Treatment/prevention of prolonged pregnancy

Cervical stretch and sweep

Membrane sweeping involves inserting a finger through the cervix and sweeping it round to strip the membranes from the decidua. This causes local release of prostaglandins which sometimes is suffi­cient to induce labour (15).

Coitus

It is widely believed that penetrative sex with intravaginal ejacula­tion induces labour. Several plausible mechanisms have been pos­tulated: mechanical pressure on the lower uterine segment, cervical ripening induced by prostaglandin- rich semen, and oxytocin release secondary to orgasm, and observational studies support the idea. The latter suggest that sexual intercourse at term reduces postdates preg­nancy (adjusted odds ratio (AOR) 0.28) and the need for postdates induction (AOR 0.28) (17). A rather outdated Cochrane review pub­lished in 2001, which included only one trial of 28 women found no differences in delivery within 3 days but was too small to rule out clinically important effects (18).

Induction

Indications for induction

Women with uncomplicated pregnancies are generally given every opportunity to go into spontaneous labour. However, when com­plications arise either for the mother or the baby whereby de­livery is deemed favourable (to either the mother, baby, or both) to continuing the pregnancy then induction of labour is indicated. Examples of indications for induction include prolonged pregnancy, prelabour rupture of membranes (at term and preterm), fetal growth restriction, fetal macrosomia, maternal medical disorders (e.g. pre-existing or gestational diabetes or hypertensive disease), intra­uterine fetal death, multiple gestations, advanced maternal age, and even maternal request.

Timing of induction

It is a widely accepted practice in the United Kingdom to offer in­duction of labour for a pregnancy that continues beyond 41 weeks. This is because the rate of stillbirth increases sixfold from 0.35 per 1000 ongoing pregnancies at 37 weeks to 2.12 per 1000 ongoing pregnancies at 43 weeks (19). The rate of neonatal (up to 28 days) and postneonatal (from 28 days to 1 year of age) mortality falls sig­nificantly with advancing gestation up until 41 weeks, when it plat­eaus and then increases with prolonged pregnancy. As such, the overall risk of pregnancy loss (stillbirth plus death occurring up to the age of 1 year) increases eightfold between 37 weeks and 43 weeks, justifying induction of labour at 41 weeks.

Methods of induction

Traditionally, methods of labour induction have been divided into those which include amniotomy, and therefore irreversibly commit the woman to delivery within a few days, and those which do not, such as vaginal prostaglandins and mechanical methods such as cer­vical balloons. Many evaluations of the latter describe them as cer­vical ripening agents for use prior to the definitive labour induction.

Although this is a rather artificial distinction, in that many women labour and deliver after prostaglandins alone, we retain the distinc­tion in what follows.

Prostaglandins

Prostaglandins aid cervical ripening and myometrial contractility and can therefore be used to induce labour. They can be given by a variety of routes: oral, vaginal, intracervical, extra-amniotic, or intravenously. Vaginal preparations of prostaglandins may be in a tablet, gel, or a sustained-release pessary form. Two prostaglandins are in common use: PGE2 (dinoprostone) and PGEi (misoprostol).

PGE2 is the prostaglandin most commonly used. A systematic re­view comparing the use of vaginal preparations of PGE2 with pla­cebo or no treatment for women with an unfavourable cervix, found an increase in successful vaginal delivery rates within 24 hours, an improvement in cervical status within 24 hours, and a reduction in the need for oxytocin augmentation in the treatment group (20).

PGEi has also been used for labour induction, although gener­ally ‘off label’. This is because until recently no formulations licensed for this indication were available. Instead, drugs licensed for treated gastric ulcer disease, such as Cytotec, were used. Recently, a con- trolled-release misoprostol formulation has been approved in the European Union and United States. Systematic reviews suggest that misoprostol may have a slightly better efficacy/side effect profile than dinoprostone.

Mechanical methods

Various instruments have been used to mechanically dilate the cervix to allow amniotomy to be performed. Hygroscopic dilators are placed in the cervical canal and work by slowly absorbing water and increasing in size and shape to dilate the cervix mechanically. Foley catheters have also been used.

Hygroscopic dilators Laminaria tents, made from sterile seaweed or synthetic materials (Dilapan), have been compared to prostaglan­dins for induction of labour and have similar caesarean section rates (RR 1.11; 95% confidence interval (CI) 0.92-1.32) but lower rates of uterine hyperstimulation with fetal heart rate changes (RR 0.13; 95% CI 0.04-0.48).

Balloon catheters Balloon catheters may be used in two ways: (a) inserted through the cervical canal, the balloon is inflated and then traction may or may not be applied, or (b) inserted into the extra- amniotic space and used to instil saline or prostaglandins.

Twenty-three studies comparing balloon catheters with prosta­glandins for induction of labour showed similar results to hygro­scopic dilators; there were similar rates of caesarean section (RR 1.01; 95% CI 0.90-1.13) and lower rates of uterine hyperstimulation (RR 0.19; 95% CI 0.08-0.43) associated with balloon catheters.

The use of a balloon catheter to infuse saline into the extra- amniotic space when compared to prostaglandins had similar rates of both caesarean section (RR 1.37; 95% CI 0.90-2.08) and uterine hyperstimulation with fetal heart rate changes (RR 0.66; 95% CI 0.30-1.46).

Amniotomy Rupturing the amniotic membrane is performed using an amniotomy hook during a vaginal examination. As mentioned previously, it is more effective as an adjunct in inducing labour when performed following the administration of vaginal prostaglandins. Women with a favourable cervix who have an amniotomy are more likely to require oxytocin augmentation than women who have an amniotomy performed following vaginal prostaglandins. A Drew- Smythe’s catheter was used historically to rupture the hind waters in women at high risk of cord prolapse, although concerns about fetal, placental and maternal trauma have rendered it a tool of the past.

Oxytocin

Oxytocin is a hormone secreted by the hypothalamus, then trans­ported to the posterior pituitary from where it is released in a pul­satile manner and acts on oxytocin receptors in the myometrium, causing uterine contractions (21). The frequency and force of con­tractions is directly proportional to the plasma oxytocin concen­tration. Oxytocin is given as an intravenous infusion and slowly increased incrementally to achieve three to four contractions in every 10-minute interval. Oxytocin has been used alone, in com­bination with amniotomy, and following cervical ripening with pro staglandins as an induction agent. One systematic review has compared the effectiveness of oxytocin, when used as a primary agent for induction of labour, in comparison to vaginal prostaglan­dins. This has shown that when oxytocin is used alone in women with an unfavourable cervix and intact membranes, those women are more likely to have no change to their cervical status after 24 hours, and have a higher risk of needing a caesarean section than women for whom vaginal prostaglandins are used (22). There is in­sufficient evidence to reach a conclusion whether oxytocin used in combination with amniotomy or vaginal prostaglandins are more likely to result in vaginal delivery within 24 hours (23) but the use of amniotomy and oxytocin in women with a favourable cervix versus vaginal prostaglandins is associated with a significantly increased risk of postpartum haemorrhage (16% vs 2%) and maternal dissat­isfaction (RR 53).

Complications of induction

Maternal

Induction of labour is associated with potential complications for the mother and her fetus. Risks to the mother include opera­tive intervention, failure to induce labour, cord prolapse during an amniotomy with a high head, placental abruption with rapid decompression of the uterine cavity at amniotomy, and uterine hyperstimulation. Fortunately, these complications are rare.

Operative intervention It is a commonly held belief that induc­tion of labour results in an increased risk of caesarean delivery. Indeed, examining maternity statistics for England 2010-2011, of those women who laboured spontaneously, 75% achieved a spon­taneous vaginal delivery, 12% had an instrumental delivery, and 11% had a caesarean delivery. Of those women who were induced, 59% achieved a spontaneous vaginal delivery, 17% had an instrumental delivery, and 23% had a caesarean delivery. However, it is unscien­tific to compare these two groups of women, when the reasons for induction of labour (prolonged pregnancy, fetal growth restriction, reduced fetal movements) will all contribute to an increased risk of caesarean delivery.

There is a growing body of evidence that induction of labour at term does not increase emergency caesarean section rates, and does not increase intrapartum deaths.

Cord prolapse Umbilical cord prolapse (UCP) is a rare obstet­rical emergency, complicating 1.25-2.1 in 1000 deliveries (24, 25). In one retrospective study of 57 cases over a 10-year period, cord prolapse occurred with amniotomy in 42% of cases (24). However, does amniotomy increase the risk of cord prolapse? A retrospective case-control study of 37 cases of intrapartum UCP and 74 matched control patients with intact membranes found no statistically sig­nificant increase in the use of amniotomy in patients who had a cord prolapse (26). A larger retrospective case-control study in which 80 cases of UCP were matched with 800 controls, found that although 63% of the cases of UCP followed amniotomy, in 87% of the con­trol group amniotomy was also performed (27).

Uterine hyperstimulation Uterine hyperstimulation is generally defined as contractions occurring more than five in 10 minutes or contractions lasting more than 2 minutes. It is a complication that arises in 1-5% of cases where pharmacological agents are used to induce labour (15). It may occur with or without fetal heart rate changes. It can also occur in spontaneous labour. During a uterine contraction there is an interruption in the blood flow to the intervillous space where oxygen exchange between the mother and the fetus occurs (28). During the relaxation phase, blood flow is restored. If the interval between contractions reduces or the dur­ation of a contraction increases, then a critical level may be exceeded where fetal hypoxaemia ensues. Simpson et al. found that uterine hyperstimulation was associated with significant fetal oxygen desat­uration and non-reassuring fetal heart rate changes (28).

Failed induction of labour There is a lack of consensus in the literature as to what constitutes a failed induction of labour (29). While some authors consider a ‘failed induction' to be an induction which doesn't result in a vaginal birth, most consider a failed in­duction to result either when there is a failure to achieve the active phase of labour following a set period of oxytocin administration, following artificial rupture of the membranes, or an unfavourable cervix despite prostaglandin regimens making artificial rupture of the membranes impossible. NICE defines a failed induction of la­bour as ‘failure to establish labour after one cycle of treatment, con­sisting of the insertion of two vaginal prostaglandin tablets or gel at 6-hourly intervals or one prostaglandin controlled released pessary over 24 hours' (15).

A recently published randomized controlled trial comparing the efficacy of prostaglandin tablets versus gel found a rate of failed induction (defined as failure of sufficient ripening of the cervix to allow amniotomy following the use of repeated doses of prosta­glandin leading to delivery by caesarean section) in primiparous women of 2.78% with gel and 20% with tablets and in multiparous women a rate of 0% with gel and a 16% with tablets (30).

Rouse et al. set out to determine the duration of oxytocin admin­istration following artificial rupture of the membranes, at which if active labour was not achieved, vaginal delivery became unlikely (31). They found that at 6 hours post amniotomy and oxytocin, 14% of nulliparous women remained in the latent phase, of whom 39% went on to deliver vaginally. At 9 hours, 7% remained in the latent phase, of whom 28% went on to deliver vaginally and at 12 hours, 4% remained in the latent phase, of whom 13% then delivered vagi- nally. Only 4% of nulliparous women had a caesarean section for a failed induction of labour (i.e. lack of cervical dilatation in the latent phase).

Duration There is a paucity of evidence on the duration of active labour in spontaneous versus induced labour. In a retrospective observational study of low-risk women comparing approximately 10,000 women who laboured spontaneously with 1000 women who underwent labour induction for no apparent medical indication, in­duction was not associated with a prolonged labour. Induction was, however, associated with a longer admission-to-delivery interval and maternal total length of stay was 0.34 days longer with induc­tion compared to spontaneous labour (32).

The findings were similar in a retrospective study of 2681 low-risk multiparous women, where women who were induced had a signifi­cantly shorter labour than those who laboured spontaneously (99 vs 161 minutes) (33).

Pain There is evidence that induced labour is more painful than spontaneous labour. United Kingdom data on analgesia in labour reveals that women who have an induced labour are twice as likely to request epidural anaesthesia as women in spontaneous labour (23% vs 11%, NHS Maternity Statistics 2005-2006). A small study (n = 61) by Capogna et al. found that the minimum effective anal­gesic dosage of sufentanil given via an epidural for women with an induced labour was 1.3 times greater than in women with a spontan­eous onset of labour (P = 0.0014) (34).

Fetal outcomes

Iatrogenic prematurity Most elective caesarean sections are per­formed at 39 weeks' gestation (35). The timing of this is advised as the risk of neonatal respiratory morbidity falls with advancing ges­tation until 40 weeks. The risk of respiratory morbidity in infants delivered by elective caesarean at 37 weeks is fourfold higher than infants delivered by mothers intended to have a vaginal delivery at 40 weeks, threefold higher at 38 weeks, and twofold higher at 39 weeks. The risk of developing neonatal respiratory symptoms for ba­bies born by vaginal delivery falls from a probability of 0.07 at 37 weeks to 0.04 at 39 weeks and thereafter plateaus (36). Thus, induc­tion of labour at 39 weeks is the optimal balance between the risk of respiratory morbidity for the neonate and the risk of antepartum stillbirth for the fetus.

Hyperbilirubinaemia There is a reported association be­tween the use of oxytocin in labour and prolonged unconjugated hyperbilirubinaemia in neonates. In a prospective study in Northern India, the use of oxytocin in labour was found to increase the risk of prolonged jaundice in neonates (odds ratio 3.4) (37). In all infants the jaundice disappeared by 8 weeks. However, a randomized con­trolled trial of the elective use of oxytocin infusion for women using epidural analgesia found no increased risk of neonatal jaundice in the treatment group (38).

Randomized controlled trials of induction

Numerous large, multicentre randomized controlled trials com­paring induction of labour with expectant management at term for prelabour rupture of membranes (39), prolonged pregnancies (40), gestational hypertension or mild pre-eclampsia (41), intrauterine growth restriction (42), advanced maternal age (43), and fetal macrosomia (44) found no difference in caesarean section rates be­tween the two groups.

A recent trial of induction of labour at term for women identi­fied as high risk for emergency caesarean section (the higher the risk score, the earlier the induction), found in the treatment group a similar caesarean section rate, a higher uncomplicated vaginal birth rate, and a reduced neonatal intensive care unit admission and ad­verse outcome rate (45).

Wood et al. published a systematic review of induction of labour versus expectant management in women with intact membranes at term and found that a policy of induction of labour was associated with a 17% reduction in the risk of caesarean section (odds ratio 0.83; 95% CI 0.76-0.92) (46). Two further systematic reviews have confirmed Wood et al.'s findings. Unfortunately, none of the many trials of labour induction around term have followed up either the babies or the mothers longer than the point of discharge from hospital.