In 1972 Liggins and Howie(1) demonstrated that maternal administration of corticosteroids before preterm delivery resulted in a significant decrease in neonatal mortality and respiratory distress syndrome (RDS). Many studies have since then demonstrated a reduction in the incidence of adverse effects associated with premature birth such as mortality, RDS and intraventricular hemorrhage (IVH)(2-5). Studies have also shown a strong association between hypotension and IVH(6-8). The protective effect of steroids in the prevention of IVH may be related to its effect on the blood pressure of premature neonates. This retrospective cohort study was performed to evaluate the effect, if any, of antenatal steroids on blood pressure in the first 48 hours of life of babies born at less than 33 weeks gestation. We also examined the effect of antenatal steroids on mortality, requirement for ventilatory support, IVH and retinopathy of prematurity (ROP).

All infants born at less than or equal to 32 weeks gestation and admitted to the Neonatal Intensive Care Unit (NICU) between July 1994 and June 1996 were eligible for this study. Babies needing early transfer to other NICU’s because of non availability of beds and babies with life threatening congenital malformations were excluded.

Medical records were reviewed to obtain maternal data including the incidence of pregnancy induced hypertension, antepartum hemorrhage, prolonged rupture of membranes, chorioamnionitis, mode of delivery and use of tocolytic agents. Gestational age was estimated by the date of the last menstrual period and if unavailable by antenatal ultrasound examina-tion. If both details were missing then the gestation was estimated by clinical examina-tion. Other infant details recorded were weight, sex, evidence of early onset infection or hemorrhage, perinatal asphyxia and whether sedation (fentanyl/morphine) was used after delivery.

For this study a complete course of antenatal steroids was defined as the period commencing 24 hours after beginning maternal steroid treatment and lasting 7 days. Two doses of steroids had to be repeated after 7 days to be considered as a complete course. An in-complete course was defined as delivery within 24 hours of starting therapy/or more than 7 days after completion of steroids. Women at risk of preterm delivery received steroids at the discretion of the treating obstetrician.

Data was collected on the total amount of colloid and inotrope infused in the first 48 hours for each infant as an indicator of the blood pressure support required. Colloid included all infusions of 5% albumin in normal saline, 20% concentrated albumin, fresh frozen plasma, packed red blood cells or platelets. The total amount of inotrope infused was assessed separately for dopamine and dobutamine. The inotrope score was calculated as described by Moise et al.(9). The score was obtained by summing the rate in micrograms per kg/min multiplied by the number of hours delivered during the first 48 hours after birth. For example if an infant received 5 mcg/kg per minute of dopamine every hour for 24 hours and then 3 mcg/kg per minute every hour for the next 24 hours then the dopamine score would be (5 ´ 24) + (3 ´ 24) = 192.

All babies had an initial recording of their heart rate, temperature, respiratory rate and blood pressure on admission to the NICU. If the baby was hypotensive, 10-20 ml/kg of colloid was infused. Hypotension was defined according to data from Versmold HT et al.(10). We aimed to maintain a MAP between 25-30 for babies weighing below 1000, 30-35 for babies weighing between 1000 to 1500 g and above 35 for babies weighing above 1500. Dopamine, Dobutamine or both were commenced if the MAP remained below the desired level. The choice of the inotrope depended on the attending neonatologist. Babies requiring ventilation had indwelling arterial catheters and had hourly recording of the systolic, diastolic and mean arterial blood pressures. Babies not requiring ventilatory support had their blood pressures monitored 4-6 hourly.

Data was collected on the ventilatory support required for each infant. The oxygenation index (mean airway pressure  ´  fraction of inspired oxygen  ´  100  ¸ PaO2) was used as a measure of amount of ventilatory support required. This was calculated at 24 and 48 hours for all ventilated infants. Surfactant (Exosurf) was administered within the first 4 hours and repeated after 12 hours if the initial chest X-ray was suggestive of Respiratory Distress Syndrome (RDS) and the baby required ventilation with a FiO2 greater than 0.4. Patients were evaluated for patent ductus arteriosus (PDA), retinopathy of prematurity (ROP) and intraventricular hemorrhage (IVH) using standard protocols and techniques.

Data were analyzed by using the Epi Info Version 6.1 software packages (CDC, Epi-demiology program office, Atlanta Georgia). Categoric data were analyzed by Chi square or Fisher exact probability test. Continuous data that were not normally distributed were analyzed with the Kruskal-Wallis test. Normally distributed numeric data were analyzed by unpaired ‘t’ test. Cornfield 95% confidence intervals were used for the odds ratios.

A total of 174 neonates with a gestation less than 33 completed weeks were admitted to the unit between July 1994 to June 1996. Nine babies were transferred to other Neonatal Intensive Care Units in New South Wales after their initial stabilization, due to non availability of beds (5 had received a complete antenatal steroid course, 2 incomplete and 2 none). These infants were not included in the analysis. Four more babies were excluded due to life threatening congenital malformations (2 with Trisomy 21 and congenital heart defects, 1 with myotonic dystrophy who died within 18 hours and 1 with multiple ileal and jejunal atresias). The remaining 161 babies were analyzed. Of these 161 infants, a total of 77 (48%) received a complete course of antenatal steroid. Forty two (26%) babies received an incomplete course and 42 (26%) did not receive steroids prior to delivery.

Clinical data of the 140 mothers and 161 infants are shown in Table I. There were 17 twin pregnancies and 2 triplet pregnancies in the study group. There was no significant difference found in the maternal characteristics between the 3 groups apart from the occurrence of prolonged rupture of membranes, being higher in the mothers receiving a complete course of steroids (p = 0.02). This difference could be expected as these would be mothers in whom there was sufficient time prior to delivery for the administration of steroids.

There were no significant differences in infant birth weight, gestation or sex between the 3 groups. The incidence of asphyxia, early infection and requirement for sedation with narcotics was not different between the groups. The group receiving no antenatal steroids did have a slightly higher incidence of blood loss during or immediately prior to delivery as determined by the need for urgent colloid or blood administration at the time of delivery or immediately after admission to the NICU (p = 0.05). However the initial volume replacement used for treating the early blood loss in these 4 patients was not included in the analysis for colloid volume.

Babies exposed to a complete course of steroid had a significantly lower mortality (4%) as compared to babies in the incomplete group (10%) and the no steroid group (19%). All deaths in the three groups occurred in ventilated infants and the beneficial effect of steroids on the mortality may be due to its effect on RDS in preterm neonates. Respiratory failure was not a cause of death in the complete steroid group and was responsible for seven deaths in the none/incomplete group. Babies receiving a complete course of steroid required less support to maintain their blood pressure in the first 48 hours after birth. A total of 65% of babies in this group required colloid infusions in the first 48 hours compared to 81% in the incomplete group and 83% in the no steroid group (p = 0.04). Also significantly fewer babies in the complete group required inotropic support as compared to the other two groups (complete 30%, incomplete 48%, none 52%; p = 0.03).

As the primary analysis showed no difference between babies in the no steroid and incomplete steroid groups with respect to their clinical outcomes, further analysis was done with these 2 groups combined together. These results are shown in Table II. Babies who had received a complete course of antenatal steroid were significantly more hemodynamically stable (Table II). Babies in this group were less likely to need colloid to maintain their blood pressure (p = 0.013) and if they did, they required a smaller volume (p = 0.0003). Despite the higher percentage of babies receiving larger volume of colloid in the none/incomplete group, more babies in this group required inotropic support as compared to the complete steroid group (p = 0.009). Since Dobutamine is associated with a higher failur rate in increasing the blood pressure, infants having received dopamine as the solitary inotrope in the two groups was also analyzed. Although, this comparison was also significant (p = 0.032), the difference between the two groups was narrower as compared to the difference with combined inotropic require-ment. There was no significant difference detected in the inotrope scores between the two groups. However, six babies in the none/incomplete group died within 24 hours and as these would have been the sickest babies with the highest inotrope requirements, this may have skewed these results away from statistical significance.

No differences were detected in the number of babies requiring any form of ventilatory support or oxygen therapy (Table II). However, significantly more babies in the none/incomplete group required surfactant (p 0.008). Also, babies in the complete group had significantly lower oxygenation and ventilation indices at 24 and 48 hours of age. Thus, although the proportion of babies in the two groups requiring ventilation was similar, babies in the none/incomplete group required more respiratory support in the form of higher ventilatory pressures and surfactant.

Only 9 of 157 (6%) babies developed severe IVH (grade III or IV IVH). Four babies died prior to head ultrasound examinations, 3 in the no steroid group and 1 in the incomplete group. However, the number of IVH’s in the study group is too small to show any significant difference. No significant difference was detected in the incidence of PDA or ROP in the three groups.

Table I: Clinical Dala

APH–antepartum hemorrhage, LSCS–lower segment Caesarian section, PIH–pregnancy induced hypertension, IUGR–intra uterine growth retardation PROM–prolonged rupture of membranes, GBS–Group B hemolytic streptococci isolated from high vaginal swab.
Values for weight and gestation are mean ± SD.
Figures in parantheses represent percentages.

Table II: Outcomes: Blood Pressure and Respiratory Support