Editorials

How to manage term breech deliveries

Avoid vaginal breech deliveries but offer external cephalic version

At term 3-4% of all babies will present by the breech. The Term Breech Trial has recently clarified whether a vaginal breech delivery at term should be avoided.¹ Until now this argument has been muddied by the emotive debate about natural versus caesarean delivery, against a background of poor evidence to support or refute the safety of breech delivery for mother or baby.

The Term Breech Trial showed a significant increase in perinatal mortality and morbidity (3.4%) with planned vaginal delivery. As breech presentation is itself significantly associated with poor perinatal outcome, previous observational studies have been too seriously confounded to be able to inform clinical decisions. Until this trial there had been only two randomised controlled trials comparing planned caesarean section and vaginal breech delivery at term. ^{2 3} These studies were small (only 313 women) but suggested a worse outcome for the mother and a better outcome for the baby if caesarean section was planned. Some obstetricians routinely performed caesarean sections for breech at term, while others selected appropriate term breech babies for vaginal delivery.

The term breech trial provides unequivocal evidence that women with a breech presentation at term who plan a caesarean section will have a baby less likely to die or have a serious outcome (in the neonatal period) than those who plan a vaginal delivery (relative risk 0.33, 95% confidence interval 0.19 to 0.56). The results showed a 1% increased risk of perinatal death and a 2.4% increased risk of serious neonatal morbidity when a vaginal birth was planned.

This randomised controlled trial was carried out in 121 centres in 26 countries and involved 2088 women with a non-footling singleton breech presentation. Selection of cases was on the basis of size (<4000 g), no obvious contraindication to vaginal delivery (such as placenta praevia), and no identified anomaly in the fetus. An experienced obstetrician was available for delivery in each centre. Predefined labour management allowed induction and the use of syntocinon for normal obstetric indications; acceptable progress was up to 18 hours for the first stage, 2 hours before pushing, and 1.5 hours of pushing in the second stage. The results were similar whether labour was induced, augmented, or prolonged, and in women with different levels of attendants' experience.

A meta-analysis that has pooled these results with those from the two other randomised controlled trials shows that the benefit to the baby is similar (relative risk of death or morbidity 0.31, 0.19 to 0.52)⁴ because the estimates of effect are compatible in all three trials. In the term breech trial study serious morbidity (or death) in the mother was not increased significantly (relative risk 1.24, 0.79 to 1.95). However, the risk to the mother becomes significant in this meta-analysis: relative risk of maternal morbidity 1.29, 1.03 to 1.61.

There is therefore a definite cost in immediate maternal morbidity with planned caesarean section. No study has considered longer term outcome. Future morbidity has not been assessed beyond the index pregnancy and is particularly a concern in pregnancies with a scarred uterus. Longer term effects on the babies are also unknown, but this analysis is planned. In some settings the risk of caesarean section may still outweigh the risk of vaginal birth, and almost 97% of babies will not be seriously compromised as a result of planning a vaginal breech. The resource implications of performing more caesarean sections in some societies may also be significant and prohibitive. Also, the number needed to treat to show benefit is higher where perinatal mortality is high.¹

As caesarean sections are recognised to have an increased mortality and morbidity compared with vaginal delivery,⁵ clinicians must not be tempted to extrapolate these findings about term breech deliveries to other breech deliveries, such as twin pregnancies and premature deliveries (the commonest cause of breech presentation). The need to provide expertise in breech delivery will not disappear: the term breech trial showed that nearly 6% of women with breech presentation still have a vaginal breech delivery because they present too late, even with a policy of planned caesarean section. Moreover, some women will still choose a vaginal breech delivery even when evidence of harm is conclusive. Indeed, some women with HIV and even with fetal distress, where the benefits are even greater, refuse caesarean section. Reassuringly the level of experience in the obstetrician does not seem to be a factor in determining outcome, and this should not be used as an excuse to perform caesarean sections for other indications.

There is good evidence that external cephalic version for breech at term will reduce non-cephalic births by nearly 60%.⁶ However this technique is far from universally offered. Even in the term breech study, with enthusiastic participating units, nearly 80% of participants had not had an attempt at external cephalic version. There is now a pressing justification for implementing this simple, apparently safe alternative to planned caesarean section in all obstetric units and to offer it universally while continuing assessment of its safety and use, including in labour. A planned caesarean, though beneficial to the term breech fetus, increases maternal morbidity and should not be the first or only obstetric intervention.

Andrew Shennan, senior lecturer, maternal and fetal research unit. 
Susan Bewley, clinical director, women's health directorate. 

St Thomas's Hospital, London SE1 7EH

Footnotes

   SB was a fast track reviewer for the Lancet for the Term Breech Trial paper.¹