Vol. 16 No 4 | Summer 2014
Should inductions be pushed?
Prof Steve Robson

This article is 10 years old and may no longer reflect current clinical practice.

The latest clinical evidence suggests perhaps more labours should be induced.

The Australian National Maternity Services Plan1, published by the Australian Health Ministers’ Conference in 2011, provided a five-year vision: ‘All Australian women will have access to high-quality, evidence-based, culturally competent maternity care in a range of settings close to where they live.’ The document stresses that woman-centred maternity care is responsive to women’s needs and preferences, and enables them to access objective, evidence-based information that supports informed choices about their maternity care.

The National Maternity Service Plan offers plaudits – ‘[Australia] is now one of the safest countries in the world in which to give birth or to be born’ – but it also notes, ‘although the majority of Australian women have vaginal births, there appears to be a trend away from normal birth…Australia has high rates of birth by caesarean section compared with the OECD average.’

An issue singled out for particular criticism in the document is induction of labour (IOL): ‘…forms of intervention, including induction of labour, are also high: 25.3 per cent of mothers had an induced labour in 2007, while a further 20 per cent of all mothers had an augmented labour…The rise in interventions, including the reason and their impact on women, babies, and the health system, is the subject of considerable debate…’

So what’s all the fuss about? Labours are generally induced when the risks, to either mother or baby, associated with continuing a pregnancy seem to be greater than those of delivery. For the purposes of benchmarking, the Australian Council on Healthcare Standards (ACHS) previously used a number of ‘defined’ indications for induction in its clinical indicators: diabetes, premature rupture of membranes, hypertensive disorders (including chronic renal disease), fetal growth restriction, isoimmunisation, fetal distress, fetal death, infection and prolonged pregnancy. Such lists are, however, designed for large-scale benchmarking and are by necessity rather narrow and there are obviously many more reasonable indications for IOL. The UK National Institute for Health and Care Excellence (NICE) guidelines on IOL make the following observation:

Although a variety of specific clinical circumstances may indicate the need for induction of labour with a greater or lesser degree of urgency, the essential judgement that the clinician and the pregnant woman must make is whether the interest of the mother or the baby, or both, will be better served by ending or continuing the pregnancy. In making the judgement, it is necessary to factor in the attitude and wishes of the woman in response to her understanding of the actual risk of continuing the pregnancy, as well as the possible consequences of the method employed and the response to the induction of labour.2

Over the last 20 years in Australia, the proportion of labours that are induced has increased from 21.6 per cent in 1991, to 32.2 per cent in 2011 (OR 1.72, 95 per cent CI 1.70, 1.74. P < 0.05).3, 4 National data detailing the primary indications for IOL have only been published since birth year 20065, and although there have been statistically significant changes over the five years, it is possible that these represent changes in reporting rather than true shifts (see Table 1). It is also important to bear in mind that IOL is commonly undertaken for a combination of reasons rather than one single indication, and this type of decision-making is very difficult to capture in mandated jurisdictional data collections.

What does IOL achieve?

Going to the trouble of inducing labour suggests that it ought to yield some tangible benefit and in Australia at present almost one-third of women who attempt vaginal birth have their labours induced.4 That’s a lot of effort. Few would argue that perinatal conditions such as hypertensive disorders, diabetes, infection, and concerns about fetal growth and wellbeing (which together constitute more than one third of inductions4, 5) are unequivocal reasons for delivery. What about prolonged pregnancy? What about ‘social’ induction?

Many public hospital maternity services face great pressures in running their birth suites. Experienced midwifery staff are expensive to employ, difficult to roster equitably and safely, and indeed it can be difficult to actually find enough suitable staff. Hospitals also face challenges in providing a physical infrastructure with enough rooms and beds to meet demand. There is commonly an institutional ‘attitude’ against what are often called ‘soft’ indications for induction. All of these issues are important and need to be carefully considered. There would have to be very good reasons to consider adding even more inductions to already over-stretched maternity services.

For a generation of midwives and doctors, it was an article of faith that routine IOL in an ‘uncomplicated pregnancy’ any earlier than 41 weeks (and often longer) was of no benefit to the mother or baby. Patient information commonly includes statements such as:

Studies show that babies may be at an increased risk of stillbirth after 42 completed weeks … This risk is small for women with a healthy pregnancy and no other risk factors …induction of labour will be offered after you are 10 days overdue. You may choose not to have your labour induced…[if] you do not want to have an induction of labour, we recommend that you have an ultrasound scan to assess the well-being of your baby…There is every chance you will go into labour spontaneously prior to 42 weeks or prior to a booked induction…

Table 1. The commonest reasons for induction of labour in Australia. Data extracted from references 4 and 5.

Birth year
Indication for IOL 2006 2011 OR 95% CI P Value
Prolonged pregnancy 30.1% 23.3 0.70 0.69 0.72 <0.05
Prelabour ROM 10.4% 13.0 1.39 1.24 1.33 <0.05
Hypertension/PE 10.9% 10.0 0.89 0.86 0.93 <0.05
Suspected IUGR 4.0% 4.7 1.18 1.11 1.25 <0.05
Other* 27.8% 36.7 1.50 1.47 1.54 <0.05


Patient information such as this is written from a perspective of patient autonomy – that patients have the right to make decisions about their medical care without their ‘healthcare provider’ trying to influence their decisions. However, with rights come responsibilities and it is important that ‘healthcare providers’ actually provide accurate information.

Perhaps the most important thing that women and their families, and indeed maternity services in general, need to understand is that major differences exist between women of different ages. As stillbirth authority Dr Ruth Fretts expressed last year, in the pages of O&G Magazine:

Both maternal age and parity are significant risk factors in the rates of late pregnancy loss. In a large US study by Reddy et al, they found the most notable difference between younger and older women occurred after 38 weeks gestation. At 39 weeks, the risk of stillbirth for women 40 years of age or older was the equivalent to that of younger women (less than 34 years of age) who reached 41 weeks of gestation. This effect was modified by primiparity, with the risk of a stillbirth after 37 weeks of pregnancy with multiparous women young than 35 having the lowest risk 1.29/1000 per ongoing pregnancies, whereas the risk for a primiparous women 40 years of age or older was 8.65/1000 ongoing pregnancies (a 6.7 fold difference).6

Over the last two decades, the proportion of women aged 40 or more having babies in Australia has more than tripled (OR 3.0, 95 per cent CI 2.9, 3.1, p<0.05).3, 4 Indeed, there has been an enormous shift in age distribution of women having babies: the number of births to women aged 35 or more has increased by an additional 150 per cent (OR 2.4, 95 per cent CI 2.36, 2.44, P<0.05) since 1991.3, 4

The most recently published systematic review and meta-analysis7, reviewing randomised controlled trials in which IOL was compared with either placebo or expectant management among women with a viable singleton pregnancy, revealed a startling finding. More than 30 000 women had participated in more than 150 individual trials, and the risk of caesarean delivery was 12 per cent lower with labour induction than with expectant management, with a pooled relative risk of 0.88 (95 per cent CI 0.84, 0.93), a significant result at both term and post-term gestations. Furthermore, the authors found that the initial cervical score, indication for induction, and method of induction did not alter the main result. Their analysis also revealed a reduced risk of fetal death (RR 0.50, 95 per cent CI 0.25, 0.99) and admission to NICU (RR 0.86, 95 per cent CI 0.79, 0.94). The conclusion was that caesarean section rates were lower among women undergoing IOL than among those managed expectantly from term onwards, with additional benefits for the fetus. This completely turns the traditional paradigm on its head.

Two years before, Stock and colleagues had reported the results of a study of more than a million singleton pregnancies of 37 weeks or more from the UK.8 They compared the outcomes of ‘elective’ IOL (which they defined as IOL with no recognised medical indication) with those of expectant management, and compared outcomes such as perinatal mortality, mode of delivery, postpartum haemorrhage, obstetric anal sphincter injury, and admission to NICU. Adjustment was made for the women’s age at delivery, parity, year of birth, birthweight, deprivation category, and, where appropriate, mode of delivery. They found that, at each gestation between 37 and 41 completed weeks, elective IOL was associated with a decreased odds of perinatal mortality compared with expectant management, but no reduction in the chance of a spontaneous vertex delivery.

Where to from here?

The Australian National Maternity Plan1 takes as its linchpin the principle that, ‘maternity care should be evidence-based and woman-centred, and acknowledge pregnancy, birth and parenting as significant life events for women.’ It is completely unrealistic to dismiss the broader range of perceptions and expectations around IOL, pertaining to women, midwives, doctors and hospital administrators, as well as the community in general.

One of the main causes of ‘access block’ for many maternity services is that women recovering from caesarean section spend longer in hospital than women who have a vaginal birth. A perception that IOLs tie up staff and birthing rooms, and that ‘failed’ IOL is a major contributor to caesarean section rates will naturally cause anxiety to maternity staff and administrators. For this reason, many women and their families are given the impression that a request for IOL is selfish and risky, even if the impression is given inadvertently.

If it were true that a liberal approach to IOL increased the rate of caesarean section or led to worse outcomes for babies, then these would be completely legitimate concerns. Caesarean birth obviously influences the mode of delivery in the next pregnancy and may limit family size. Downstream, the risk of abnormal placentation increases after three caesarean sections. Babies who suffer from complications after birth incur higher costs for their care and admission to a NICU or SCBU may mean mothers spend more time as maternity inpatients, even if they are not delivered by caesarean section.

However, these fears appear to be unfounded. Indeed, the best evidence we have for the current demographic profile of women having babies in Australia, is that a low threshold for IOL at term actually reduces the risk of caesarean section quite substantially.7 Similarly, babies seem to be less likely to suffer ill effects, including the disaster of stillbirth.

Should more labours be induced?

As the National Maternity Plan1 emphasises, it is critical that women and their families be engaged in decision-making about their birth. There is a symbolism associated with birth that is rarely present in other areas of hospital activity, with the exception perhaps of palliative care. Maternity care providers commonly have strongly held views that have to be taken into account. Furthermore, birth care is expensive, with extensive resourcing required in terms of staffing and physical infrastructure. When things don’t go to plan, adverse maternity outcomes often expose services to embarrassing publicity and costs, and can have a severe effect on morale within a unit. For all of these reasons, it can be very difficult for clinicians and administrators – at both a local and a jurisdictional level – to steer a course with which everybody is happy.

When guidelines are being developed for clinical care within maternity services, it is critical that a balance be struck that will optimise outcomes for women and their babies. The available evidence supports a relatively liberal consideration of IOL at term, and it may be that the current policies in many units are actually contributing to the very problems that people are struggling to solve.


  1. Australian Health Ministers’ Conference. National Maternity Service Plan 2010. Commonwealth Government, Canberra, 2011 www. .
  2. National Institute of Health and Clinical Excellence. 2008. Induction of labour. NICE clinical guideline 70.
  3. Lancaster P, Huang J, Pedisich E & AIHW National Perinatal Statistics Unit. 1994. Australia’s mothers and babies 1991. Perinatal Statistics Series no. 1. Cat. no. AIHW 240. Canberra: AIHW. Viewed 2 October 2014,
  4. Li Z, Zeki R, Hilder L, Sullivan EA. Australia’s mothers and babies 2011. Perinatal statistics series no. 28. Cat. No. PER 59. 2013, Canberra: AIHW National Perinatal Epidemiology and Statistics Unit.
  5. Laws P, Hilder L. Australia’s mothers and babies 2006. Perinatal statistics series no. 22. Cat. No. PER 46. 2008, Canberra: AIHW National Perinatal Epidemiology and Statistics Unit.
  6. Fretts RC. The stillbirth scandal. O&G Magazine 2013; 15(4): 13-15.
  7. Mishanina E, Rogozinska E, Thatthi T, et al. Use of labour induction and risk of cesarean delivery: a systematic review and meta-analysis. CMAJ 2014.
  8. Stock SJ, Ferguson E, Duffy A, et al. Outcomes of elective induction of labour compared with expectant management: population based study. BMJ 2012; 344: e2838.

Leave a Reply

Your email address will not be published. Required fields are marked *