Recent guidance on antenatal corticosteroids in prematurityBMJ 2016; 352 doi: http://dx.doi.org/10.1136/bmj.i1655 (Published 23 March 2016) Cite this as: BMJ 2016;352:i1655
- Melissa K Whitworth, consultant obstetrician
- St Mary’s Hospital, Central Manchester University Hospitals NHS Foundation Trust, Manchester Academic Health Science Centre, Manchester M13 9WL
Preterm birth—birth before 37 weeks of pregnancy—is the leading cause of perinatal and neonatal mortality and morbidity. Globally, 15 million babies are born preterm each year; for England and Wales the figure is around 50 000, costing the UK economy almost £1bn a year.1
Maternal antenatal corticosteroids are one of the most effective obstetric interventions to reduce neonatal morbidity and mortality in premature babies. The most recent Cochrane review of this treatment, a meta-analysis of 21 randomised trials including 4269 women, reported that neonatal respiratory distress syndrome occurred in 17.3% of the group receiving corticosteroids compared with 26% of the control group and that neonatal mortality was 10.1% in the corticosteroid group compared with 14.8% in the control group.2 This and other benefits have led to a focus on ensuring that corticosteroids are administered promptly.
Last year the World Health Organization (WHO) and the UK National Institute for Health and Care Excellence (NICE) published guidelines on preterm labour and birth. Both documents recommend the administration of antenatal corticosteroids to women at risk of early delivery.
WHO recommends giving corticosteroids from 24 to 34 weeks’ gestation, while NICE recommends that clinicians consider steroids at 24-25 weeks but offer corticosteroids from 26 weeks to 33 weeks. Both agree the maximum benefit occurs from 24 hours to 7 days after initial administration.3 4 We know that compared with a single course, multiple courses of corticosteroids do not alter risk of death or disability at age 5 years.5 However, evidence suggests that multiple courses may adversely affect in utero growth, with possible harms such as obesity and increased cardiovascular risk manifesting in adult life.6
Timing of the administration of corticosteroids is crucial, with the key phrase being “at risk of preterm delivery.” WHO’s recommendations provide no guidance on how to determine whether a woman is at imminent risk of this. The implied assumption is that it is easy to diagnose. However, several studies have showed that obstetricians are poor at differentiating between actual and suspected preterm labour, with 30-80% of women who present with symptoms suggestive of preterm labour remaining pregnant 14 days later.7 8
Diagnosis of preterm labour was prioritised by NICE for health economic analysis, and the apparent reliance on these results to determine who should receive corticosteroids has been questioned. The complex economic modelling has resulted in a policy of treating all at less than 30 weeks’ gestation if clinical assessment alone suggests the woman is in preterm labour. At gestations of 30 or more weeks, NICE recommends transvaginal ultrasonography to measure cervical length. If that’s not available, cervicovaginal fetal fibronectin concentrations could be measured to determine likelihood of birth within 48 hours.
The treat all policy could greatly increase the transfer rates of women in threatened preterm labour, as doctors move women to units with the facilities to optimise neonatal care. Some evidence suggests that the negative predictive value of fetal fibronectin is such that it should be used at all gestations.9 Many units anecdotally report substantial reductions in rates of in utero transfer after introduction of fetal fibronectin testing before 30 weeks. The main reason that economic modelling favours a treat all strategy at less than 30 weeks’ gestation is the enormous extra costs associated with treating a preterm baby whose mother did not receive corticosteroids.
Clinicians often struggle with decisions based primarily on economics. However, within this academic and economic argument is a woman who may be in preterm labour. For the individual patient the decision is often reduced to: “Have corticosteroids and reduce the risk of your baby dying if you deliver within the next 24-48 hours, or wait and miss the opportunity for a lifesaving intervention.” At that moment, women are likely to hear only “reduce the risk of dying” and opt for corticosteroids rather than weighing up the likelihood of not delivering within 24-48 hours and the potential for harm from unnecessary treatment.
Disappointingly, the NICE guideline fails to recommend research to determine which interventions are most effective to predict or prevent preterm births (ranked as the number one research question on preterm birth by the James Lind Alliance10). If clinicians had an accurate test to diagnose preterm labour the risk of missing the narrow window of opportunity for timely administration of maternal corticosteroids would be reduced. In the meantime clinicians need to consider how they effectively communicate the complex risk-benefit analysis of administering corticosteroids to pregnant women in suspected preterm labour. In addition, they need to push for large trials of combinations of the current tests available to diagnose preterm labour.
Competing interests: I have read and understood BMJ policy on declaration of interests and declare I receive National Institute for Health Research funding for a pilot study of management options for women with a high risk of preterm labour and a short cervix.
Provenance and peer review: Commissioned; not externally peer reviewed.