis of RDS required clinical signs and tachypnea and retractions, supplemental oxygen therapy for at least 24 hours, and a chest x-ray film consistent with hyaline membrane disease. Secondary end points considered included maternal and neonatal infectious complications and other neonatal morbidity associated with prematurity, as well as length of neonatal hospitalization. Statistical comparisons were made with t tests, X2 analysis, and Fisher's exact tests where appropriate. Results
The study was conducted between December 1984 and May 1990. During this time 76 patients were successfully randomized, 37 to the betamethasone group and 39 to the placebo group. Of the 37 randomized to betamethasone, four were delivered elsewhere and were lost to follow-up; similarly, one patient was lost from the placebo group. Thus 33 received betamethasone and 38 placebo. All patients lost to follow-up had progressed beyond 29 weeks and would not have affected the analysis of outcome within 1 week of the last drug administered. Each group included three sets of twins. There were three stillbirths in the betamethasone group and one stillbirth in the placebo group. When twins were included and stillbirths were excluded, there were 33 neonates in the betamethasone group and 40 in the placebo group for the final neonatal analysis. Demographic and matching comparisons for the two groups are shown in Table I. As shown, there were no statistically significant differences between the two groups at the time of randomization in the interval to delivery, route of delivery, birth weight, or Apgar scores. Of the 33 mothers in the betamethasone group, six were delivered
