Value of antenatal corticosteroid therapy in preterm birth

Value of antenatal corticosteroid therapy in preterm birth

Arsenio Spinillo*‘, Ezio Capuzzu”, Alessslra Chn&tob, Mauro Stronatib, Federica 13altaroa, Angda Iascia

pCiiniw Ostetrico-Ginewlogiw. Vniveesiti a? Pavk. Pa*, Iroly bDivisione di Patologia Neomtale, IRCCS Policlitdco San Mafteo. P. le Go&i 2, 271W Pavio. Italy

Received 22 October 1994, revision received 14 February 1995: accepted 28 Feimary 1995

This prospective observational study was des@ed to evaluate the mqp&de af the corticosteroid-related reduction in the risk of respiratory distress syndrome (RDS), ,&raven- tricular hemorrhage and neonatal death according to different etiologic subgroups of preteen delivery. Of 380 patients delivered before 35 weeks’ gestation, 155 received a con&@&e ccnu%e of dexamethasone or betamethasoae to promote fetal lung maturatioa. In logistic &, the steroid-related reduction of RDS was greater among patients with intact as oppod to patients with premature ruptan of metnm (-lraars risk = 0.3 I; ce intervaI(C.l.)= 0.13-0.73; P=O.~)~ia~~~~~~~ lrrbara5mm- pared with other etiologic groups (excess risk 3: 0.33; 95% C.I. = other hand, the steroid-related rcduetion of severe (grade ill-iv) was more marked in growth retarded as opposed to w@&grow 95%C.I.=O.O3-0.96, P=0.04),aadin~as~slpe*sdwithapw(etocourprataaardcEiv" eriea (excess risk = 0.15; 95% C.I. = 0.03-O.%, P = 0.04). Results show &at the mtcsit be=- fit from antenatal steroids appears to be in preterm deliveries with intact membranes and in planned preterm &liveries.

Keywords: Corticosteroids; Respiratory distress syndrome; Intracranial hemorrage

Respiratory distress syndrome (RDS) is a leading ause of mortality and mo&idi- ty in premature neonates [l). Many controlled trials have shown that maternal ad-

* Cortwpondiag author, Clinka Oste&o-Ginccologica, IRCCS Policlinico San Matteo, P. le Go@ 2, 27100 Pavia, Italy. Tel.: +39 382 526215; Fax: +39 382 423233.

0378-3782/95/$09.50 0 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0378-3782(95)01638-J

38 A. Spinillo et al. /Early Human Development 42 (1995) 37-47

ministration of corticosteroids reduces the occurrence of neonatal RDS in preterm deliveries [2-51. However, 22 years after the classic work of Liggins and Howie [2], differences of opinion still exist among obstetricians about the place of antenatal steroid therapy. An overview of the data from controlled trials [6] concluded that antenatal corticosteroid administration reduces the risk of RDS by - 50%. However, the majority of the trials examined in this meta-analysis included only patients with spontaneous preterm labor or premature rupture of the membranes (PROM) [2-61. The effect of steroids on RDS has not been adequately studied in major etiologic categories of preterm delivery such as pre-eclampsia, fetal growth retardation and twin pregnancy [7]. In addition, there are still differences of opinion regarding the routine use of steroids in patients with PROM [8,9]. In clinical practice, it is still not clear whether fetuses of planned preterm deliveries because of pregnancy complications may benefit from antenatal steroid therapy (81.

In an attempt to identify those subgroups of preterm delivery that may benefit from antenatal corticosteroid administration, we conducted a prospective observational study to evaluate the effect of steroids on three different neonatal outcome measures (respiratory distress syndrome, intraventricular hemorrhage and neonatal death). This effect was studied by a multivariate approach in six different clinical subgroups (singleton versus twin infants, induced versus spontaneous preterm delivery, well-grown versus growth-retarded fetuses, normotensive versus hypertensive pregnancies, preterm premature rupture of fetal membranes versus intact membranes, and spontaneous preterm delivery with intact membranes versus ‘other’ subgroups).

2. Subjects aod methods

The population under study comprised selected patients delivered preterm at our institution in the period 1988-1993. The specific inclusion criteria into the study were as follows: (1) preterm delivery between 24-34 weeks’ gestation; (2) undeter- mined ( < 2) or immature lecithinlsphingomyelin ratio; (3) livebom, non-malformed infant; (4) either optimal corticoateroid treatment (cases) or no corticosteroid treatment (controls); and (5) planned delivery for patients with medical complications of pregnancy (e.g. pre-eclampsia, fetal growth retardation, placenta praevia). We considered as ‘optimal’ corticosteroid treatment the administration of two doses of 12 mg of betamethasone or dexamethasone, with the first dose 248 h and d 7 days before delivery. Although an ‘incomplete’ course of steroids may have an impact on the occurrence of RDS, this category was excluded from analysis to avoid the com- plexity of a three-group interaction study. Triplets and pregnancies complicated by severe abruption or eclampsia were excluded from the analysis. In the period of the study, it was the policy of our institution to administer corticosteroids routinely to patients in spontaneous preterm labor with intact membranes and cervical dilatation ~3 cm, including twin pregnancies. In all other cases (e.g. PROM, fetal growth retardation, pre-eclampsia), steroids were administered at the discretion of the atten- ding physician. In spontaneous preterm labor or PROM, tocolytics (ritodrine or isoxsuprine) were usually added to glucocorticoids in order to delay delivery. As

A. Spinillo et al. /Early Human Development 42 (1995) 37-47 39

repmted in Table 1, there was no evidence of an overall inomasing tmnd in steroid use dtiag the period of the study. FIowever, in the last 2 years of the study, cor- ticotids were administered more frequently in moat of the

Fetal growth retardation was diagnosed when circumference was below the 10th centile of our graphic examinations and birthweight was below the se All patients had undergone sonography within 1 tional age. Maternal pre-eclampsia was defkked et al. Ill]. Of an initial population of 435 eligible patients who received an incomplete course of corti cause of fetal malformations. The final population of 380 pregnancies and 424 fetuses (44 twin data were collected prospectively and stored responsible for neonatal diagnoses were steroid administration. The diagnosis of c~~o~~itis was temperature > 37.8”C plus two of the followings: (1) maternal beatshin; (2) basehne fetai heart rate > 160 beat&mm; (3) positi endocervix or the inner third of the vagina; (4) maternal white blood eel1 mt > 18.0Q#mm3; (5) foul-smelling amniotic fluid. After birth, all infants were to our Pkonatal Intensive Care Unit which has the faeihtks of a te&uy ca All the infants with diagnosis of RDS had physioal signs of respiratoq (grunting, chest retraction, tachypnea) and required vent&tory support fur >48 h. Each case was confirmed radiologically.

Table 1 Rates of optimal steroid treatment during the study period

Steroid 1988-1989 1990-1991 1992-1993 x, for P treatment trend

Twins No 5 7

Yes No

Indicated delivery Yes No

Fetal growth retardation Yes No

Pre-echunpsia Yes No

Premature rupture of membranes YeS No

Spontaneous pretcrm labor YeS

6 3 8 52

16 10 8 38

8 4 2 22

8 2 24 26

22 10 16 48

8 12

11 0.009 0.92

12 30

1.81 0.18 50 18

1.99 0.16 20 14

1.1 0.29 26 36

2.34 0.13 50 14

2.4 0.12 20

40 A. Spinillo et al. /l%rly Human Development 42 (1995) 37-47

In all the infants, a standard transfontanellar echoencephalogram was obtained within 24-48 h of admission. Subsequently, serial cranial ultrasound examinations were carried out every 48-72 h during the first 2 weeks of life and then weekly until discharge according to a definite protocol including: (1) very-low-birthweight (< 1500 g) infants; (2) infants with acidosis or resuscitated in the delivery room or on ventilatory support; (3) infants from high-risk pregnancies (i.e. pre-eclampsia, prolonged PROM, severe fetal growth retardation); (4) infants with suspect ultrasound brain imaging on the first examination or with abnormal neurological findings. Infant intracranial hemorrhage was graded into four categories according to Papile et al. [12]. In this type of categorization, grade I-II hemorrhages are not associated with increased prevalence of moderate or severe infant handicap [ 12,131.

Continuous variables were compared using the Mann-Whitney test. Univariate associations between corticosteroid treatment and the three outcome measures studied were tested by the computation of crude odds ratios (ORs) and 95% conti- dence intervals (C.I.). The corticosteroid-related reduction in the risk of the outcomes (excess risk) studied in each of the six subgroups was evaluated by a multivariate study of interactions. First-order interactions between corticosteroid treatment and the six clinical groups were tested with unconditional logistic models [14]. According to the hierarchy principle [14], a term each for corticosteroid treatment and interaction factor (clinical group), together with an interaction term (product coefftcient term), was included in the logistic model to analyze the effect of each interaction on the outcome. Subsequently, the corresponding coefficients were adjusted for the effect of other confounding factors, such as gestational age, birthweight and sex of the infant. Each interaction was considered statistically significant if its inclusion (product coefficient term) improved the model fit (P < 0.05) based on the likelihood ratio test.

3. Results

The mean maternal age was 29.1 years & 4.9 in the steroid group (cases) and 28.7 years f 4.8 (S.D.) in the non-treated group (controls) (P = 0.48 by Mann-Whitney test). The main maternal and pregnancy characteristics of the population under study are reported in Table 2. Patients treated with corticosteroids were more likely to have been treated with tocolytic drugs and to have ruptured membranes than controls. Among patients with PROM, the median time from rupture to delivery was longer in cases (4.1 days, range l-52) than in controls (2.1 days, range l-77; P < 0.0001 by Mann-Whitney test). In the subgroup of PROM, however, the mean gestational age at delivery was similar between the steroid and the control group (30.2 weeks f 2.3 in PROM receiving steroids as opposed to 30.2 weeks f 2.8 in PROM receiving no steroids; P = 0.89 by Mann-Whitney test).

Table 3 shows the main neonatal variables. There were no differences regarding gestational age and birthweight between the corticosteroid and the control group. Infants with a diagnosis of fetal growth retardation were treated to promote lung maturation less frequently than infants with appropriate growth. In univariate analysis there were no differences regarding the three outcome measures studied between

A. Spinillo et al. /Eddy Human Development 42 (1995) 37-47 41

Table 2 Maternaland prcgwwy chaNcteristics of 155 pmlgraacies, treated with oosticosrcroid and 225 controls

- Steroid group control group (N = 155) (N = 225)

n % n % --

Nulliparity 98 63.2 143 63.6 Prwious aboftions 54 34.8 52 23.1 Previous low birth weight 23 14.8 30 13.3 smdciapinpreeaancy 50 32.2 65 28.9 First t&nester hemorrhage 40 25.8 50 22.2 Diabetes 3 1.9 2 0.9 Placenta praevia 15 9.7 10 4.4 Tocolysis 109 70.3 72 32.0 Wksnpsia 30 19.3 37 16.4 PreUWuc Rlpture of membranes 80 51.6 90 40.0 cwafeansection 74 47.7 120 53.3 Spontaneous prctcrm labor 30 19.3 7i 31.6 Twittpqplancy 21 13.5 23 10.2 Chorioamnionitis 10 6.5 13 5.8

All dR&ences betweem the steroid and the control group were not statistically significant at P < 0.05.

Table 3 ChamcWstics of 176 infants of corticostcroid-treated mothers and 248 controls

Steroid group Control group (N = 176) (N = 248)

MWtl S.D. MCiUl SD. -

Gestational age (weeks) 30.4 2.6 30.5 2.9 Birdwei&t (g) 1423 445 1468 541

Mak sex 98 55.7 110 44.3 Fetal growth retardation 32 18.2 64 25.8 Sepsis 20 11.4 28 11.3 Acidosis (pH < 7.2) 32 18.2 52 21.0 Mechan&l ventilation 54 30.7 82 33.1 Bradycardia 78 44.3 102 41.1 seb 4 2.3 10 4.0 Patent ductus arteriosus 8 4.6 8 3.2 Respiratory distress syndrome 66 31.5 108 43.5 Mild in~nial hemorrhage 18 10.2 24 9.7 Severe intracranial hemorrhage 12 6.8 30 12.1 Bronchopubnonary dysplasia 6 3.4 12 4.8 Cystic periventricular kucomalacia 6 3.4 10 4.0 Neonatal death (within 28 days) 24 13.6 36 14.5

All difkrences between the steroid and the control group were not statistically signifkant at P < 0.05.

n % n %


infants of steroid-treated mothers and controls. In the steroid group, the median Apgar score at 1 and 5 min was 7.5 (range O-10) and 8.5 (range 3-lo), respectively, compared with 7 (range O-10, P = 0.40 by Mann-Whitney test) and 8 (range 4-10, P = 0.25 by Mann-Whitney test) in the controls. After adjustment for the effect of gestational age, birthweight and sex of the infant, which were the main confounding factors for the outcome, the adjusted odds ratio of respiratory distress syndrome was 0.72 (95% C.I. = 0.48-1.10; P = 0.13), whereas for severe (grade III-IV) hemorrhage and neonatal death the adjusted odds ratios were 0.53 (95% C.I. = O-28-0.98; P = 0.045) and 1.02 (95% C.I. = 0.54-1.94; P = 0.94), respectively.

Table 4 reports the prevalence of RDS according to a course of glucocorticoid therapy in the main subgroups of preterm delivery. In stratified, univariate analysis, glucocorticoid administration was associated with a greater reduction in the risk of respiratory distress syndrome in twins as opposed to singleton infants, in patients with intact membranes with respect to PROM and in patients with spontaneous preterm labor with intact membranes than in other etiologic subgroups of preterm delivery. Interaction factors indicating the glucocorticoid-associated reduction in the risk of RDS (excess risk) were subsequently corrected for the confounding effect of gestational age and birthweight using logistic models. In these models the glucocorticoid-associated reduction in the risk of RDS remained significantly greater among twins as compared with singleton infants (excess risk = 0.31; 95%

Table 4 Odds ratios of respiratory distress syndrome (RDS) according to cortkostaroid treatment and subgroup of preterm delivery

Twins Singleton Indiited delivery Other than indicated

delivery

Steroid group

RDS Unatfected (n = 66) (n = 110)

IO 32 56 78 26 50 40 60

Control group

RDS Unaffe&d (n = 108) (n = 140)

30 16 78 124 46 44 62 96

OR (95% CL)”

0.1.7 (0.06-0.42) 1.14 (0.73-1.78) 0.50 (0.26-0.83) I .03 (0.62-l .72)

Fetal growth retardation

Normal fetal growth Pre-e-clampsia Normotensive

Intact membranes PPROM Spontaneous preterm

delivery Others

IO 22 30 34 0.51 (0.21-1.26)

56 88 78 106 0.86 (0.55-1.35) 14 22 20 18 0.57 (0.23-1.44) 52 88 88 122 0.81 (0.53-1.27)

28 66 72 80 0.47 (0.27-0.81) 38 44 36 60 1 A4 (0.79-2.62) 8 32 34 44 0.32 (0.13-0.79)

58 78 74 96 0.96 (0.61-1.52)

PPROM, preterm premature NptUlT of membranes. ‘odds ratio (OR) and 95% contkkncc interval (CL).

A. Spiniilo et al. /Early Human Lkvekqmmt 42 Cl*S) 37-47 43

C.I. = 0.10-0.94; P = 0.038), in patients with intact membranes as opposed to PR0M (excess risk = 0.31; 95% C.I. = 0.13-0.73 P = 0.007) and in q@ieMs with spontaneous pruterm labor with intact membranes than in ‘other’ subgroups (ex@ss risk = 0.33; 95% C.I. = 0.11-0.98; P = 0.04).

TaMe 5 reports the stratified analyses regarding the effect ofcortico&eroids on the risk of severe (grade III-IV) intracranial hemorrhage. In univariate an&sis, the glucocorticoid-associated reduction in the risk of severe hemorrhage was more marked among growth retarded than in appropriately grown infants. After adjustment for the effect of gestational age and birthweigbt, this risk r&u&on was signiticantly greater among growth retarded as opposed to well-grown iafants (excess risk = 0.15; 95% C.I. = 0.03-O.%; P = O&t), and in planned than in spontaneous preterm deliveries (excess risk = 0.15; 95% C.I. = 0.03-0.96; P = 0.04). Since no case of severe hemorrhage was recorded among twins or in of pfe- eclamptic mothers who received glucocorticoid therapy, it was not possible to carry out a reliable multivariate analysis of these two subgroups. The rate of severe bemor- rhage was lower among twins and infants born to preedampti~ mothers who had been given corticosteroids when compared with controls (none of 40 versus 8 of 90;

Table 5 Odds ratios of severe (grade III-IV) intraventricular hemorrhage (Sev IVH) according to corticosteroid treatment and subgroup of preterm delivery

Steroid groupa Control groupa OR (95% C.IJb -

Sev IVH Unaffected Sev NH Unaffected (n = 12) (n = 146) (n=30) (n= 194)

Twins Singleton Indicated delivery Other than indicated

delivery

Fetal growth retardation

Normal fetal growth Pm-edampsia Normotensive

Intact membranes PPROM Spontaneous preterm

delivery others

0 40 8 32 Undefined 12 106 22 162 0.80 {0.38- 1.69) 2 64 10 68 0.21 (O&a-l.Qo)

10 82 20 126 0.77 (0.34-I .72)

2 26 18 40

8 110 22 128 0.42 (0.18-0.99)

120 12 154 28 6 28

118 24 166

78 18 120 68 12 74 36 8 66

0.17 (0.04-0.80)

1.07 (Q.45-2.55) Undefined 0.70 (0.34- 1.46)

0.51 (0.19-1.35) 0.54 (0.19-1.53) 0.92 (0.26-3.25)

PPROM, preterm premature rupture of membranes. ‘42 patients (18 in the steroid group and 24 in the control group) with mild (grade I-II) intracranial

hemo&age were not included in the analysis. bonds ratio (OR) and 95% confidence interval (CL).

44 A. Spinillo ef al. /Early Human Development 42 (1995) 37-47

P = 0.003, and none of 28 versus 6 of 34; P = 0.02, respectively, by Fisher exact test). However, twins in the glucocorticoid group had slightly higher birthweights and longer gestations compared with control twins (3 1.2 weeks f 1.8 versus 29.8 f 2.4; P < 0.0001, and 1590 g * 426 versus 1450 f 482; P = 0.0013, respectively). On the other hand, birthweights and gestational ages of infants of pre-eclamptic mothers were similar between glucocorticoid-treated and control mothers (30.8 weeks i 2.2 versus 31.2 f 2.6; P = 0.32, and 1273 g f 415 versus 1284 +Z 464, P =-0.71, respectively).

Finally, in multivariate analysis, there were no significant interactions between steroid therapy and subgroups of preterm birth with regard to overall (grade I-IV) hemorrhage or neonatal death.

4. -on

The rationale of this study is based on the hypothesis that the etiology of preterm delivery affects the corticosteroid-related reduction in the risk of RDS, intraventricular hemorrhage or neonatal death. This hypothesis is supported by data from previous studies. Both randomized studies and meta-analyses [2-81 have shown that in preterm deliveries, maternal corticosteroid administration reduces the risks of neonatal RDS, necrotizing enterocolitis, intracranial hemorrhage and, although to a lesser extent, neonatal death. These effects have been demonstrated mainly for singleton pregnancies and spontaneous preterm labor [5,6]. In fact, subgroups such as diabetes, hypertension or fetal growth retardation had been barred from analysis in most of the randomized studies of glucocorticoid administration [6,7]. Within these high risk groups, the effectiveness of steroid administration is still uncertain [6,7]. In addition, antenatal steroid therapy seems to have a reduced efficacy after membrane rupture, and also no significant respiratory benefit from steroids has been shown for multiple pregnancies [5,8,9]. The implications of these potential limita- tions of glucocorticoid use in obstetric practice are very important. The etiology of preterm birth is very heterogeneous; in two recent studies [15,16] reviewing this problem, the rate of indicated preterm delivery because of hypertension or fetal growth retardation accounted for 18-30% of births before 35 weeks gestation. In addition, PROM and twin pregnancies account for - 30% and lo- 15% of preterm deliveries, respectively [17]. On the basis of these data, when twins and planned preterm deliveries are grouped together as a subset in which the effectiveness of steroids is unclear, -4O-50% of preterm babies may receive no benefit from treatment. As a result, there continues to be variation in obstetricians’ use of antenatal corticosteroids. Studies from Europe and the USA [ 18,191 indicate that corticosteroids are used routinely in lo-30% of preterm deliveries. In addition, some major perinatal centers in the USA do not recommend antenatal steroid use at all [20]. At our institution the policy regarding corticosteroid use is in agreement with current recommendations [6,20-221 and it has not been changed during the study period.

The interaction between the various subgroups of preterm delivery and antenatal steroid therapy has not been adequately addressed in previous studies. Most analyses

A. SpihilIo et al. /Early Hmmm Devdqmtvt 42 /NW) 37-47

have been carried out separately for subgrou trial (e-g. twins, PRUM, planned dalivery) ologieal approach in such an in~estigatiun sh teractions between steroid use and subgroups of preterm delivery on the cu&ome studied [ 14,231. In the present investigation we sought tu c?tThwy in standard obstetrical practice by taking into account intera&ion between steroid administration and etiofogy of the observational nature of the study, the possibility of a excluded. However, the entry criteria utilized and the use o to account for the effect of potential confounders shoukl rniniftliae thi$

Overall, steroid use was associated only with a weakly sign&cant red&&on in RDS. However, the design of the analysis, the exchrsion criteria used (e.g. exclusion of babies receiving an ‘incomplete’ steroid course) and the limit& santp#e s&e can well explain this fmding. Regarding RDS, our study confirms that the ben&iial ef&ct of corticosteroids is greater in patients with intact ~~ and in those with spontaneous preterm labor. In these two groups, ghicocorticoid ~~~~o~ was asgdated with a 70% reduction in the risk of RDS in ~, w&k other categories at risk. We were not able to demonstrate a ben&cial e&I&t of saps in patients with PROM. Although two meta-analyses of p&&bad data (&&Q have reported that antenatal steroids are associated with a reduction in the risk o-f RDS in such a group, this finding has been criticined [9,22& AU but one of the sevarr randomized studies examined in a me&t-analysis ]6] showed no benefit from corticosteroid administration in patients with PROM. The study of Morales et 1. ]24]* which alone included 28% of the patients considered in that meta-analysis, has been widely criticized, and, according to other authors 19,221, it may have biased the con- clusions drawn from the overall analysis. In addition, the evaluation of the &&et of steroids in patients with PROM is more complex than in patients with intact membranes 168). Additional treatments such as induction of labor or expectant nmnage- ment can affect the response to steroids [21,22]. In agreement with several authors [9,21,22], we believe that additional data are needed to clarify the efficacy of steroid treatment in patients with PROM.

With regard to the prevention of RDS, twin pregnancy is considered a condition of uncertain steroid efficacy [6]. A secondary analysis of the data from the Collaborative Group on Antenatal Steroid Therapy showed that the reduction in the risk of RDS after antenatal dexamethasone administration was not statisticaily significant among twins [S]. However, other studies contradict these results Q5]. In our analysis, the overall corticosteroid-related reduction in the risk of RDS was more marked among twin than singleton infants. Although this effect can be at- tributed to the higher prevalence of spontaneous preterm labor with intact membranes among twin as compared with singleton pregnancies (50/88 versus 681336; P < O.OOl), our results strongly support the routine use of steroids in multiple pregnancies.

In the present study, antenatal steroid use was associated with an overall reduction in the risk of intraventricular/intraparenchymal hemorrhage. This confirms the results of previous investigations on preterm deliveries with different etiology [26]. In-


terestingly, the beneficial effect of corticosteroids on the occurrence of intracranial hemorrhage was more pronounced in growth retarded as compared with well-grown fetuses, and in planned as opposed to spontaneous preterm deliveries. Although it was not possible to correct for confounders, this beneficial effect was also more evi- dent among hypertensive than among normotensive pregnancies. It has been sug- gested that the steroid-related reduction in the risk of intracranial hemorrhage could be related both to an indirect effect on neonatal lung function and a direct effect on the immature brain vasculature [6,7,26]. In our population, the use of steroids did not reduce the rate of RDS among growth-retarded fetuses or infants of pre- eclamptic mothers. Although it has been reported that steroids can improve neonatal pulmonary function without reducing the overall rate of RDS [26], our findings support the possibility of a direct effect on brain vasculature.

Finally, steroid administration did not significantly affect the risk of neonatal death in any of the subgroups studied. Given the dimension of the study, this result was not surprising. For example, considering the neonatal mortality rate in the subgroup of pre-eclamptic pregnancies, our data suggest that a study of 241 pre- eclamptic patients and an equal number of normotensive controls would give a 90% chance of detecting a 40% steroid-related reduction of the incidence of neonatal death among hypertensive pregnancies, with a probability of < 0.05.

In conclusion, the results of this study have shown that the steroid-related reduction in the risk of RDS is greater in patients with intact membranes, in those with spontaneous preterm labor and in twin pregnancies as opposed to other etiologic categories of preterm birth. On the other hand, the corticosteroid-related reduction in the risk of neonatal intraventricular/intraparenchymal hemorrhage is greater among growth-retarded compared with well-grown fetuses and in planned than spontaneous preterm deliveries.

References

[I] Robertson, P.A., Sniderman, S.H., Laros, R.K. et al. (1992): Neonatal morbidity according to ger- tational age and birthweight from five tertiary care centers in the United States, 1983 through 1986. Am. I. Obstet. Gynecol., 166, 1629-1645.

[2] Liggins, G.C. and Howie, R.N. (1972): A controlled trial of antepartum ghtcocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics, 50, 515-525.

[3] Papageorgiou, A.N., Desgranges, M.F., Masson, M., Colle, E., Shatx, R. and Gelfand, M.M. (1979): The antenatal use of betamethasone in the prevention of respiratory distress syndrome. A controlled double blind study. Pediatrics, 63, 73-79.

[4] Taeusch, H.W., Frigoletto, F., Kitzntiller, J. et al. (1979): Risk of respiratory distress syndrome after prenatal dexamethasone treatment. Pediatrics, 63, 64-72.

[S] Collaborative Group on Antenatal Steroid Therapy (1981): Effects of antenatal dexamethasone administration on the prevention of respiratory distress syndrome. Am. J. Obstet. Gynecol., 141, 276-287.

[6] Crowley, P., Chalmers, I. and Keirse, M.J.N.C. (1998): The effects of corticosteroid administration before preterm delivery: an overview of the evidence from controlled trials. Br. J. Obstet. Gynaed, 97, I l-25.

[7] Crowley, P. (1991): Promoting pulmonary maturity. In: Effective Care in Pregnancy and Childbirth, pp. 746-764. Editors: I. Chalmers, M. Enkin and M.J.N.C. Keirse, Oxford University Press, New York.

A. Spinillo et al. /Early Human Development 42 (1995) 37-47 47

[8] Ohlsson, A. (1989): Treatments of preterm premature rupture of membranes: a me&analysis. Am J. Obstet. Gynecol., 160, 890-898.

[9] Eriksen, N.L. and Blanco, J.D. (1991): The role of corticosteroids in the maaaspanent of patients with preterm premature rupture of the membranes. Clin. Obstet. Gyoecol., 34, 694-701.

[IO] Bossi, A., Caccamo, ML., De Scrilli, A.M. and Milani, S. (1980): Standard de1 peso de1 neonato italiano @alla 32a alla 43a settimana di gestaxione). Riv. Ital. Ped., 6, 153-170.

[I I] Davey, D.A. and MacGilhvray, I.M. (1988): The classif~tion and definition of the hypertensive disorders of pregnancy. Am. J. Obstet. Gynecol., 158, 892-898.

(12) Papile, L., Burstein, J., Burstein, R. and Kotller, H. (1978): Incidenoe and evolution of subepen- dymal and intraventricular hemorrhage: a study of infants with birth weights less than 1500 g. J. Pediatr., 92, 529-534.

[13] Stewart, A.L., Reynolds, E.O.R., Hope, P.L. et al. (1987): Probability of neurodevelopmental disorders estimated from ultrasound appearance of brains of very preterm infants. Dev. Med. Child Neurol., 29, 3-11.

[14] Breslow, N.E. and Day, N.E. (1980): The analysis of case-control studies. In: Statistical Methods in Cancer Research. IARC Scientific Publication, Lyon, France.

I151 Tucker, M.J., Goldenberg, R.L., Davis, R.O. et al. (1991): Etiologies of preterm birth in an indigent population: is prevention a logical expectation? Obstet. Gynecol., 77, 343-347.

(161 Savitx, D.A., Blackmore, C.A. and Thorp, J.M. (1991): Epidemiologic characteristics of preterm delivery: etiologic heterogeneity. Am. J. Obstet. Gynecol., 164, 467-471.

1171 lams, J.D., Johnson, F.F. and Cmasy, R.K. (1988): Prevention of preterm birth. Clin. Obstet. Gynecol., 31, 599-615.

[18] Keirse, M.J.N.C. (1984): Obstetrical attitudes to glucocorticoid treatment for lung maturation: time for a change? Eur. J. Obstet. Gynecol. Reprod. Biol., 17, 247-255.

119) Hack, M., Ho&u, J.D., Malloy, M.H., Tyson, J.E., Wright, E. and Wright, L. (1991): Very low birth weight outcomes of the National Institute of Child Health and Human Development neonatal network. Pediatrics, 87, 587-597.

1201 Cunningham, F.G., MacDonald, P.C., Leveno, K.J., Gant, N.F. and Gilstrap III, L.C. (1993): Williams Obstetrics. Prentice-Hall International, London, pp. 864-865.

121) Svigos, J.M., Robinson, J.S. and Vigneswaran, R. (1994): Threatened and actual preterm Labour including mode of delivery. In: High Risk Pregnancy. Managemen t Options, pp. 151-164. Editors: D.K. James, P.J. Steer, C.P. Weiner and B. Gonik, WB Saunders Co., London.

1221 Asrat, T. and Garite, T.J. (1991): Management of preterm ptcmature rupture of membranes. Clin. Obstet. Gynrcol., 34, 730-741.

[23] Altman, D.G. (1994): Interaction. Brit. J. Obstet. Gynaecol., 101, 959. 1241 Morales, W.J., Diebel, N.D., Laxar, A.J. and Zadroxny, D. ( 1986): The effect of antunatal dex-

amethasone on the prevention of respiratory distress syndrome in preterm gestations with premature rupture of membranes. Am. J. Obstet. Gynecol., 154, 591-595.

[25] Ardila, J., Le Guennec, J. and Papageorgiou, A. (1994): Influence of antenatal betamethasone and gender cohabitation on outcome of twin pregnancies 24 to 34 weeks of gestation. Sem. Perinatol., 18, 15-18.

[26] Maher, J.E., Cliver, S.P., Goldenberg, R.L., Davis, R.O. and Copper, R.L. (1994): The effect of corticosteroid therapy in the very premature infant. Am. J. Obstet. Gynecol., 170, 869-873.

Documents

Value of antenatal corticosteroid therapy in preterm birth