Description of the health problem
There were 698,512 live births in England and Wales in 2013.1 More than one in five births followed labour induction; this represents > 150,000 pregnant women in England2 and Wales3 per year. There is evidence that the number of labour inductions has been steadily increasing over the past two decades. NHS England maternity statistics for 2010 noted that 21.3% of births followed induction of labour, and by 2012–13 this figure had increased to 23.3%.4
Induction of labour is carried out for a number of clinical indications.5,6 The most common reasons include post-term pregnancy (defined as 41+0 weeks’ gestation), prelabour rupture of the amniotic membranes (PROM) or when the well-being of the woman or baby may be compromised by prolonging the pregnancy (e.g. in cases of fetal growth restriction or pre-eclampsia).
There is a broad range of methods available for induction of labour. The choice of method may depend on national guidelines and local protocol, as well as individual clinical factors. The advantages and disadvantages of different methods vary, and the choice of method has implications for women and the UK NHS.
From a clinical perspective, the decision about which method to use for induction of labour can be influenced by the woman’s readiness for labour, for example whether or not membranes have ruptured spontaneously or whether or not the cervix remains undilated at the start of the induction process. Different methods used for inducing labour have different mechanisms of action, and vary in terms of how quickly birth is achieved and the likelihood of causing complications in women with different clinical characteristics. Thus, the choice of method will take into account the reason for induction and its urgency. The woman’s obstetric and medical history is also considered. For example, there is evidence that women may be more sensitive to drugs that stimulate the uterus if they have had a previous birth, and women who have a scar from a previous caesarean birth are at increased risk of uterine rupture, which can result in hysterectomy and fetal death.7
Different methods also have different direct costs, and some methods require continuous monitoring of the woman throughout labour. Consequently, the choice of induction method may have significant implications for NHS resources, especially if the method is known to increase the risk of complications requiring a caesarean section (CS).
Women may wish to experience a natural onset of labour, and there is evidence that an induced labour can have a negative impact on their overall experience of childbirth.8 Some methods of induction are painful or unpleasant, and some are associated with distressing side effects, such as headache or nausea. Women may also have preferences about which method is used and may prefer non-pharmacological approaches. On the other hand, women will want their baby to be born safely, and timely induction may improve outcomes for women and babies.5 Women facing decisions about induction of labour require up-to-date information about the range of options available, including alternative and complementary methods.
Description of available interventions and current service provision/policy
In the NHS context, choice of induction method is typically between prostaglandins and oxytocin combined with artificial rupture of membranes. UK clinical guidelines published in 20089 identified vaginal prostaglandin E2 (PGE2) as ‘the preferred method of induction’. We note that this recommendation was not based on a quantitative overview of the evidence of the effects and safety of all available methods, or from the synthesis and analysis of data from a range of comparisons. Furthermore, this guideline9 did not recommend any particular type (gel, tablet or pessary) or dose of PGE2 because trial evidence has rarely compared different PGE2 preparations. Potential updating of the current guidance is awaiting the publication of this report.10
Despite its importance, the question of resource use for the NHS has been relatively under-studied, and uncertainty remains about the costs that are associated with induction of labour. There is evidence that inducing labour in women with complications is associated with lower health-service costs than costs associated with expectant management.11–13 However, there is little evidence on the costs associated with specific methods of induction compared with others. Randomised trials in which one method of induction has been compared with another have only rarely included economic analyses.14
A broad range of pharmacological, mechanical, complementary and alternative methods have been used to induce labour. In the remaining sections of this chapter, we describe all of the pharmacological and mechanical methods for third-trimester induction of labour or cervical ripening which have been used in clinical practice and that have been examined in randomised trials. Complementary or alternative methods have been less commonly used in NHS settings but have been used in comparable settings in other countries. Complementary and alternative methods are included here, as information on the effects and safety of such methods may be important for women who prefer a less medicalised birth.
Pharmacological methods for the induction of labour
Prostaglandins: prostaglandin E2 and prostaglandin F2 alpha
Prostaglandins are hormones produced naturally by the body that are important in the onset of labour. Synthetically manufactured prostaglandins have been used in clinical practice since the 1960s to ripen the cervix and induce uterine contractions. They are more frequently used in women when the cervix is unripe (i.e. with a Bishop score < 6). Prostaglandins promote cervical ripening and encourage the onset of labour by acting on cervical collagen so as to encourage the cervix to soften and stretch in preparation for childbirth. Prostaglandins may also stimulate uterine contractions.
Despite the widespread use of prostaglandins as part of labour induction, they can cause a number of side effects, including nausea, vomiting, diarrhoea and fever. In addition, because of their effect on the uterus, prostaglandins can cause contractions that last too long, or are too frequent or are too strong. Excessive uterine activity, or hyperstimulation, may be associated with fetal distress, and in a small number of cases can lead to uterine rupture, especially in those women who have uterine scarring from surgery or a previous caesarean birth.
A large number of prostaglandin preparations have been available for labour induction, including prostaglandin F2 alpha (PGF2α, dinoprost), prostaglandin E2 (PGE2), prostaglandin E (PGE1) and misoprostol (a synthetic analogue of PGE1, which is described separately: see Misoprostol). In the past, PGF2α was frequently used in clinical practice but, more recently, PGE2 (dinoprostone) has become the most commonly used formulation. Commercially produced PGE2 analogues are expensive and require refrigeration. These factors have limited use in low-resource settings.
Prostaglandins are available in a variety of formulations and doses, and may be given via various routes of administration, including vaginally, intracervically, orally and, less frequently, intravenously.
Vaginal and intracervical administration
Prostaglandin preparations for vaginal and intracervical administration include gels, lactose-based vaginal tablets, suppositories, pessaries or inserts.15,16 Dosages of prostaglandins (mainly PGE2) vary, depending on route and local protocol (frequently 0.5 mg for intracervical use, 2–3 mg for intravaginal use and 10 mg for sustained-release pessaries). There is also variation in terms of the number of applications and time intervals between repeated doses. Sustained-release vaginal pessaries have been developed to reduce the number of applications and vaginal examinations that are needed during induction of labour. Vaginal and intracervical administration are the most common forms of administration in current practice.
In the meta-analysis we have treated different types of vaginal and intracervical PGE2 as different interventions as different preparations may vary in terms of rate of absorption, safety and cost. We have therefore included as separate interventions:
PGE2 vaginal tablets (lactose based).
PGE2 vaginal pessaries normal release (also sometimes referred to as suppositories), manufactured using various base materials, including wax and glycerine. [Note that this intervention includes a heterogeneous group of vaginal PGE2 preparations of varying composition. The base material used was not always clear, and pessaries were frequently produced in local pharmacies (i.e. not commercially available). We included this group of interventions in the network meta-analysis (NMA) and the cost analysis for completeness, even though they are not generally reproducible or available in the UK NHS.]
PGE2 vaginal pessaries sustained release (10- to 12-mg pessaries, single application).
PGE2 gel introduced via vaginal applicator.
PGE2 for intracervical administration.
Extra-amniotic administration
The administration of extra-amniotic prostaglandin gel was first carried out in the early 1970s. The gel is administered via a Foley catheter inserted through the cervix into the extra-amniotic space. The catheter is frequently left in place with the balloon inflated, and light traction may also be applied by taping the catheter to the woman’s leg. Extra-amniotic administration is no longer common in current practice.17
Intravenous administration
Intravenous (i.v.) prostaglandins are associated with increased rates of maternal vomiting and diarrhoea and are rarely used in current practice.18
Oral administration
Oral PGE2 and PGF2α have been available since the early 1970s. Oral administration is associated with gastrointestinal side effects and is seldom used nowadays.19
Misoprostol
Misoprostol is a PGE1 analogue that is known to be effective in stimulating uterine contractions. Misoprostol is inexpensive and requires no special storage facilities. Several routes of administration and regimens of misoprostol have been studied, including oral (swallowed as a tablet or dissolved in a titrated solution), vaginal (inserted into the vagina as a tablet or gel), rectal (inserted into the rectum as a tablet) and buccal or sublingual (the tablet is dissolved in the cheek or under the tongue, respectively).20–22 Different routes of administration have advantages and disadvantages. Oral misoprostol achieves rapid onset of action, whereas vaginal administration is associated with slower absorption but more prolonged action. Over the past decade, slow-release misoprostol vaginal pessaries have also been tested in trials.
Although misoprostol is widely used in obstetric practice for other indications (e.g. abortion), there have been concerns about its use due to the increased risk of serious adverse effects, such as uterine rupture. Several small studies have reported excessive uterine activity that is associated with the use of misoprostol, such as uterine tachysystole (more than five contractions per 10 minutes for at least 20 minutes), uterine hypersystole/hypertonus (a contraction lasting ≥ 2 minutes) and/or uterine hyperstimulation syndrome [uterine tachysystole or hypersystole with fetal heart rate (FHR) changes such as persistent decelerations]. A meta-analysis examining the use of vaginal misoprostol suggested that despite excess uterine activity, misoprostol was not associated with adverse fetal outcomes especially at a lower dose (< 25 µg).23
Oxytocin
Oxytocin is a hormone that is produced naturally by the body, and which has a range of functions, including the stimulation of uterine contractions in the second and third stages of labour. Oxytocin analogues, administered intravenously, are the commonest induction agents used worldwide. Oxytocin is frequently administered when the cervix is dilated (or favourable) and may be combined with artificial rupture of the amniotic membranes (amniotomy). Oxytocin may cause excess uterine activity, especially in settings where equipment is not available to titrate doses accurately and monitor contractions.
Current i.v. oxytocin regimens usually involve incremental increases in dosage. Lower-dose regimens typically involve 0.5–2.0 milliunits (mU)/minute starting doses, with incremental increases of 1.0–2.0 mU/minute every 15–60 minutes. Higher-dose regimens have starting doses up to 6.0 mU/minute, with incremental increases of 2.0–6.0 mU/minute every 15–40 minutes. There are advantages and disadvantages of high- or low-dose regimens; higher doses may lead to a shorter period to delivery, but may increase the risk of hyperstimulation, whereas lower doses may increase risk of infection if labour is prolonged.24–27
Nitric oxide donors
Nitric oxide (NO) is thought to be involved in cervical ripening, and in recent years NO donors [isosorbide mononitrate (ISMN), isosorbide dinitrate, nitroglycerin and sodium nitroprusside] have been used to promote cervical ripening. NO is administered as a vaginal tablet.28
Mifepristone
Mifepristone is a progesterone antagonist that has been used in the past in combination with prostaglandins in first trimester and early second trimester pregnancy terminations. Mifepristone has been proposed as a method to induce labour because it acts to increase uterine contractions. Mifepristone is administered as an oral tablet.29
Oestrogens, corticosteroids, relaxin and hyaluronidase
Oestrogens have a role in promoting cervical ripening and, historically, have been administered intravenously or into the extra-amniotic space. There are no commercially available preparations for use in cervical ripening or induction of labour, and the two included trials of this agent date back to 196730 and 1981.31
The role of corticosteroids in the process of labour is not well understood, and they are currently not used in clinical practice for the induction of labour.32
Relaxin is a hormone that is thought to encourage cervical ripening, which has been tested in a very small number of trials.33 Similarly, hyaluronidase is also thought to be implicated in cervical ripening.34 Both agents have been administered in vaginal or intracervical gel, but neither is common in current practice.
Mechanical and physical methods for induction of labour
Mechanical methods to induce labour have been available for many years. Mechanical devices include various types of catheters and laminaria tents, introduced into or through the cervix and into the extra-amniotic space. The introduction of devices into the cervix may cause the cervix to dilate. Their presence may also increase prostaglandin or oxytocin secretion, which, in turn, may increase cervical dilatation and stimulate uterine contractions.35 Here we also include descriptions of membrane sweep and amniotomy since they may be considered a physical method of inducing labour.
Catheters
Foley urinary catheters have been used for the induction of labour, as have double-balloon and other catheters that are specifically designed for use in induction of labour (e.g. Cook catheter). The catheter is introduced into the extra-amniotic space, and then the balloon(s) is (are) inflated to keep the catheter in place. Traction may be applied by taping the catheter to the woman’s leg. Catheters are usually left in situ until they are expelled. In some cases a saline infusion is introduced into the extra-amniotic space via the catheter.
Laminaria tents
Laminaria tents are made from sterile seaweed or synthetic materials. These devices are introduced into the cervical canal and expand to gradually stretch the cervix.
Membrane sweep
Stripping or sweeping of the membranes has been used for many years to induce labour, and continues to be carried out in many clinical settings. Membrane sweeping involves the clinician detaching the membranes from the lower uterine segment by a circular movement of the examining finger. Membrane sweeping is thought to lead to an increased production of prostaglandins. When the cervix is closed, a cervical massage may be carried out instead of a membrane sweep to stimulate the production of prostaglandins.36
Amniotomy
During labour the amniotic membranes usually rupture spontaneously as the cervix dilates and stretches in preparation for the descent of the fetus. Amniotomy refers to rupture of the membranes using a plastic hooked instrument or, occasionally, surgical forceps.
Amniotomy may be carried out alone or in combination with oxytocin or prostaglandins to induce labour. It can be carried out only if the amniotic membranes are accessible to the midwife or doctor, and this may not happen until the cervix has started to dilate.
Amniotomy may cause some potentially serious adverse effects, including cord prolapse. The procedure may introduce infection. For women known to be human immunodeficiency virus (HIV) positive the procedure is avoided because it may increase the risk of mother-to-child transmission of HIV.25
Breast stimulation
Manual breast stimulation has been used in the past to stimulate uterine contractions.37 It is thought that it may trigger the release of oxytocin.
Sexual intercourse
Sexual intercourse at term has been thought to lead to the onset of labour.38 The hypothesised mechanism of action here is the prostaglandin contained within semen.
Complementary and alternative methods for induction of labour
Castor oil
Castor oil is derived from the bean of the castor plant, and has been used in oral form as a method of stimulating labour.39 Castor oil has laxative properties, stimulating the intestines and bowel. It is this stimulation that is hypothesised to initiate uterine contractions and labour as a secondary effect.
Acupuncture
Acupuncture involves the insertion of fine needles by trained staff into the skin at specified points on the body. Stimulation of particular acupuncture points is intended to initiate uterine contractions and labour.40
Homeopathy
Homoeopathy involves the use of highly diluted solutions that contain tiny amounts of the original substance. Homeopathic preparations are popular and are available over the counter in pharmacies and health food shops. Some homeopathic preparations have been recommended to promote the onset of labour.41
Overall aims and objectives of assessment
Given the broad range of methods used to induce labour, the main research question addressed by this review is ‘what is the best method for induction of labour?‘. The specific objectives were to:
assess the effectiveness and safety of a range of induction methods to determine which method or methods achieves the best outcomes
provide a quantitative summary of the evidence on the relative effects of a broad range of induction methods to identify which method works best
develop a decision model to evaluate the cost-effectiveness of the different methods for induction
explore, if sufficient evidence is available, the effect of different clinical subgroups [with intact or ruptured membranes, at different gestational ages, in women following a previous CS and with low (< 6) or higher Bishop scores] on effectiveness and cost-effectiveness.
Specification of the PICO research question
Population Pregnant women carrying a viable fetus and who are eligible for any method of third-trimester cervical ripening or labour induction.
Intervention and relevant comparators No treatment, placebo, all pharmacological (all routes and doses), mechanical and complementary methods used for the induction of labour.
Outcomes Our primary effectiveness outcome was (1) vaginal delivery (VD) not achieved within 24 hours, and our primary measures of safety were (2) uterine hyperstimulation with FHR changes and (3) CS. Our secondary outcomes for serious adverse events were (4) serious neonatal morbidity or perinatal death and (5) serious maternal morbidity or death. Other outcomes included were (6) maternal satisfaction with the induction method used, and, for use in the economic model, (7) cost, resource use and utilities.
Definition of the decision problem for the economic evaluation
Our aim was to answer the following question: what is the most cost-effective method (from the interventions described above), for third-trimester cervical ripening or labour induction? Outputs from the economic evaluation include expected costs, expected benefits, incremental cost-effectiveness ratios (ICERs), expected net benefit and cost-effectiveness acceptability curves (CEACs).
Stakeholder involvement in project
The steering group (listed in Appendix 1) and project team included a consumer representative, a health economist, a midwife and an obstetrician engaged in clinical practice.
A consumer representative was included as a collaborator on the project, and she contributed to the early discussions on this project and drafting the application. Induction of labour is known to be of great interest to pregnant women. In particular, women are interested in self-administered ways of initiating labour and for this reason these methods were examined in the proposed work. The consumer representative co-ordinated the involvement of members of the CPCG (Cochrane Pregnancy and Childbirth Group) consumer panel, National Childbirth Trust and the Association for Improvements in Maternity Services (AIMS) who expressed an interest in participating. Members of these groups were asked for comments to inform steering group meetings, to determine the final outcomes, to aid in the interpretation of the findings and to shape the papers to be published. The authors of this report include a consumer representative (GG).
The steering group commented on the study design, selection of outcomes, methods for the cost-effectiveness analysis and dissemination strategies.
Overview of report
In Chapter 2 we describe the methods used for the assessment of clinical effectiveness, including the methods for the systematic review to identify relevant evidence on clinical effectiveness, and the methods for the NMA. In Chapter 3 we present the results from the systematic review and NMA, including the relative effectiveness of interventions that have been used to induce labour in women at or near term. In Chapter 4 we describe methods and present results of the cost-effectiveness analysis, taking a UK NHS perspective. In Chapter 5 we summarise findings, set out the strengths and limitations of our approach, consider the implications of our results on recommended practice, and indicate areas for which future research would be beneficial.
