Home > Clinical Guides > Induction of Labour > Methods of induction of labour

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Induction of Labour. London: RCOG Press; 2008 Jul. (NICE Clinical Guidelines, No. 70.)

5Methods of induction of labour

5.1. Pharmacological-based methods

Clinical question

Prostaglandins (PGE2)

Prostaglandins are capable of stimulating uterine contractions resulting in labour. Prostaglandins can be administered by various routes: vaginal, oral, intravenous, extra-amniotic and intracervical.

5.1.1. Vaginal PGE2

Overview of available evidence

One systematic review and several additional RCTs were identified.

One systematic review (57 RCTs involving 10 039 women) compared the effects of prostaglandin gel (PGE2, 2–5 mg) versus placebo/no treatment (35 RCTs); versus PGE2 tablet (five RCTs); versus PGE2 pessary/suppository (two RCTs); PGE2 tablet versus PGE2 pessary/suppository (three RCTs); PGE2 (slow release) versus PGE2 (any vehicle) (seven RCTs); PGE2 low dose versus PGE2 high dose (seven RCTs); PGF versus placebo (three RCTs) and PGF versus PGE2 (two RCTs).109 [EL = 1++]

As PGF is associated with unpleasant gastrointestinal effects, and intracervical PGE2 was considered to be too invasive, only studies comparing different preparations of vaginal PGE2 were considered by the GDG.

The vaginal preparations of PGE2 in these trials varied and dosage of PGE2 was presented as described in the trials.

In women with an unfavourable cervix, compared with placebo/no treatment, all regimens of vaginal PGE2 are significantly associated with uterine hyperstimulation with fetal heart rate (FHR) changes (RR 4.47, 95% CI 2.01 to 9.93; 12 RCTs, 1143 women), improved cervical status within 24 hours (RR 1.45, 95% CI 1.16 to 1.86; two RCTs, 172 women), reduction in the need for oxytocin augmentation (RR 0.72, 95% CI 0.61 to 0.85; eight RCTs, 813 women) and reduced incidence of meconium-stained liquor (RR 0.65, 95% CI 0.47 to 0.89; five RCTs, 697 women). The rates of vaginal birth not achieved within 24 hours (no data available on birth within 36 or 48 hours) (RR 0.88, 95% CI 0.67 to 1.15; one RCT, 39 women), caesarean section (RR 0.87, 95% CI 0.75 to 1.02; 22 RCTs, 2173 women), postpartum haemorrhage (RR 0.99, 95% CI 0.47 to 2.05; seven RCTs, 917 women) and maternal side effects (RR 0.97, 95% CI 0.62 to 1.51; seven RCTs, 871 women) were comparable between the two groups. There was no perinatal mortality.

Comparisons between PGE2 gel (2 mg) and PGE2 tablets (3 mg) did not show any significant differences in vaginal birth not achieved within 24 hours (RR 1.28, 95% CI 0.87 to 1.87; one RCT, 73 women), uterine hyperstimulation with FHR changes (RR 2.00, 95% CI 0.18 to 21.71; one RCT, 200 women), caesarean section (RR 0.93, 95% CI 0.63 to 1.38; three RCTs, 352 women) or oxytocin augmentation (RR 0.85, 95% CI 0.71 to 1.02; four RCTs, 377 women). Comparisons between PGE2 gel (2.5–5 mg) and PGE2 suppositories (3.5–5 mg) found that uterine hyperstimulation with FHR changes was significantly less likely to occur with PGE2 gel (RR 0.16, 95% CI 0.03 to 0.87; two RCTs, 159 women); there were no data on oxytocin augmentation. Comparisons between PGE2 tablets (3 mg) and PGE2 suppositories (0.75 mg × 4 (3 mg)) suggested that oxytocin augmentation was significantly less likely to be required with PGE2 tablets (RR 0.35, 95% CI 0.19 to 0.64; one RCT, 200 women); there were no data on uterine hyperstimulation with FHR changes.

For all women, the difference in oxytocin augmentation between controlled release PGE2 pessaries (10 mg) and vaginal PGE2 gel (1–2.5 mg) was not significant (RR 0.83, 95% CI 0.65 to 1.06; three RCTs, 361 women). For women with an unfavourable cervix, oxytocin augmentation was significantly less likely to be required with the pessaries (RR 0.55, 95% CI 0.35 to 0.88; two RCTs, 161 women), but there was considerable heterogeneity in these studies and the regimens of oxytocin augmentation protocols were unclear. One additional RCT110 [EL = 1−] not included in this review reported comparable maternal and fetal outcomes between these two methods.

Compared with high-dose PGE2 (3.5–10 mg), uterine hyperstimulation with FHR changes was significantly less likely to occur with the use of low-dose PGE2 (1–2.5 mg) (RR 0.18, 95% CI 0.03 to 0.99; two RCTs, 140 women).109 [EL = 1++]

In women with a favourable cervix, all regimens of vaginal PGE2 significantly reduced the likelihood of vaginal birth not being achieved within 24 hours when compared with placebo/no treatment (RR 0.12, 95% CI 0.08 to 0.17; one RCT, 345 women). Comparisons between PGE2 gel (2 mg) and PGE2 tablet (3 mg) (one RCT) found similar maternal and fetal outcomes.109 [EL = 1++]

Cost of vaginal PGE2

No published study was identified that examined the cost-effectiveness of controlled release pessary compared with vaginal tablets or gel and no studies were identified comparing vaginal tablets with vaginal gel. The drug cost of the controlled release pessary tablet is similar to either the tablet or the gel − £30 per pessary compared with £26.56 for either tablet or gel (assuming two doses per induction).111

Previous guidelines on induction of labour112 concluded that vaginal tablets should be recommended rather than vaginal gel on the grounds that, given equal efficacy, the tablets were less costly and therefore more likely to be cost-effective. A large increase in the price of the vaginal tablets and a small decrease in the price of the vaginal gel in September 2007 has annulled any difference in drug costs between these two options. 111

A simple cost analysis suggested that the costs of vaginal tablets, vaginal gel and the controlled release pessary tablet are broadly comparable at 2007 prices. While the drug cost for the controlled release pessary is higher, there may be some offsetting ‘downstream’ cost savings as a result of reduced oxytocin augmentation109 and a reduced need for vaginal examination. However, the magnitude of any such downstream saving is uncertain. The vaginal tablet and vaginal gel are likely to be relatively more cost-effective in women with a favourable cervix as a result of both lower drug and downstream costs. The cost analysis is described in more detail in Appendix C.

Evidence statements

Evidence from reasonably sized trials suggested that, in women with an unfavourable cervix, all regimens of vaginal PGE2 are effective in improving cervical status and reducing oxytocin augmentation and meconium staining, when compared with placebo or no treatment. However, one very small trial reported no difference between vaginal PGE2 and placebo in achieving vaginal birth within 24 hours. All regimens of vaginal PGE2 are associated with increased uterine hyperstimulation. [EL = 1++]

In women with an unfavourable cervix, PGE2 gel (2 mg) and PGE2 tablets (3 mg) result in comparable maternal and fetal outcomes. Uterine hyperstimulation with FHR changes is less likely with the use of PGE2 gel (2.5–5 mg) when compared with PGE2 suppositories (3–5 mg). Maternal and fetal outcomes are comparable between controlled release PGE2 pessaries and PGE2 gel. The need for oxytocin augmentation between controlled release PGE2 pessaries and PGE2 gel could not be determined owing to the heterogeneity of the studies included. Compared with PGE2 high dose, PGE2 low dose is associated with a reduced likelihood of uterine hyperstimulation with FHR. [EL = 1++] When compared with intravenous oxytocin and amniotomy, vaginal PGE2 is less likely to be associated with postpartum haemorrhage (see Section 5.1.7). [EL = 1++]

In women with a favourable cervix, all regimens of vaginal PGE2 are more effective than placebo/no treatment in achieving vaginal birth within 24 hours. One small RCT found that vaginal PGE2 gel (2 mg) and PGE2 tablet (3 mg) are comparable in the need for oxytocin augmentation. [EL = 1++] Comparison with intravenous oxytocin and amniotomy reported similar maternal and fetal outcomes (see Section 5.1.7). [EL = 1++]

Cost of vaginal tablet and gel

The drug cost of vaginal PGE2 tablets, gel and slow-release pessaries are similar at 2007 prices (£26.56 versus £26.56 and £30, respectively) but slow-release pessaries may be cheaper overall as a result of reduced rates of oxytocin augmentation and vaginal examination.

Interpretation of evidence

In women with an unfavourable cervix, the GDG recognised that the evidence base for vaginal PGE2 in the primary outcome of achieving vaginal birth within 24 hours is limited to one very small RCT, which reported no difference when compared with placebo. However, there were a number of RCTs with larger samples, which showed that all regimens of vaginal PGE2 were significantly more effective than placebo in improving cervical status and reducing oxytocin augmentation and meconium staining. In women with a favourable cervix, vaginal PGE2 is more effective than placebo in achieving vaginal birth within 24 hours. The risk of uterine hyperstimulation is significantly associated with the use of all regimens of vaginal PGE2.

The evidence comparing vaginal PGE2 with amniotomy plus intravenous oxytocin is based on small RCTs, which found comparable maternal and fetal outcomes between the two groups in women with an unfavourable cervix. In women with a favourable cervix, there was a risk of postpartum haemorrhage associated with the use of amniotomy plus intravenous oxytocin (see Section 5.1.7).

On balance, the GDG considered the evidence for the use of vaginal PGE2 in women with an unfavourable cervix to be persuasive, as vaginal PGE2 was more effective than placebo in a number of secondary outcomes. In this group of women, controlled release PGE2 pessary may be more appropriate because the induction time may be prolonged and it is more likely that repeated use of tablet or gel will be required. In women with a favourable cervix, there is robust evidence that vaginal PGE2 is an effective induction agent, and vaginal PGE2 tablet or gel may be more appropriate. In addition, the GDG also considered comfort, convenience and acceptability to be important to women (vaginal PGE2 is less invasive than amniotomy, and oxytocin requires intravenous access and continuous EFM, thus reducing women’s mobility during induction) and the balance of evidence supported the GDG’s view that vaginal PGE2 should be the preferred method of induction of labour irrespective of cervical status.

The optimal frequency of use and the maximum dose are not clear from the evidence. The GDG considered that vaginal PGE2 products should be used in accordance with the manufacturers’ instructions.

The PGE2 tablet and gel drug costs are slightly cheaper (assuming two doses per induction) than controlled release pessaries at 2007 prices. Overall costs are broadly comparable as there may be ‘downstream’ savings with controlled release pessaries as a result of reduced oxytocin augmentation and vaginal examination. These downstream savings relate to the number of doses of PGE2 tablet/gel that are required to initiate labour. Therefore, the relative cost-effectiveness of the PGE2 tablet and gel is likely to be greater in women with a favourable cervix.

Recommendations on vaginal PGE2

Vaginal PGE2 is the preferred method of induction of labour, unless there are specific clinical reasons for not using it (in particular, the risk of uterine hyperstimulation). It should be administered as a gel, tablet or controlled release pessary. Costs may vary over time and trusts/units should take this into consideration when prescribing PGE2. For doses, refer to the SPCs. The recommended regimens are:

  • one cycle of vaginal PGE2 tablets or gel: one dose, followed by a second dose after 6 hours if labour is not established (up to a maximum of two doses)
  • one cycle of vaginal PGE2 controlled release pessary: one dose over 24 hours.

When offering PGE2 for induction of labour, healthcare professionals should inform women about the associated risks of uterine hyperstimulation.

Research recommendation on vaginal PGE2

Research is needed to assess the effectiveness, safety, maternal satisfaction and acceptability of different regimens of vaginal PGE2, stratified by clinical indications, cervical and membrane status, parity and previous caesarean section.

Research question

What are the effectiveness, safety and maternal acceptability of:

Why is this important?

Despite extensive studies carried out over the past 30 years to determine the most effective ways of inducing labour with vaginal PGE2, uncertainties remain about how best to apply these agents in terms of their dosage and timing. It would be particularly useful to understand more clearly why vaginal PGE2 fails to induce labour in some women.

5.1.2. Oral PGE2

Overview of available evidence

One systematic review was identified.

One systematic review (19 RCTs involving 2688 women, Bishop score ≤ 3 to 7) assessed the effects of oral PGE2 versus no treatment or placebo (three RCTs); versus vaginal PGE2 (three RCTs); versus cervical PGE2 (two RCTs); versus intravenous oxytocin (seven RCTs); versus intravenous oxytocin plus amniotomy (four RCTs); versus oral oxytocin (four RCTs); versus oral oxytocin plus amniotomy (two RCTs); and oral PGE2 with incremental doses or high dose versus oral PGE2 constant or low dose (two RCTs). All maternal and fetal outcomes were similar between women undergoing induction of labour with oral PGE2 and the modalities described above. However, nausea and vomiting are significantly more likely to be reported in the oral PGE2 groups.

For women with an unfavourable cervix, caesarean birth was significantly less likely with oral PGE2 than placebo (RR 0.54, 95% CI 0.29 to 0.98; three RCTs)

For women with a favourable cervix, the available evidence suggested no significant differences in maternal or fetal outcomes in the comparisons between oral PGE2 and oral oxytocin and oral oxytocin plus amniotomy.113 [EL = 1++]

Evidence statements

Evidence suggested that, for women with an unfavourable cervix, oral PGE2 is associated with a reduction in caesarean birth rate when compared with placebo. However, oral PGE2 is no more effective as a cervical priming method than vaginal/intracervical PGE2, or oral/intravenous oxytocin. For women with a favourable cervix, oral PGE2 achieved similar maternal and fetal outcomes to oral oxytocin or oral oxytocin plus amniotomy. Gastrointestinal side effects including vomiting were frequently reported by women treated with oral PGE2. [EL = 1++]

Interpretation of evidence

For women with an unfavourable and favourable cervix, oral prostaglandins do not appear to offer any benefit over other routes of prostaglandin administration or intravenous oxytocin in women requiring cervical priming and induction of labour. There is a higher incidence of gastrointestinal side effects.

Recommendation on oral PGE2

Oral PGE2 should not be used for induction of labour.

5.1.3. Intravenous PGE2

Overview of available evidence

One systematic review was identified.

One systematic review (13 RCTs involving 1165 women, mixed Bishop score) compared the effects of intravenous prostaglandins (PGE2, 1–6.7 micrograms/minute; PGF, 6–40 micrograms/minute) versus intravenous oxytocin (four RCTs); versus extra-amniotic prostaglandin infusion (one RCT); and intravenous PGF versus intravenous oxytocin (eight RCTs). Overall, the use of intravenous prostaglandins was associated with higher rates of uterine hyperstimulation both with changes in the FHR (RR 6.76, 95% CI 1.23 to 37.11) and without changes in the FHR (RR 4.25, 95% CI 1.48 to 12.24) compared with oxytocin. There were significantly more maternal side effects (such as gastrointestinal side effects, thrombophlebitis and pyrexia) with the use of intravenous prostaglandins than oxytocin (RR 3.75, 95% CI 2.46 to 5.70). Trials comparing combinations of oxytocin/PGF and oxytocin or extra-amniotic PGE2 did not report any significant differences in maternal or fetal outcomes. In women with an unfavourable cervix, there was no significant difference between intravenous PGE2 and intravenous oxytocin in maternal outcomes. No fetal outcomes were reported in this group of women. There were very limited data available for women with favourable cervix.114 [EL = 1++]

Evidence statements

Evidence suggested that, overall, intravenous prostaglandin is associated with uterine hyperstimulation and gastrointestinal side effects when compared with intravenous oxytocin. For women with an unfavourable cervix, intravenous prostaglandins and intravenous oxytocin, used for induction of labour, appear to achieve similar maternal outcomes. There are very limited data available for women with a favourable cervix. [EL = 1++]

Interpretation of evidence

The use of intravenous prostaglandins is associated with significant uterine hyperstimulation with and without FHR changes, and with maternal complications such as thrombophlebitis, pyrexia and gastrointestinal side effects. Intravenous prostaglandin is no more likely to result in vaginal birth than oxytocin.

Recommendation on intravenous PGE2

Intravenous PGE2 should not be used for induction of labour.

5.1.4. Extra-amniotic PGE2

Overview of available evidence

One systematic review was identified.

One systematic review (10 RCTs involving 920 women, mixed parity and Bishop score) compared the effects of extra-amniotic PGE2 (250–500 micrograms) versus extra-amniotic placebo (three RCTs); versus vaginal PGE2 (four RCTs); versus intravenous oxytocin (one RCT); and extra-amniotic PGF versus extra-amniotic placebo gel (one RCT); and versus mechanical method (one RCT). For women with an unfavourable cervix, oxytocin augmentation was significantly less likely to be required with extra-amniotic prostaglandins when compared with placebo (RR 0.50, 95% CI 0.38 to 0.66; three RCTs). Comparisons with vaginal PGE2 found no significant difference in caesarean birth rates (RR 0.89, 95% CI 0.42 to 1.89; three RCTs). There were no other significant differences in maternal or fetal outcomes when compared with other methods. However, the small sample size of the studies included made interpretation difficult. In women with a favourable cervix, the likelihood of achieving vaginal birth was similar for extra-amniotic PGE2 and vaginal PGE2.115 [EL = 1++]

Evidence statements

Evidence suggested that, for women with an unfavourable cervix, extra-amniotic prostaglandins lessen the requirement for oxytocin augmentation when compared with placebo. There are insufficient data to determine its effectiveness when compared with intravenous oxytocin and mechanical methods. For women with a favourable cervix, extra-amniotic PGE2 is comparable to vaginal PGE2 in achieving vaginal birth within 24 hours. [EL = 1++]

Interpretation of evidence

Evidence is not clear whether the placebo comparison is similar to an extra-amniotic catheter without drug, which can mimic the effects of ‘cervical priming’. Outcomes such as caesarean birth rates are comparable to vaginal PGE2. Extra-amniotic PGE2 is no more effective than vaginal PGE2 and is a more invasive procedure.

Recommendation on extra-amniotic PGE2

Extra-amniotic PGE2 should not be used for induction of labour.

5.1.5. Intracervical PGE2

Overview of available evidence

One systematic review was identified.

One systematic review (56 RCTs involving 7738 women) assessed the effects of intracervical PGE2 (mixed parity, mixed Bishop scores) versus placebo/no treatment (28 RCTs); versus vaginal PGE2 (29 RCTs); and of different doses of intracervical PGE2 (two RCTs). In women with an unfavourable cervix, intracervical PGE2 was significantly associated with vaginal birth within 24 hours (RR 2.00, 95% CI 1.39 to 2.87; four RCTs) and no difference in caesarean birth (RR 0.88, 95% CI 0.77 to 1.01; 27 RCTs) when compared with placebo/no treatment. Intracervical PGE2 was significantly more likely not to achieve vaginal birth within 24 hours (RR 1.26, 95% CI 1.12 to 1.42; ten RCTs) when compared with vaginal PGE2. In women with a favourable cervix, no significant differences were found between intracervical PGE2 and vaginal PGE2 in caesarean and instrumental vaginal birth rates.116 [EL = 1++]

Evidence statements

Evidence suggested that, in women with an unfavourable cervix, intracervical PGE2 is more effective than placebo as an induction agent. Intracervical PGE2 is less effective than vaginal PGE2 in achieving vaginal birth within 24 hours. In women with a favourable cervix, maternal and fetal outcomes are comparable between intracervical and vaginal PGE2. [EL = 1++]

Interpretation of evidence

For women with an unfavourable cervix, intracervical PGE2 is less effective than vaginal PGE2 and confers no benefit. For women with a favourable cervix, it achieves similar maternal outcomes as vaginal PGE2. Intracervical administration is invasive. Intracervical PGE2 is not commonly used in the UK.

Recommendation on intracervical PGE2

Intracervical PGE2 should not be used for induction of labour.

5.1.6. Intravenous oxytocin alone

Oxytocin has been used alone, in combination with amniotomy, or following cervical ripening with other pharmacological or non-pharmacological methods. However, it is important to distinguish its role as an induction agent, i.e. to initiate labour, from its very frequent use in the augmentation of labour.

Overview of available evidence

One systematic review was identified. Several additional RCTs comparing different combinations of intravenous oxytocin with other methods were excluded as they were not considered appropriate by the GDG.

One systematic review (58 RCTs involving 11 129 women, mixed parity and Bishop score) evaluated the effects of intravenous oxytocin alone versus expectant management (26 RCTs); versus vaginal PGE2 (27 RCTs); and versus intracervical PGE2 (13 RCTs).117 [EL = 1++]

For this clinical question, the GDG considered the comparisons between intravenous oxytocin alone and vaginal PGE2 to be appropriate and relevant.

Studies of women with an unfavourable cervix and intact membranes reported that intravenous oxytocin alone was significantly associated with an unchanged cervical status after 12–24 hours (RR 2.67, 95% CI 1.21 to 5.88; one RCT) and an increased caesarean birth rate (RR 2.08, 95% CI 1.14 to 3.81; three RCTs) when compared with vaginal PGE2. In women with ruptured membranes, women given intravenous oxytocin alone were significantly less likely to give birth vaginally within 24 hours when compared with vaginal PGE2 (RR 1.70, 95% CI 1.29 to 2.25; three RCTs).117 [EL = 1++]

In women with a favourable cervix, vaginal birth was significantly less likely to be achieved within 24 hours when compared with vaginal PGE2 (RR 1.50, 95% CI 1.08 to 2.09; one RCT). Other maternal and fetal outcomes were similar.117 [EL = 1++]

Evidence statements

Evidence suggested that, in women with an unfavourable cervix and intact membranes, intravenous oxytocin alone is less effective than vaginal PGE2 in improving cervical status and in reducing the caesarean birth rate. [EL = 1++]

In women with an unfavourable cervix and ruptured membranes, intravenous oxytocin was less effective than vaginal PGE2 in achieving vaginal birth within 24 hours. [EL = 1++]

In women with a favourable cervix, intravenous oxytocin alone was less effective than vaginal PGE2 in achieving vaginal birth within 24 hours. [EL = 1++]

Interpretation of evidence

In women with an unfavourable cervix and intact membranes, the use of intravenous oxytocin alone when compared with vaginal PGE2 as an inducing agent results in fewer vaginal births within 24 hours, a lower Bishop score at 24 hours and more caesarean births.

In women with a favourable cervix, the use of intravenous oxytocin alone when compared with vaginal PGE2 as an inducing agent results in fewer vaginal births within 24 hours.

Recommendation on intravenous oxytocin alone

Intravenous oxytocin alone should not be used for induction of labour. (Refer to Section 5.1.7.)

5.1.7. Amniotomy with intravenous oxytocin

Overview of available evidence

One systematic review was identified.

One systematic review (17 RCTs involving 2566 women, mixed parity and mixed Bishop score) evaluated the effects of amniotomy plus oxytocin versus placebo/no treatment (one RCT); versus vaginal PGE2 (11 RCTs); versus cervical PGE2 (one RCT); versus oxytocin alone (two RCTs); and versus amniotomy alone (2 RCT).118 [EL = 1++]

Of the 17 RCTs reported in this review, four RCTs included women with an unfavourable cervix. There were no significant differences between amniotomy plus intravenous oxytocin and vaginal PGE2 in not achieving vaginal birth within 24 hours (9/21 (43%) versus 10/21 (48%); RR 0.90, 95% CI 0.46 to 1.75; one RCT, 42 women) or caesarean birth rate (11/51 (22%) versus 12/55 (22%); RR 0.98, 95% CI 0.48 to 2.03; two RCTs, 106 women). However, the number of cases in the studies was very small. Comparisons between amniotomy plus intravenous oxytocin and intracervical PGE2 found similar maternal and fetal outcomes.118 [EL = 1++]

For women with a favourable cervix, there was one RCT included in the review that compared amniotomy plus oxytocin with vaginal PGE2 and which reported significant increases in postpartum haemorrhage (8/50 (16%) versus 1/50 (2%); RR 8.00, 95% CI 1.04 to 61.62; one RCT, 100 women) and the proportion of women not satisfied (RR 53.00, 95% CI 3.32 to 846.47; one RCT, 100 women) in the amniotomy group.118 [EL = 1+] The rates of vaginal birth achieved within 24 hours were not reported in this trial. Compared with amniotomy alone, intravenous oxytocin plus amniotomy was significantly associated with achieving vaginal birth within 24 hours (RR1.17, 95% CI 1.09 to 1.26; two RCTs).118 [EL = 1++]

Evidence statements

Evidence from small trials suggested that, for women with an unfavourable cervix, amniotomy plus intravenous oxytocin achieves similar maternal and fetal outcomes as vaginal PGE2. [EL = 1++]

In women with a favourable cervix, one RCT from the review reported that amniotomy and intravenous oxytocin is significantly associated with postpartum haemorrhage and dissatisfaction with treatment, when compared with vaginal PGE2. Compared with oxytocin alone, women undergoing amniotomy and intravenous oxytocin are more likely to give birth vaginally within 24 hours. [EL = 1++]

Interpretation of evidence

In women with an unfavourable cervix, the studies investigating the effectiveness of amniotomy plus oxytocin compared with vaginal PGE2 were considered by the GDG to be small and therefore would be underpowered to give a reliable estimate of the effect size in the primary outcomes concerned.

In women with a favourable cervix, one trial reported that the use of intravenous oxytocin with amniotomy was associated with postpartum haemorrhage and reduced women’s satisfaction. This is likely to apply to women with unfavourable cervix as well. In addition, as this method required intravenous access and continuous monitoring, it is necessarily more invasive than the use of vaginal PGE2 and will limit women’s mobility during induction.

Recommendation on amniotomy with intravenous oxytocin

Amniotomy with oxytocin should not be used as a primary method of induction of labour unless there are specific contraindications to the use of vaginal PGE2, in particular the risk of uterine hyperstimulation.

5.1.8. Misoprostol

Misoprostol is a synthetic prostaglandin that can be given orally, vaginally or sublingually. It is effective in causing uterine contractions. However, misoprostol is not licensed for use in pregnancy in the UK. Oral misoprostol usually comes in tablets of 100 micrograms or 200 micrograms. Using small doses (50 micrograms) will involve dividing the tablet using a pill cutter, a technique that makes accurate dosage difficult.

Overview of available evidence

Four systematic reviews on oral, vaginal and sublingual misoprotol, additional RCTs and one unpublished RCT were identified.

Oral misoprostol (20–200 micrograms)

One systematic review (41 RCTs involving 8606 women, mixed parity and mixed Bishop score) assessed the effects of oral misoprostol (20–200 micrograms) versus placebo (four RCTs); versus vaginal dinoprostone (nine RCTs); versus intracervical prostaglandin (two RCTs); versus intravenous oxytocin (seven RCTs); and versus vaginal misoprostol (16 RCTs). Compared with placebo, oral misoprostol was effective as an induction agent.119 [EL = 1++]

For all women irrespective of parity, membranes and cervical status, caesarean birth was less likely to occur with oral misoprostol (50–100 micrograms) when compared with vaginal PGE2 (RR 0.88, 95% CI 0.76 to 1.01; nine RCTs) although this was not statistically significant. Maternal and fetal outcomes were comparable between oral misoprostol (50–200 micrograms) and intracervical PGE2. Meconium-stained liquor was more likely to occur with oral misoprostol than with oxytocin (RR 1.72, 95% CI 1.08 to 2.74; six RCTs). Similar maternal and fetal outcomes were achieved between oral misoprostol of different doses and regimens (three RCTs).

Compared with vaginal misoprostol (25 micrograms every 4 hours, maximum dose 150 micrograms), primiparous women with an unfavourable cervix given oral misoprostol (50 micrograms every 4 hours; max dose 300 micrograms) were significantly less likely to achieve vaginal birth within 24 hours (RR 1.25, 95% CI 1.01 to 1.55, one RCT). However, maternal and fetal outcomes were comparable between oral and vaginal misoprostol in multiparous women with an unfavourable cervix. Comparisons between oral misoprostol (20 micrograms every 2 hours × 2, then 40 micrograms every 2 hours × 10 until three contractions every 10 minutes, maximum dose 475 micrograms) and vaginal PGE2 gel (2 mg 6-hourly) found no significant difference in achieving vaginal birth within 24 hours between the two groups (one RCT). Analyses of outcomes of all women suggested that oral misoprostol (50–100 micrograms) may be associated with a reduced risk of caesarean birth (RR 0.88, 95% CI 0.76 to 1.01; nine RCTs). There were no perinatal deaths.119 [EL = 1++]

Additional RCTs identified found vaginal misoprostol 50 micrograms to have a higher incidence of uterine hyperstimulation when compared with oral misoprostol 100 micrograms.120 [EL = 1+] Oral misoprostol 50 micrograms was more effective than 25 micrograms in shortening the mean initiation-to-birth interval.121 [EL = 1+] Birth within 48 hours was significantly more likely with oral misoprostol 50 micrograms than vaginal prostaglandin 4 mg.122 [EL = 1+]

Titrated low-dose oral misoprostol (25 micrograms) was more effective than standard regimen (vaginal PGE2 plus intravenous oxytocin) in terms of achieving vaginal birth within 24 hours and reduced the caesarean birth rate, in women with prelabour rupture of membranes. There were significantly more maternal side effects reported with the use of misoprostol (19% versus 13%, RR 1.42, 95% CI 1.02 to 1.98). These side effects included nausea, vomiting, diarrhoea, shivering and pyrexia during labour.123 [EL = 1+] (unpublished)

Vaginal misoprostol (25–100 micrograms)

One systematic review (70 RCTs involving 10 524 women, with both mixed parity and Bishop score) compared the effects of vaginal misoprostol (25–100 micrograms) versus placebo (five RCTs); versus vaginal PGE2 (25 RCTs); versus intracervical PGE2 (17 RCTs); versus oxytocin (13 RCTs); vaginal misoprostol lower dose regimen versus higher dose (13 RCTs); and misoprostol gel versus tablets (one RCT). For women with an unfavourable cervix, compared with placebo, vaginal misoprostol showed effectiveness as an induction agent. Compared with vaginal PGE2 (gel, tablet or controlled release pessary), vaginal misoprostol (50 micrograms) was significantly more likely to achieve a favourable cervix within 12–24 hours (RR 1.15, 95% CI 1.01 to 1.31; one RCT), vaginal birth within 24 hours (RR 1.19, 95% CI 1.11 to 1.26; 13 RCTs), and was associated with uterine hyperstimulation both with FHR changes (RR 2.32, 95% CI 1.62 to 3.32; 17 RCTs) and without FHR changes (RR 2.93, 95% CI 2.04 to 4.20; seven RCTs) and a reduced need for oxytocin augmentation (RR 0.64, 95% CI 0.56 to 0.73; 11 RCTs).124 [EL = 1++]

Compared with intracervical PGE2, vaginal misoprostol (44–88 micrograms) was significantly more likely to achieve an improved cervical status after 12–24 hours and vaginal birth within 24 hours. Vaginal misoprostol was significantly associated with uterine hyperstimulation with and without FHR changes (RR 2.19, 95% CI 1.47 to 3.27; 14 RCTs and RR 1.90, 95% CI 1.44 to 2.49; nine RCTs, respectively) and a reduced need for oxytocin augmentation (RR 0.57, 95% CI 0.51 to 0.62; 11 RCTs). There was no significant difference in caesarean birth rates between the two groups (RR 1.04, 95% CI 0.88 to 1.23; 16 RCTs).

Compared with intravenous oxytocin, vaginal misoprostol was significantly associated with uterine hyperstimulation without FHR changes (RR 2.52, 95% CI 1.45 to 4.36; four RCTs). Other maternal and fetal outcomes were similar between the two groups.

Compared with vaginal misoprostol high dose (maximum 50 micrograms), low-dose regimens (minimum 12.5 micrograms) were significantly associated with reduced uterine hyperstimulation with and without FHR changes (RR 0.55, 95% CI 0.38 to 0.79’ nine RCTs and RR 0.66, 95% CI 0.50 to 0.85; four RCTs, respectively) and an increased need for oxytocin augmentation (RR 1.30, 95% 1.14 to 1.49; five RCTs). Maternal side effects (nausea, vomiting and diarrhoea) were less likely to be reported with low dose (RR 0.77, 95% CI 0.45 to 1.30; four RCTs), although the difference was not statistically significant.

Compared with a vaginal misoprostol tablet (50 micrograms), vaginal misoprostol gel (50 micrograms) was significantly less likely to cause uterine hyperstimulation with FHR changes (RR 0.49, 95% CI 0.29 to 0.83; one RCT) but more likely to need oxytocin augmentation (RR 1.26, 95% 1.13 to 1.41; one RCT) and epidural analgesia (RR 1.19, 95% CI 1.03 to 1.38).124 [EL = 1++]

Additional RCTs identified reported that vaginal misoprostol 50 micrograms was associated with increased likelihood of birth within 24 hours and reduced need for oxytocin augmentation when compared with vaginal PGE2 (3 mg).125 [EL = 1+] Vaginal misoprostol 50 micrograms was significantly more likely than vaginal PGE2 (10 mg) to cause uterine hyperstimulation.126 [EL = 1+] Vaginal misoprostol 25 micrograms and PGE2 gel 1–2 mg achieved similar maternal and fetal outcomes.127 [EL = 1+]

A drug company-sponsored multicentre phase III RCT (unpublished) in 19 UK cities compared the effects of vaginal misoprostol 25 micrograms 4-hourly for up to three doses (n = 318, 56% nulliparous) versus vaginal PGE2 3 mg 6-hourly for up to two doses (n = 308, 58% nulliparous) in women at term with an unfavourable cervix. Both methods were similar in achieving vaginal births within 24 hours (43% versus 47%; absolute difference 3.74%, 95% CI −3.58% to 11.05%), with vaginal misoprostol significantly associated with birth within 12 hours (11% versus 18%, P = 0.0067). However, a significantly higher caesarean birth rate (28% versus 22%, P = 0.037) and lower incidence of maternal nausea (13% versus 20%, P = 0.025) was reported in the vaginal misoprostol group. All other maternal and fetal outcomes were comparable between the two groups.128 [EL = 1+]

Additional RCTs identified suggested that vaginal misoprostol 25–50 micrograms was more effective than intravenous oxytocin in achieving vaginal birth within 24 hours129 [EL = 1+], and with lower caesarean birth rates but increased uterine hyperstimulation.130 [EL = 1+] One RCT did not show any difference between the two interventions for maternal and fetal outcomes.131 [EL = 1+]

Vaginal misoprostol 100 micrograms and 50 micrograms achieved comparable maternal and fetal outcomes.132 [EL = 1+] Vaginal misoprostol 50 micrograms was associated with shorter initiation-to-birth interval and reduced need for oxytocin augmentation when compared with isosorbide mononitrate (IMN) 40 mg. However, uterine hyperstimulation and reports of headaches, nausea and dizziness were more likely in the IMN group.133 [EL = 1+]

Buccal misoprostol (50–100 micrograms)

One systematic review (three RCTs, 502 women, mixed parity and Bishop score) compared the effects of buccal or sublingual misoprostol (50–100 micrograms) versus vaginal misoprostol (one RCT); and versus oral misoprostol (two RCTs).

Overall, there were no significant differences in maternal or fetal outcomes between buccal/sublingual and vaginal misoprostol. There were no valid outcomes reported in this review for women with an unfavourable cervix.134 [EL = 1++]

One additional systematic review was identified that evaluated the use of misoprostol orally, vaginally, sublingually or buccally, compared with PGE2, vaginally or intracervically, for induction of labour in women at term with an unfavourable cervix and intact membranes. It included 14 RCTs, some of which were included in reviews in previous sections. The comparisons included oral misoprostol versus vaginal PGE2 gel (one RCT); versus intracervical PGE2 gel (one RCT); vaginal misoprostol versus vaginal PGE2 gel (four RCTs); versus vaginal PGE2 controlled release (two RCTs); versus vaginal PGE2 tablet (one RCT); versus vaginal PGE2 pessary (one RCT); and versus intracervical PGE2 gel (four RCTs).135 [EL = 1++]

This review reported that, compared with PGE2, any misoprostol was associated with a higher risk of tachysystole (RR 1,86, 95% CI 1.01 to 3.43), hyperstimulation (RR 3.72, 95% CI 2.00 to 6.88), higher rate of vaginal birth within 24 hours (RR 1.14, 95% CI 1.00 to 1.31), a lower rate of oxytocin use (RR 0.71, 95% CI 0.60 to 0.95) and a trend towards increased meconium staining (RR 1.22, 95% CI 0.96 to 1.55). There was no significant difference between the two groups in the incidence of caesarean birth (RR 0.99, 95% CI 0.83 to 1.17). The use of misoprostol at starting dosages above 25 micrograms had similar findings to the primary analysis. Lower misoprostol doses (starting at 25 micrograms) did not show any significant differences in maternal or fetal outcomes.135 [EL = 1++]

Evidence statements

Oral misoprostol

Evidence suggested that, irrespective of cervical status, oral misoprostol is more effective than placebo as an induction agent. There is no significant difference in maternal and fetal outcomes between oral misoprostol (200 micrograms) and intracervical PGE2. [EL = 1++]

The use of oral misoprostol (100 micrograms) is more likely than oxytocin to be associated with meconium-stained liquor. Oral misoprostol 50 micrograms or 100 micrograms achieve similar maternal and fetal outcomes. Oral misoprostol (50–100 micrograms) is less likely than vaginal PGE2 to result in caesarean birth (borderline significance). [EL = 1++]

In women with an unfavourable cervix, oral misoprostol 50 micrograms is less likely than vaginal misoprostol 25 micrograms to achieve vaginal birth within 24 hours. Oral misoprostol has similar efficacy to vaginal PGE2 gel in terms of vaginal birth within 24 hours. [EL = 1++]

Vaginal misoprostol

Evidence suggested that, for women with an unfavourable cervix, vaginal misoprostol is more effective than placebo as an induction agent. Vaginal misoprostol (50–100 micrograms) is more likely than vaginal PGE2 to produce a favourable cervix within 24 hours, achieve birth within 24 hours, and cause uterine hyperstimulation. [EL = 1++]

Vaginal misoprostol (50–100 micrograms) is more likely than intravenous oxytocin to cause uterine hyperstimulation without FHR changes. Vaginal misoprostol at lower dose (minimum 25 micrograms) was less likely than high dose (maximum 50 micrograms) to cause uterine hyperstimulation with and without FHR changes. [EL = 1++]

Vaginal misoprostol gel (50 micrograms) is less likely than vaginal misoprostol tablet to cause uterine hyperstimulation with FHR changes, but more likely to need oxytocin augmentation and epidural analgesia. [EL = 1++]

Vaginal misoprostol is more likely than IMN to achieve earlier birth and not need oxytocin augmentation. Tachysystole and uterine hyperstimulation are less likely in women given vaginal IMN. There were more reports of headaches, nausea and dizziness in the IMN group. [EL = 1+]

Buccal misoprostol

For women with an unfavourable cervix, there were insufficient data to determine the effectiveness of buccal/sublingual misoprostol as compared with oral and vaginal misoprostol. [EL = 1++]

Compared with PGE2, any misoprostol is more effective in achieving vaginal birth within 24 hours and lessening the need for oxytocin use, but any misoprostol is associated with higher risks of hyperstimulation and increased meconium staining. Caesarean birth rates were similar between the two interventions. [EL = 1++]

A review conducted by the World Health Organization (in press) of the evidence from the four systematic reviews above119,124,134,135 concluded that the currently available studies are not large enough to have adequate statistical power to assess the safety issues of the induction process with misoprostol and the long-term follow up of babies exposed to misoprostol. Trials or meta-analyses that have adequate power to address rare adverse fetal outcomes will need to include in excess of 30 000 women.

Interpretation of evidence

Misoprostol is currently unlicensed for use in pregnancy in the UK.

Oral and vaginal misoprostol (for women with undefined, variable and unfavourable cervix)
Vaginal misoprostol (favourable cervix)

There were insufficient data comparing this route with other regimens to reach a conclusion.

Buccal/sublingual misoprostol (both unfavourable and favourable cervix)

There were insufficient data comparing this route with other regimens to reach a conclusion.

Recommendation on misoprostol

Misoprostol* should only be offered as a method of induction of labour to women who have intrauterine fetal death (see Section 4.9) or in the context of a clinical trial.

*

At the time of publication (July 2008), misoprostol was not licensed for use for labour induction in fetal death in utero in the UK. Informed consent should therefore be obtained and documented.

5.1.9. Mifepristone

Mifepristone, also known as RU 486, is an antiprogestin and has been developed to antagonise the action of progesterone. Mifepristone now has an established role in the termination of pregnancy, in combination with prostaglandins, during the first and second trimester.

Overview of available evidence

One systematic review was identified. The GDG was alerted to one recent study from China that reported serious neonatal side effects associated with the use of mifepristone.

One systematic review (seven RCTs, 594 women, mixed parity and Bishop score < 6) that evaluated the effects of mifepristone versus placebo/no treatment in women at term found insufficient information to support the use of mifepristone to induce labour.136 [EL = 1++] However, there is recent evidence of serious neonatal side effects involving renal function in the form of ischaemic hypoxic changes in the fetal kidney ultrastructure when labour was induced by mifepristone between 16 and 28 weeks of gestation. The smaller the fetus, the more obvious the changes.137 [EL = 2+]

Evidence statement

There is insufficient information to support the use of mifepristone to induce labour. [EL = 1++}

One study in China found ischaemic changes in the fetal kidney when labour was induced using mifepristone at between 16 and 28 weeks of gestation. [EL = 2+]

Interpretation of evidence

There is concern from the latest evidence that mifepristone may be associated with fetal kidney damage. The efficacy and safety of mifepristone as an induction agents needs to be established.

Recommendation on mifepristone

Mifepristone should only be offered as a method of induction of labour to women with intrauterine fetal death (see Section 4.9).

5.1.10. Hyaluronidase

The level of hyaluronic acid increases markedly after the onset of labour. Cervical injection of hyaluronidase was postulated to increase cervical ripening.

Overview of available evidence

One systematic review was identified.

This systematic review (one RCT involving 168 women, Bishop score unknown) assessed the effects of intracervical hyaluronidase in women undergoing induction of labour. Women given hyaluronidase were reported to achieve significant improvement in cervical status (RR 0.62, 95% CI 0.52 to 0.74) and there were significantly fewer caesarean births (RR 0.37, 95% CI 0.22 to 0.61) when compared with placebo. No side effects for mother or baby were reported.138 [EL = 1++]

Evidence statements

Evidence suggested that intracervical hyaluronidase is likely to improve cervical ripening and reduce caesarean rates when compared with placebo. [EL = 1++]

Interpretation of evidence

Although intracervical hyaluronidase may be effective in improving cervical ripening and reducing caesarean birth rates, it is an invasive procedure that women may find unacceptable when alternative available methods such as vaginal PGE2 are less invasive.

Recommendation on hyaluronidase

Hyaluronidase should not be used for induction of labour.

5.1.11. Corticosteroids

Corticosteroids are postulated to have a promoting effect in induction of labour but their role in the process of labour is not well understood.

Overview of available evidence

One systematic review was identified.

This systematic review (one RCT involving 66 women, favourable cervix) assessed the effects of corticosteroids versus intravenous oxytocin in cervical priming and induction of labour. Vaginal birth within 24 hours was not reported. There were no reports of uterine hyperstimulation, Apgar score < 7 or maternal fever in either group, and caesarean birth rates were not significantly different (RR 0.40, 95% CI 0.08 to 1.92).139 [EL = 1++]

Evidence statements

The available evidence relating to the effects of corticosteroids for cervical priming and induction of labour is limited. [EL = 1++]

Recommendation on corticosteroids

Corticosteroids should not be used for induction of labour.

5.1.12. Oestrogens

The increase in the serum oestrogen-to-progesterone ratio that occurs before the onset of labour is believed to activate prostaglandin production, which in turn stimulates cervical ripening.

Overview of available evidence

One systematic review was identified.

This systematic review (six RCTs involving 341 women, Bishop score < 3) assessed the effects of oestrogens in women undergoing induction of labour. The included studies compared oestrogen (intravenous, oral, vaginal or extra-amniotic) versus placebo (four RCTs); versus vaginal PGE2 (one RCT); versus intracervical PGE2 (one RCT); versus oxytocin (one RCT); and versus extra-amniotic PGF (one RCT). It reported no significant differences between the oestrogens and the placebo groups in the rates of caesarean births, instrumental vaginal births or uterine hyperstimulation with or without FHR changes. There were insufficient data for the remaining comparisons. Overall, there were insufficient data to make any meaningful conclusions.140 [EL = 1++]

Evidence statements

Limited evidence suggested that oestrogen and placebo achieve similar maternal and fetal outcomes. There was insufficient data available for the comparisons between oestrogen and vaginal PGE2, oxytocin alone or extra-amniotic PGF. [EL = 1++]

Interpretation of evidence

Oestrogens and placebo achieved similar maternal and fetal outcome There was insufficient evidence to determine the effectiveness of oestrogen for cervical ripening.

Recommendation on oestrogens

Oestrogen should not be used for induction of labour.

5.1.13. Vaginal nitric oxide donors

Nitric oxide is considered a fundamental mediator of cervical ripening without causing uterine contractions or adverse effects on the mother and fetus.

Overview of available evidence

Four RCTs were identified.

Vaginal glyceryl trinitrate versus vaginal prostaglandins

One RCT in Thailand compared the effects of 6-hourly vaginal glyceryl trinitrate 500 micrograms (n = 54) versus 6-hourly vaginal PGE2 tablet 3 mg (n = 56) in women with 40 weeks or more of gestation and unfavourable cervix (Bishop score = 6). Women in the glyceryl trinitrate group were more likely than the PGE2 group to have a longer duration from start of medication to birth (26 versus 22 hours, P = 0.01), a lower incidence of uterine hyperstimulation (0% versus 9%, P = 0.02) and an increased need for oxytocin (78% versus 43%, P < 0.001). There were more side effects (headaches and palpitation) reported in the glyceryl trinitrate group. Other maternal and fetal outcomes were similar between the two groups.141 [EL = 1+]

Vaginal isosorbide mononitrate (IMN) versus placebo

A double-blind RCT in Sweden compared the effects of vaginal nitric oxide donor IMN 40 mg (n = 100) versus placebo (n = 100) in women with uncomplicated pregnancy at 42 weeks or more of gestation and a Bishop score < 6. Compared with placebo, vaginal IMN was significantly associated with onset of labour within 24 hours (22% versus 8%, P = 0.01) and headaches (88% versus 8%).142 [EL = 1+]

Vaginal IMN versus vaginal misoprostol

One RCT in Thailand compared the effects of vaginal IMN 40 mg (n = 55) versus vaginal misoprostol 50 micrograms (n = 52) in women at term. Vaginal IMN was associated with a lower incidence of uterine hyperstimulation (0% versus 15%, P < 0.01) but a longer induction-to-birth interval (26 versus 14 hours, P < 0.01) and increased need for oxytocin (92% versus 11%). Caesarean birth rates were similar between the two groups.133 [EL = 1+]

Vaginal IMN versus vaginal PGE2 gel

An RCT in the UK compared the effects of vaginal IMN 40 mg (n = 199) versus vaginal PGE2 gel 2 mg (n = 199) for cervical priming in nulliparous women at 38 weeks of gestation with a modified Bishop score < 6. It reported a significantly longer treatment-to-birth interval (39.7 versus 26.9 hours, P < 0.001) and a lower mean change in Bishop score at 24 hours (1.35 versus 2.79, P < 0.001) in the IMN group when compared with the PGE2 group. Modes of birth were similar between the two groups. However, abnormal FHR was significantly more likely in the PGE2 group (0% versus 7%, P = 0.0002). There were significantly more side effects (nausea, hot flushes, headaches, faintness and abdominal pain) reported in the IMN group. Maternal satisfaction was significantly higher in the IMN group (mean VAS 7.0 versus 5.8, P < 0.0001). Women in the IMN group were significantly more likely to prefer IMN as an outpatient treatment (55% versus 17%, P < 0.0001).143 [EL = 1+]

Evidence statements

Evidence suggested that, in women with an unfavourable cervix, vaginal glyceryl trinitrate is associated with a longer induction-to-birth interval but a lowered incidence of uterine hyperstimulation, compared with vaginal PGE2. There were more side effects reported with the use of vaginal glyceryl trinitrate, such as headaches and palpitation. [EL = 1+]

Vaginal IMN is effective in initiating labour within 24 hours when compared with placebo. However, headaches were more frequently reported with its use. Compared with vaginal misoprostol, vaginal IMN results in fewer adverse effects but is less effective in shortening the induction-to-birth interval. Compared with vaginal PGE2, IMN is associated with a longer induction-to-birth interval and a lower Bishop score at 24 hours. There was a higher incidence of gastrointestinal side effects in the IMN group. However, maternal satisfaction was high in the IMN group. [EL = 1+]

Interpretation of evidence

Vaginal glyceryl trinitrate and nitric oxide donors have not been shown to be of any particular benefit when compared with vaginal PGE2 as induction agents, although they seem to be associated with less uterine hyperstimulation. However, there are significant side effects associated with its use. Results of a recent trial are in the process of being published.

Recommendation on nitric oxide donors

Vaginal nitric oxide donors should not be used for induction of labour.

5.2. Non-pharmacological methods

Clinical question

5.2.1. Membrane sweeping

Stripping/sweeping of the membranes was used as a method for inducing labour at least as early as 1810.144 Increased local production of prostaglandins following vaginal examination for membrane sweeping provides a plausible explanation for the effect of this procedure on pregnancy duration.145 Vaginal examination allows an assessment of the condition of the cervix which informs clinical decision making. Carried out in late pregnancy, when consideration is being given to induction, it offers the opportunity to undertake membrane sweeping. If the woman is on the threshold of spontaneous labour, a membrane sweep may be all that is required to initiate it, thus reducing the need for formal induction of labour. The procedure entails passage of the examining finger through the cervix so that it can be rotated against the wall of the uterus beyond the internal cervical os, thereby stripping the chorion away from the decidua (the decidua is the richest source of PGF within the uterus). Clearly if the cervix will not admit a finger it may not be possible to strip the membranes but in such cases massaging around the cervix in the vaginal fornices may achieve a similar effect.

For the purpose of this guideline, membrane sweeping is regarded as an adjunct to induction of labour rather than as a method per se.

Overview of available evidence

One systematic review and one additional RCT were identified. Reference is made to the NICE clinical guideline on antenatal care as supplementary evidence.

One systematic review (22 RCTs involving 2797 women, Bishop score ranged from ‘closed’ to 6 or less, mixed parity) compared sweeping of membranes with no treatment (20 RCTs) and compared membrane sweeping with prostaglandins (three RCTs) and oxytocin (one RCT). Two studies reported more than one comparison. Women at 37–40 weeks and those at 40 weeks or more of gestation were included in 16 studies and six studies, respectively. Unfavourable cervix (as defined by triallists) was reported in seven studies. The interventions included weekly membrane sweeping (seven RCTs), sweeping every 3 days (one RCT) and daily sweeping (two RCTs). The control groups received cervical assessment or gentle vaginal examination.146

All studies in this review,147–153 irrespective of sweeping frequency, reported that membrane sweeping was associated with a reduced number of pregnancies beyond 41 weeks (RR 0.59, 95% CI 0.46 to 0.74) and 42 weeks (RR 0.28, 95% CI 0.15 to 0.50). To avoid one formal induction of labour, sweeping of membranes would be performed in eight women (number needed to treat (NNT) = 8). There were no significant differences between the sweeping and no-treatment groups in terms of caesarean births (RR 0.90, 95% CI 0.70 to 1.15) or risks of maternal or neonatal infection. There were four perinatal deaths (two in each group, one stillbirth with meconium-stained liquor in the sweeping group, one with double nuchal cord in the control group and two from congenital heart defects). More women in the sweeping group reported discomfort during vaginal examination and other adverse effects such as bleeding and irregular contractions.146 [EL = 1++]

Women with an unfavourable cervix and gestational age between 38 and 42 weeks154–157 were significantly less likely to require formal induction of labour (RR 0.51, 95% CI 0.37 to 0.71; three RCTs, 226 women) when they underwent membrane sweeping. There were no significant differences between sweeping and no sweeping for caesarean births (RR 0.98, 95% CI 0.49 to 1.95; three RCTs, 200 women), epidural usage (RR 0.70, 95% CI 0.42 to 1.18; one RCT, 65 women), instrumental vaginal births (RR 0.87, 95% CI 0.33 to 2.24; two RCTs, 135 women), 5 minute Apgar score < 7 (RR 0.97, 95% CI 0.06 to 4.85; one RCT, 65 women) or neonatal intensive care unit admissions (RR 0.97, 95% CI 0.15 to 6.47; one RCT, 65 women). There was no maternal or perinatal mortality.146 [EL = 1++]

There were limited data available in studies comparing membrane sweeping versus vaginal prostaglandins (two RCTs) or intravenous oxytocin (one RCT) in women with an unfavourable cervix. These studies did not show any significant differences in the need for formal induction, caesarean birth rates or other maternal and fetal outcomes.146 [EL = 1++]

One additional RCT from the Netherlands,151 not included in the review,146 evaluated the effects of membrane sweeping, repeated every 48 hours (n = 375) and no membrane sweeping (routine monitoring) (n = 367) in women with low-risk pregnancy at 41 weeks of gestation and a median Bishop score of 4. Serial sweeping significantly reduced the proportion of post-term pregnancies (defined as 42 weeks or more gestational age) (23% versus 41%; RR 0.57, 95% CI 0.46 to 0.71) in both nulliparous and multiparous women. The need for induction of labour at or after 42 weeks was 15% in the sweeping group and 26% in the control group (RR 0.56, 95% CI 0.42 to 0.75). Sweeping significantly increased the likelihood of birth in a primary care setting in parous women (67% versus 51%; RR 1.32, 95% CI 1.11 to 1.58) but not in nulliparous women. Sweeping reduced the incidence of induction of labour in parous women (15% versus 27%; RR 0.57, 95% CI 0.37 to 0.86) with no effect in nulliparous women (29% versus 31%; RR 0.92, 9%% CI 0.68 to 1.25). Adverse effects were similar in both the sweeping and control groups in analgesia use and fever during labour, mode of birth and adverse neonatal outcomes. However, uncomplicated bleeding was reported significantly more frequently in the sweeping group (34% versus 5%; RR 6.58, 95% CI 3.98 to 10.87). There were two perinatal deaths in each group, one due to possible group B streptococcal infection in the sweeping group and one unexplained death at 42 weeks after a failed vacuum extraction. Membrane sweeping was reported to be ‘not painful’ in 31%, ‘somewhat painful’ in 51%, and ‘painful’ or ‘very painful’ in 17% of women. After birth, 88% of them would choose this procedure in a next pregnancy. Of the women who described sweeping as painful (all ‘painful’ categories), 88% reported that they would choose sweeping again in the next pregnancy.151 [EL = 1+]

Evidence statements

In women with an unfavourable cervix, evidence suggested that membrane sweeping and no membrane sweeping achieve comparable maternal and fetal outcomes including sanalgesia use. However, membrane sweeping is associated with:

  • reduced need for formal induction of labour, especially in multiparous women
  • increased rate of spontaneous labour, if performed more than once from 38 weeks of gestation; the most appropriate regimen is not clear from the evidence
  • increased incidence of uncomplicated bleeding
  • increased reports of pain but most women would still choose sweeping in a future pregnancy and recommend it to friends.

Evidence also suggests benefits for repeated sweeping attempts. There is also evidence that one attempt may be sufficient.

Data were limited with regard to providing evidence of benefits in comparisons between sweeping and vaginal PGE2 or intravenous oxytocin. [EL = 1++]

Interpretation of evidence

Compared with no sweeping, sweeping reduces the need for formal induction of labour. Additional membrane sweeping may be beneficial.

Membrane sweeping is an important and integral part of preventing prolonged pregnancy, and should be scheduled to be discussed with the woman at her routine antenatal visit.

The GDG considered it important to offer women information relating to the possibility of induction of labour to prevent prolonged pregnancy at their 38 week antenatal visit, to give women time to consider the options such as vaginal examination for membrane sweeping, before their next scheduled antenatal visits. Women may accept or decline this offer of information, and the options.

Recommendations on membrane sweeping

Prior to formal induction of labour, women should be offered a vaginal examination for membrane sweeping.*

At the 40 and 41 week antenatal visits, nulliparous women should be offered a vaginal examination for membrane sweeping.

At the 41 week antenatal visit, parous women should be offered a vaginal examination for membrane sweeping.

When a vaginal examination is carried out to assess the cervix, the opportunity should be taken to offer the woman a membrane sweep.

Additional membrane sweeping may be offered if labour does not start spontaneously.

*

This recommendation is from ‘Antenatal care: routine care for the healthy pregnant woman’ (NICE clinical guideline 62). Available from www​.nice.org.uk/CG062.

Research recommendation on membrane sweeping

Research is needed to assess effectiveness, maternal satisfaction and acceptability of:

Research question

What are the effectiveness and acceptability of, and maternal satisfaction with, the following:

Why is this important?

Membrane sweeping is considered to be a relatively simple intervention that may positively influence the transition from maintenance of pregnancy to the onset of labour, reducing the need for formal induction of labour. However, there are disadvantages, such as possible vaginal bleeding and discomfort. Research into when and how frequently membrane sweeping should be carried out to maximise its effectiveness and acceptability would be of value.

5.2.2. Herbal supplements

The use of herbal supplements to promote health has become popular. It is believed by some that drinking herbal beverage teas while pregnant nourishes and tones the uterus, supporting optimal health in pregnancy.

Evidence statements

No evidence was identified relating to the effects of herbal supplements in cervical priming/induction of labour.

Interpretation of evidence

There is no evidence available to determine the effects of herbal supplements as an induction agent. The GDG considered that the unsupervised use of herbal preparations, which may contain active ingredients with undesirable effects, should be treated with caution.

Recommendation on herbal supplements

Herbal supplements as a method of induction of labour should not be offered.

Research recommendation on herbal supplements

Further research is required to evaluate the effectiveness, safety and maternal satisfaction of the use of herbal supplements as a method of induction of labour.

5.2.3. Acupuncture

Acupuncture involves the insertion of very fine needles into specific points of the body. It has been hypothesised that neuronal stimulation by acupuncture may increase uterine contractility. It is also gaining acceptance as a method to alleviate labour pain and ripen the cervix.158

Overview of available evidence

One systematic review and an additional RCT were identified.

One systematic review (one RCT involving 56 women, Bishop score < 5, mixed parity) that assessed the effects of acupuncture in women undergoing induction at term found no meaningful data on the effectiveness of acupuncture as a cervical priming method, owing to methodological limitations and drop-out rates.159 [EL = 1++]

One additional RCT was identified in the USA that compared the effects of usual medical care alone (not specified) (n = 26) and usual care plus three outpatient acupuncture treatments (n = 30) in nulliparous women with uncomplicated pregnancies at term with a median Bishop score of 4. Women continued to receive medical care in either group (for example, membrane sweeping, timing of inductions or herbal supplementation for cervical ripening). There were no significant differences between the acupuncture group and the control group in spontaneous labour (70% versus 50%; OR 2.33, 95% CI 0.78 to 6.98) or caesarean birth rates (17% versus 39%; OR 3.13, 95% CI 0.99 to 10.8).160 [EL = 1+]

Evidence statements

The available evidence is insufficient to determine the effectiveness of acupuncture in cervical priming/induction of labour. [EL = 1++]

Interpretation of evidence

In the absence of sufficient evidence that proves either effectiveness or harm, acupuncture as a method of induction is not recommended to be offered.

Recommendation on acupuncture

Acupuncture as a method of induction of labour should not be offered.

Research recommendation on acupuncture

Further research is required to evaluate the effectiveness, safety and maternal satisfaction of acupuncture as a method of induction of labour.

5.2.4. Homeopathy

Homeopathy involves the administration in dilution of substances aimed at the alleviation of symptoms that the same substances generally cause in their undiluted form. It has been suggested that the herbs belonging to the Caulophyllum genus are useful in establishing labour, when uterine contractions are short and/or irregular or when they stop.161

Overview of available evidence

One systematic review was identified.

One systematic review (two RCTs involving 133 women, cervical dilation up to 3 cm) assessed the effects of caulophyllum for cervical priming and induction of labour. There was insufficient methodological information for the studies included and clinically meaningful outcomes were limited.162 [EL = 1++]

Evidence statements

The available evidence was poor and insufficient to determine the effectiveness of homeopathy as a method of induction of labour. [EL = 1++]

Interpretation of evidence

In the absence of sufficient evidence to prove either effectiveness or harm, homeopathy as a method of induction is not recommended to be offered.

Recommendation on homeopathy

Homeopathy as a method of induction of labour should not be offered.

Research recommendation on homeopathy

Further research is required to evaluate the effectiveness, safety and maternal satisfaction of homeopathy as a method of induction of labour.

5.2.5. Castor oil, hot baths and enemas

Castor oil has been widely used as a traditional method of initiating labour in midwifery practice. However, the mechanism is poorly understood.

Overview of available evidence

One systematic review and an additional RCT were identified. No evidence was identified which assessed the effects of hot baths and enemas in induction of labour.

The systematic review (one quasi-RCT involving 103 women, Bishop score < 4, intact membranes, parity unknown) compared the effects of a 60 ml single dose of castor oil (diluted in orange or apple juice) versus no treatment in women requiring induction of labour. There was no evidence of differences between the two groups in caesarean birth rate, meconium-stained liquor or Apgar score < 7 at 5 minutes. All women who ingested castor oil felt nauseous (RR 97.08, 95% CI 6.16 to 1530.41).163 [EL = 1++]

A small RCT in Iran compared the effects of castor oil (n = 24) and control (no intervention) (n = 23) in women at 40–42 weeks of gestation (Bishop score = 4, parity unknown). It reported a significant increase in the initiation of labour in the castor oil group compared with the control group (54.2% versus 4.3%, P < 0.001) and an increase in the mean Bishop score in the castor oil group (from 2.50 ± 1.29 to 6.79 ± 3.20, P < 0.001). There were no significant differences between the two groups in Apgar scores, meconium-stained liquor or methods of birth. Women given castor oil were significantly more likely to report nausea (45.8% versus 0%).164 [EL = 1+]

Evidence statements

Evidence suggested that women given castor oil for induction of labour achieve similar maternal and fetal outcomes as women given placebo. [EL = 1++] One small RCT reported improved Bishop scores in women given castor oil. [EL = 1+] However, both studies reported that castor oil was associated with nausea. [EL = 1++]

Interpretation of evidence

There is limited and conflicting evidence relating to the effects of castor oil for cervical priming and induction of labour. Castor oil is unpleasant to ingest and causes nausea. There is no available evidence relating to hot baths or enemas as induction agents.

Recommendation on castor oil, hot baths and enemas

Castor oil, hot baths and enemas as methods of induction should not be offered.

Research recommendation on castor oil, hot bath and enemas

Further research is required to evaluate the effectiveness, safety and maternal satisfaction of the use of castor oil, hot baths and enemas as methods of induction of labour.

5.2.6. Sexual intercourse

The role of sexual intercourse in stimulating labour is not well understood. It has been suggested that human semen is a biological source of high prostaglandin concentrations and the action of sexual intercourse may stimulate uterine contractions. There may be an endogenous release of oxytocin as a result of orgasm.

Overview of available evidence

One systematic review was identified.

One systematic review (one RCT involving 28 women, Bishop score and parity not known) assessed the effects of sexual intercourse for cervical priming and induction of labour. Data were limited and the review reported no significant differences in changes in Bishop score (1.0 versus 0.5, P > 0.05), Apgar scores < 7 at 5 minutes (0% versus 0%) or number of women delivered within 3 days of intervention (46% versus 47%; RR 0.99, 95% CI 0.45 to 2.20) between the group who had sexual intercourse for three consecutive nights with vaginal sperm deposit and the control group who abstained from sexual intercourse.165 [EL = 1++]

Evidence statements

One small study with limited data found no significant difference in labour outcomes between sexual intercourse and no sexual intercourse. [EL = 1++]

Interpretation of evidence

In the absence of sufficient evidence to prove either effectiveness or harm, sexual intercourse as a method of induction of labour is not recommended.

Recommendation on sexual intercourse

Sexual intercourse as a method of induction of labour should not be used.

Research recommendation on sexual intercourse

Further research is required to evaluate the effectiveness, safety and maternal satisfaction of sexual intercourse as a method of induction of labour.

5.2.7. Breast stimulation

It is known that breast stimulation results in the production of endogenous oxytocin in both pregnant and non-pregnant women,166,167 causing uterine contractions.

Overview of available evidence

One systematic review was identified.

One systematic review (six RCTs, 719 women, Bishop score 5–7) assessed the effects of breast stimulation for cervical priming and induction of labour. Breast stimulation was significantly associated with increased numbers of women achieving labour by 72 hours (93.6% versus 62.7%; RR 5.79, 95% CI 3.41 to 9.81; four RCTs) and a reduction in the rate of postpartum haemorrhage (0.7% versus 6%; RR 0.16, 95% CI 0.03 to 0.87) when compared with no breast stimulation. No significant differences were detected in the rates of caesarean birth or meconium staining. There were no instances of uterine hyperstimulation. For women with unfavourable cervix, one small trial168 in this review169 reported three perinatal deaths in the breast stimulation group (1.8% versus 0%; RR 8.17, 95% CI 0.45 to 147.77; one RCT, 37 women).

When comparing breast stimulation with oxytocin alone, the analysis found no differences in caesarean birth rates or the number of women not in labour after 72 hours. There was one perinatal death in the oxytocin group. None of the RCTs included in this review reported on women’s satisfaction with the treatment. The methods and frequency of breast stimulation varied in these studies. [EL = 1++]

Evidence statements

Evidence suggested that breast stimulation appears to be beneficial in increasing the number of women in labour by 72 hours and in reducing postpartum haemorrhage rates when compared with control. Caesarean birth rates were similar between breast stimulation and intravenous oxytocin. [EL = 1++] One small RCT reported three perinatal deaths in the breast stimulation group and one in the oxytocin group.

Interpretation of evidence

There is evidence that breast stimulation may be effective as a method of induction. However, interpretation of the results was problematic owing to the poor quality of the studies reviewed and the heterogeneous populations, including high-risk women from developing countries. There is inconsistency in the timing, methods and frequency of breast stimulation described in these studies, making guidance on this method difficult. The GDG made a research recommendation.

Research recommendation on breast stimulation

Further research is required to evaluate the effectiveness, timing, methods, frequency, safety and maternal satisfaction of breast stimulation as a method of induction of labour.

5.3. Surgical methods

Clinical question

5.3.1. Amniotomy

Amniotomy is the deliberate artificial rupture of the membranes, used for induction of labour. The procedure is only possible if the membranes are physically accessible.

Overview of available evidence

One systematic review was identified.

One systematic review (one RCT involving 260 women, Bishop score = 6, mixed parity; and one quasi-RCT, 20 women, Bishop score = 4) evaluated the effects of amniotomy in induction of labour in women near term. There were very limited data available for women with an unfavourable cervix. For women with a favourable cervix, data were available for the comparisons between amniotomy and vaginal PGE2 (followed by amniotomy 4 hours later), which reported a significant increased likelihood of oxytocin augmentation in the amniotomy group (RR 2.85, 95% CI 1.82 to 4.46; one RCT). Other maternal and fetal outcomes were comparable.170 [EL = 1++]

Evidence statements

For women with an unfavourable cervix, there is limited evidence to determine the effects of amniotomy alone as an effective method of induction. [EL = 1++]

For women with a favourable cervix, one trial found that amniotomy was significantly associated with oxytocin augmentation when compared with vaginal PGE2. [EL = 1++]

Interpretation of evidence

Although there is limited evidence for amniotomy when the cervix is unfavourable, the practice is not recommended because of the invasiveness of the procedure and the potential risks of infection when amniotomy is performed at the start of labour.

In the case of a favourable cervix, although amniotomy appears to be effective it is associated with more frequent need for oxytocin augmentation when compared with vaginal PGE2.

Recommendations on amniotomy

Amniotomy alone should not be used as a primary method of induction of labour unless there are specific clinical reasons for not using vaginal PGE2, in particular the risk of uterine hyperstimulation.

5.3.2. Mechanical methods

Clinical question

Mechanical methods used for induction of labour include various types of balloon catheters or laminaria tents introduced into the cervical canal or into the extra-amniotic space.

Overview of available evidence

One systematic review and several additional RCTs were identified.

One systematic review (45 RCTs involving 2385 women, Bishop score 0–9, mixed parity) compared mechanical methods versus placebo/no treatment; versus vaginal or cervical PGE2; and versus misoprostol and oxytocin. The various types of mechanical methods were also compared: laminaria tents, balloon catheters and extra-amniotic infusion versus placebo/no treatment, any prostaglandins and oxytocin.171 [EL = 1++]

The GDG considered that balloon catheters are the most commonly used method in the UK and laminaria tents are sometimes used in some other European countries. Studies from this review that compared effects of balloon catheter insertion or laminaria tents with all routes of prostaglandins were thus included. The GDG, however, considered that intracervical prostaglandins are rarely, if ever, used in the UK.

For women with an unfavourable cervix, induction of labour with balloon catheter or vaginal prostaglandins and catheter versus intracervical prostaglandins achieved comparable maternal and fetal outcomes. Balloon catheters were associated with less uterine hyperstimulation with FHR changes (RR 0.04, 95% CI 0.00 to 0.67; one RCT) when compared with vaginal misoprostol 50 micrograms. Laminaria tents were less likely than vaginal prostaglandins to cause uterine hyperstimulation without FHR changes (RR 0.22, 95% CI 0.02 to 0.49; two RCTs). There were no significant differences in maternal or fetal outcomes between induction with laminaria tent and with intracervical prostaglandins. In this review, there were no data available on women with a favourable cervix.171 [EL = 1++]

Two additional RCTs were identified that compared intracervical balloon catheter insertion versus intravaginal misoprostol. One RCT found that catheter insertion was less effective than vaginal misoprostol 100 micrograms in achieving vaginal birth within 24 hours. There were two cases of uterine rupture in the misoprostol group.172 [EL = 1+] The other RCT found that catheter insertion resulted in less tachysystole and uterine hyperstimulation when compared with vaginal misoprostol and combination misoprostol–catheter.173 [EL = 1+]

Evidence statements

For women with an unfavourable cervix, there is limited evidence to assess the effectiveness of intracervical/extra-amniotic balloon catheter or laminaria tent in terms of likelihood of vaginal birth within 24 hours, or a reduction in caesarean births when compared with all routes of prostaglandins, including misoprostol. The likelihood of uterine hyperstimulation may be reduced. [EL = 1++]

Compared with intracervical balloon catheter insertion, intravaginal misoprostol 100 micrograms may be more effective as a cervical priming agent. This dosage is higher than is usually advocated and may explain the two cases of uterine rupture. [EL = 1+]

Intracervical Foley catheter, intravaginal misoprostol and a combination of Foley–misoprostol are comparable for pre-induction cervical priming. [EL = 1+]

For women with a favourable cervix, there was no available evidence to determine the effects of mechanical methods as an agent of induction of labour. [EL = 1++]

Interpretation of evidence

The evidence for the use of mechanical methods for inducing labour in women with an unfavourable cervix is confused by a large number of small studies using different comparators and protocols. When compared with all prostaglandins given by any route, mechanical methods do not improve the rate of vaginal birth within 24 hours nor do they reduce the caesarean birth rate. They may reduce the incidence of uterine hyperstimulation but increase the risk of neonatal infection. The value of mechanical methods of inducing labour in women with an unfavourable cervix is doubtful. Since these methods are associated with less hypertonicity, they may reduce the risk of uterine rupture in the presence of a previous caesarean section scar.

For women with a favourable cervix, there was no available evidence to determine the effects of mechanical methods as an induction agent.

Recommendation on mechanical methods

Mechanical procedures (balloon catheters and laminaria tents) should not be used routinely for induction of labour.

Research recommendation on mechanical methods

Future trials on the use of mechanical methods should include women in whom prostaglandins during labour would pose increased risks, such as women with previous caesarean birth. These trials should clearly stratify groups by parity, cervical status and previous vaginal birth.

Copyright © 2008, National Collaborating Centre for Women's and Children's Health.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK [www.cla.co.uk]. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Cover of Induction of Labour
Induction of Labour.
NICE Clinical Guidelines, No. 70.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2008 Jul.

Download

NICE (National Institute for Health and Care Excellence)

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...