Intended for healthcare professionals


Caesarean section or vaginal birth for breech presentation at term

BMJ 1996; 312 doi: (Published 08 June 1996) Cite this as: BMJ 1996;312:1433
  1. M Hannah,
  2. W Hannah
  1. Associate professor and director Professor emeritus Maternal, Infant, and Reproductive Health Research Unit, Centre for Research in Women's Health, Toronto, Ontario, Canada M5G 1N8

    We need better evidence as to which is better

    About 3-4% of pregnant women reach term with a fetus in breech presentation, although an active policy of external cephalic version would be expected to lower this incidence somewhat. Over the past 20 years planned caesarean section has increasingly become the favoured approach for the delivery of these infants, despite the absence of good data to support this trend. Indeed, there have been only two randomised controlled trials that have compared the policy of planned caesarean section with that of planned vaginal birth.1 2 Although the trials were too small to have reasonable power to measure clinically important differences in serious neonatal morbidity or mortality, the studies did not find caesarean section to be a better option than planned vaginal delivery.

    A meta-analysis of these two trials, undertaken as part of a Cochrane systematic review, found no significant differences between the two approaches in terms of perinatal mortality (excluding malformations) (typical odds ratio 0.22 (95% confidence interval 0.00 to 14.52)) and Apgar score <7 at 5 minutes (0.64 (0.18 to 2.34)).3 Not unexpectedly, a policy of planned caesarean section was associated with higher rates of maternal morbidity (1.63 (1.03 to 2.57)). These results, which are the best evidence that we have available, do not support a policy of planned caesarean section.

    To explore further whether planned caesarean section might be better than planned vaginal birth for the delivery of the breech presentation at term, we undertook a systematic review of retrospective and prospective studies that compared these two policies.4 Although most of the studies included in the review were retrospective hospital audits, the results from our meta-analysis showed significantly lower rates of perinatal mortality and neonatal morbidity with planned caesarean section than with planned vaginal birth.

    In contrast, the retrospective population based study reported by Danielian et al in this issue of the BMJ—in which 1387 infants, with breech presentation at term, were followed to school age—did not find planned caesarean section to be associated with better long term infant outcomes (p 1451).5 The handicap rate of almost 20% in this large study is disturbing, and the observation that it is just as high in the cohort who underwent elective caesarean section raises once again the question whether underlying fetal abnormality, from whatever cause, may be associated with an increased incidence of breech presentation at term.

    Where lies the truth? Non-randomised studies are prone to selection bias, and thus differences in outcomes may just reflect differences in the women and the infants being studied in the two policy groups. In addition it is often not clear from these studies what the selection criteria or intrapartum management protocols for vaginal breech delivery were or whether they were appropriate. The published studies also give very little information about the skill and experience of the practitioners at delivery. The question of which is better for the mother and her infant—planned caesarean section or planned vaginal birth—therefore remains unanswered.

    If planned caesarean section should be better for the infant, information on the size of the benefit is also important if parents are to be expected to make an informed choice about which approach they would prefer, since increased maternal morbidity is an inevitable consequence of abdominal delivery. Care providers would also have to weigh other possible disadvantages of a policy of planned caesarean section. These include less training for vaginal breech delivery and thus fewer experienced practitioners available to deliver these infants. This could possibly increase the risk of adverse neonatal outcome for mothers presenting in advanced labour with a breech presentation, when caesarean section is not possible.

    We believe that an appropriately large randomised controlled trial is needed to determine whether planned caesarean section is better for the infant than planned vaginal delivery and, if better, to determine the size of the benefit. Indeed, we consider that the need is urgent, as a survey of Canadian obstetricians indicated that 69% of obstetricians thought that resident physicians were not acquiring the necessary skills to manage safely a trial of labour and conduct a vaginal delivery for a mother with a frank breech presentation at term.6 The most frequently cited reasons were lack of clinical volume and staff obstetricians' inexperience or reluctance to undertake vaginal breech deliveries. If such a trial is not conducted soon the art of vaginal breech delivery may well be lost, and planned caesarean section may become standard practice by default. Not only would this limit the childbirth alternatives for women, but the costs associated with such a policy would be substantial.

    In fact, plans for a large international randomised controlled trial, coordinated in Toronto, are in an advanced stage. So far, 130 centres in Canada, the United Kingdom, the United States, Australia, Israel, South Africa, and other countries have expressed interest in participating in the trial. Additional participating centres would be welcome. This will not be an easy trial to undertake, but we agree with Danielian et al that the question of which is the better management for women at term with a breech presentation will never be answered until such a trial is completed.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.