Induction of Labour: Types, Reasons, Risks and What the Evidence Says
A brief history
People have tried to start labour for centuries. Historical accounts describe manual and mechanical methods dating back to Hippocrates, including mammary stimulation and cervical dilation. Modern pharmacological induction began in the 20th century:
1900s: Pituitary extracts were used to stimulate contractions, though dosing was unreliable and risked uterine rupture.
1953: Vincent du Vigneaud synthesised oxytocin, which became the standard synthetic drug for induction.
1967: Embrey and Mollison first described placing a balloon catheter through the cervix to ripen it.
1968: Onwards: Prostaglandins were introduced for induction, and research into misoprostol grew through the 1990s.
Today, 30% of deliveries in England begin with induction, up from 27% a decade earlier — spontaneous onset has fallen from 59% to 42% over the same period.
Types of induction
Membrane sweep
A midwife or doctor sweeps a finger around the cervix to separate the membranes from the lower uterus, releasing natural prostaglandins. A Cochrane review found sweeping increases the chance of spontaneous labour and reduces the need for formal induction, with a relative risk of 1.21 for spontaneous onset of labour and 0.73 for needing further induction. It can be uncomfortable and may cause spotting. NICE recommends offering it before formal induction.
Prostaglandins (dinoprostone or misoprostol)
These soften and ripen the cervix, given as a vaginal tablet, gel, pessary, or oral tablets. NICE lists these as first-line options when the cervix is unfavourable. The main risk is uterine hyperstimulation, which can affect the baby's heart rate.
Mechanical methods (balloon catheter)
A small balloon is placed through the cervix and gently inflated. Cochrane evidence suggests caesarean rates are broadly similar to prostaglandins, with a better safety profile; a 2023 update found the balloon may be slightly less effective than oral misoprostol. It's often used when prostaglandins are less suitable, such as after a previous caesarean.
Breaking the waters (ARM) and oxytocin
Once the cervix is favourable, the membranes can be released and/or an oxytocin drip started. It requires continuous fetal monitoring and often limits mobility, with a small risk of cord prolapse.
The ‘ Natural’ Way
Breast/nipple stimulation
Breast and nipple stimulation is considered a natural way to encourage labour because it triggers the release of oxytocin, the hormone that causes uterine contractions. It's typically only suggested for low-risk, full-term pregnancies, and it's best to discuss it with your midwife or doctor first, since it can occasionally cause overly strong or frequent contractions.
It showed a reduction in women not in labour at 72 hours, but a Cochrane review found the evidence too limited to confirm its safety.
Sexual Intercourse
The evidence for sexual intercourse as an induction method is too limited to draw conclusions, based on just one small trial. However, semen contains prostaglandins - the same hormone-like compound used in medical cervical-ripening gels. This is the theory behind why sex is sometimes suggested as a way to encourage labour to start.
Acupuncture & Acupressure
For induction, acupuncture and acupressure are based on the traditional theory that stimulating specific points — such as SP6 above the ankle or LI4 in the hand — can trigger the release of oxytocin or prostaglandins, though a Cochrane review found the clinical evidence for this actually starting labour remains limited
Caster Oil
Castor oil is a traditional labour-starter thought to work by irritating the bowel and stimulating the uterus, but a Cochrane review of three trials involving 233 women found there hasn't been enough research to show whether it ripens the cervix or induces labour, and all the women who took it by mouth felt nauseous. The trials were small and of poor methodological quality, so the results should be interpreted with caution.
Raspberry Leaf Tea
Raspberry leaf tea is traditionally described as a uterine "tonic" — believed to help the uterus contract more efficiently once labour begins, rather than to trigger it, via a compound called fragarine. However, the main randomised controlled trial testing this found no significant effect on the length or outcome of labour, though it did confirm the tea appears safe when started in the third trimester.
When induction may be offered
Prolonged pregnancy: NICE recommends offering induction between 41+0 and 42+0 weeks, as the risk of stillbirth and complications rises gradually the further a pregnancy continues past term. It's worth remembering that your estimated due date (EDD) is just that — an estimate. Even with a dating scan, EDDs can be off by several days depending on when ovulation actually happened and how the scan measurements fall on the growth charts used, so "41 weeks" isn't always as precise as it sounds on paper. Some of the research behind this guidance, including the Swedish SWEPIS trial, has also been criticised for being stopped early and for how the results were interpreted — it's a conversation worth having with your midwife rather than treating the dates as absolute.
Waters breaking before labour at term: Once the waters break, the protective barrier between the baby and the outside world is gone, which creates a small but increasing risk of infection the longer labour is delayed — this is the main reason it's discussed at all, rather than any danger from the waters breaking itself. Both induction and expectant management (waiting, typically up to around four days) are considered reasonable options, as research has found similar infection and caesarean rates either way.
Hypertension or pre-eclampsia: Induction may be offered earlier, as these conditions can affect how well the placenta is working, and research suggests earlier birth can reduce complications for both mother and baby.
Diabetes: NICE advises offering birth by 40+6 weeks for gestational diabetes, and a little earlier for pre-existing diabetes, due to the increased risk of the baby growing large or the placenta functioning less well later in pregnancy.
Elective induction at 39 weeks: Some units now offer induction at 39 weeks for first-time mothers, even without a medical reason, as some research has suggested this may be linked to lower caesarean rates. This isn't routine policy everywhere, and it's a personal choice rather than something to feel pressured into.
It's worth noting that a lower caesarean rate doesn't mean induction is risk-free. Induction still carries its own considerations — including a higher chance of uterine hyperstimulation (contractions that come too close together), a longer and often more intense labour, increased use of epidural pain relief, and reduced mobility due to continuous monitoring.
NICE's own evidence review acknowledges these trade-offs when comparing induction against waiting for labour to start naturally, which is why the decision should weigh your full circumstances rather than caesarean rate alone.
When induction may be avoided
| Situation | Considerations |
|---|---|
| Placenta praevia | Vaginal birth usually unsafe |
| Vasa praevia | Vaginal birth usually unsafe |
| Transverse or oblique lie | Vaginal birth not possible |
| Cord presentation or prolapse | Emergency situation (induction with prostaglandins itself is not linked to cord prolapse) |
| Previous classical caesarean or uterine rupture | High rupture risk |
| Active primary genital herpes | Caesarean usually advised if infection occurs within 6 weeks of birth |
| Previous caesarean (lower segment) | Possible, but rupture risk rises, especially with prostaglandins |
| Breech baby | Usually caesarean or ECV first |
| Grand multiparity | Higher risk of hyperstimulation and rupture |
| Abnormal fetal heart trace | Needs assessment first |
On rupture risk: in a large US population study of women with a prior caesarean, uterine rupture occurred in 5.2 per 1,000 women with spontaneous labour, 7.7 per 1,000 with induction using oxytocin alone, and 24.5 per 1,000 with induction using prostaglandins. A separate, larger observational study found an overall symptomatic rupture rate of 0.7% across all methods of labour after a prior caesarean.
Sources:
Grand multiparity - Evidence is mixed — some studies link it to higher rupture risk with induction, especially in an unscarred uterus; risk appears less clear-cut after a previous caesarean
Weighing it up
Induction can be lifesaving. However, it can also be an unnecessary intervention, so the reason matters. Compared with spontaneous labour, induction often means a hospital setting, more monitoring, a more intense labour, and higher use of epidurals.
Helpful questions to ask a care provider:
Benefits: What is the reason, and how strong is the evidence for my situation?
Risks: What are the risks to me and my baby?
Alternatives: Could we wait, monitor, or try a sweep first?
Intuition: What does my gut say?
Nothing: What happens if we do nothing for a day or two?
The decision is always yours, and you can decline or delay induction.
** This article is for education purposes and does not replace individual medical advice.**