Skip navigation

GRiP - Steroids in Preterm Delivery

Background

About 1.4% of babies are born prematurely before 34 weeks of gestation, for a variety of reasons. The most common complication of preterm birth is respiratory distress syndrome (RDS), and this affects 50% of babies born before 34 weeks gestation.

Other complications of preterm birth include periventricular-intraventricular haemorrhage (PIVH) which affects 15-20% of preterm infants. PIVH generally occurs within six hours of birth and is associated with an adverse neurodevelopmental outcome in preterm babies.

The treatment of RDS has been estimated to account for over half of all bed-days in neonatal intensive care units and a third of all special care baby unit bed-days.

Evidence of effectiveness

As long ago as 1969 researchers noted that lambs born preterm after exposure to corticosteroids in utero survived longer than control lambs. The randomised, placebo-controlled study of betamethasone administration in women who were expected to give birth prematurely by Liggins & Howie was published in 1972. It showed that the frequency of RDS in babies born before 32 weeks gestation was significantly less after corticosteroids, and that there was a five-fold reduction in neonatal mortality after corticosteroid.

A detailed systematic review (meta-analysis) has been published (Crowley 1990): it examined 12 controlled trials involving over 3000 participants, and demonstrated clearly that antenatal administration of corticosteroids (24 mg betamethasone, 24 mg dexamethasone or 2 g hydrocortisone) resulted in significant reduction in the incidence of RDS in preterm infants. The magnitude of the reduction was of the order of 40-60%.

The effects of steroids were most pronounced when administered 24-168 hours before delivery, and in pregnancies of less than 31 weeks gestation. Importantly, though, there was no sub-group of babies identified for which it could be concluded that corticosteroid administration was not associated with a reduction in the risk of neonatal respiratory morbidity.

Further benefits from steroids in pre-term delivery come from effects on forms of neonatal morbidity. Although the Crowley study had RDS as the main outcome measure, it also examined other measures of morbidity.

A key finding was that the incidence of PIVH and necrotising enterocolitis were reduced by 10 to 80% after steroid administration. This in turn leads to shorter mean durations of hospital stay in infants of corticosteroid-treated mothers.
A further important outcome was neonatal death. The risk of neonatal death was reduced by between 25 and 50% where corticosteroids were administered. Moving from a situation where 30% of eligible mothers received steroid to one where 70% received steroids would save 400 infants deaths a year.

Possible risks to mother or baby

Despite very considerable analysis of the data, especially looking at infection in premature rupture of membranes, no identifiable risks associated with corticosteroid administration were found.

The conclusion is that, except in special circumstances, it is a safe treatment.

What proportion should be treated?

International studies have shown that the uptake of steroids in premature delivery can be as high as 80% or more. In the UK, teaching hospitals have been shown to have an uptake of approaching 50%, but in district general hospitals it can be as low as 35%. Women in teaching hospitals or regional centres were twice as likely to have received antenatal steroids for more than 24 hours than those in DGHs.

A consensus would seem to be that at least 70% of eligible women should receive antenatal corticosteroids. There are some arguments as to whether the appropriate gestational cut-off is 23 or 34 weeks, but any figure less than 70% is causing unnecessary mortality and morbidity, and placing unnecessary costs on the health service.

What are the cost implications?

The corticosteroid drugs themselves are very cheap. Babies who develop RDS are usually treated with exogenous pulmonary surfactant, at an approximate cost of £1000 per treatment. Reducing the number of RDS cases has large implications for drugs budgets.

The implications for cost savings for the NHS are large in other areas. A neonatal intensive care cot has been estimated to cost between £340 and £1140 in different hospitals. Since a large proportion of neonatal intensive care cots are taken up by RDS cases, then, again, significant savings are likely.

Finally, reduction in RDS and other morbidity will mean that babies leave hospital early, reducing the loading on special care baby units, with future unknown cost savings.

If use of corticosteroids rose by 25%, the savings in the UK would be about £1.5 million in surfactant and £2.6 million in neonatal care. Since the exact proportion presently being treated is not known, the impact on the heath service cannot be fully estimated. Working on the basis of half of all UK births being in teaching hospitals where uptake of steroid use is about 50%, and half in DGHs where it is about 35%, the overall cost benefit could be about £6 million, together with 400 babies' lives and reduced stress on services.

The Oxford GRiP Project

In Oxfordshire, the purchasing authority is requesting that the present rate of steroid use in women in preterm labour is to be assessed by audit.

Contracts with providers will stipulate that if the rate is less than 70%, then the rate should be increased to meet that target.

previous or next story in this issue