Home > Coping with pain in labour: non-epidural...

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). Intrapartum Care: Care of Healthy Women and Their Babies During Childbirth. London: RCOG Press; 2007 Sep. (NICE Clinical Guidelines, No. 55.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

5Coping with pain in labour: non-epidural

5.1. Introduction

A woman’s desire for, and choice of, pain relief during labour are influenced by many factors, including her expectations, the complexity of her labour and the severity of her pain. To many women the pain of labour is significant and the majority require some form of pain relief. Flexible expectations and being prepared for labour may influence her psychological wellbeing after birth. Extreme pain can result in psychological trauma for some women, while for others undesirable side effects of analgesia can be detrimental to the birth experience. Effective forms of pain relief are not necessarily associated with greater satisfaction with the birth experience and, conversely, failure of a chosen method can lead to dissatisfaction.

There are two schools of thought around how women might cope with the pain of labour. The first suggests that in the 21st century there is no need to suffer unnecessarily during labour and that effective analgesia is available and should be offered. The second sees pain as part of the experience of birth and advocates that women should be supported and encouraged to ‘work with the pain’ of labour.

While individual women or carers may identify with either view, the reality for most women is probably somewhere between these two. The challenge for midwives and healthcare professionals is not only to identify where that individual woman lies on the continuum, but also, through good communication, to recognise and respond appropriately to changes in the woman’s stance during labour.

Whatever the woman’s viewpoint, it is fundamental that she should be treated with respect and as an individual. Women need to be in control of, and involved in, what is happening to them and the manner in which they are supported is key to this. Continuing communication between woman and midwife during the progress of labour about her desire for analgesia is also fundamental, as is the recognition of severe distress.

Clinical question

What is the effectiveness of the following interventions or techniques in labour on outcomes?

  • breathing and relaxation
  • massage
  • complementary therapies
  • birth balls
  • injected water papules
  • water (including temperature regulation).

5.2. Women’s views and experiences of pain and pain relief in childbirth

Description of included studies

This systematic review was undertaken to specifically address the outcome of women’s views of pain relief and the experience of childbirth in relation to intrapartum analgesia. The included studies involve women in labour at term, entering labour without complications. Outcomes include women’s views of pain relief and the overall experience of childbirth (including satisfaction with the childbirth experience).

Review findings

A systematic review of 137 reports of pain and women’s satisfaction with childbirth was identified for inclusion.67 [EL = 2++] The review includes descriptive studies, randomised controlled trials and systematic reviews of intrapartum interventions. Findings were summarised qualitatively. Thirty-five reports of 29 studies met the inclusion criteria for observational studies of childbirth satisfaction. Sample sizes ranged from 16 to 2000 and more than 14 000 women from nine countries were studied. Thirteen reports of five systematic reviews and seven randomised controlled trials were also included. More than 27 000 women were included and the methodology of the studies was generally very good. One systematic review and 20 RCTs met the inclusion criteria for studies of intrapartum pain relief that included a measure of satisfaction as an outcome. The most common method of assessment of satisfaction was a single VAS score, usually made in the immediate postnatal period. The methodological quality of these studies was quite good with generally small sample sizes. The author illustrates the complexity of the relationship between pain, pain relief and women’s experiences of childbirth with findings from two population-based surveys (one UK (n = 1150) and one Australian (n = 1336)). The UK survey found that women who were very anxious about labour pain antenatally were less satisfied after the birth. The most satisfied women postnatally were those who had used no pain-relieving drugs during labour. All effects were independent of parity or demographic variables. In the Australian survey, the odds of dissatisfaction were greater when women rated their caregivers as less than very helpful and when women felt they were not actively involved in decision making. The impact on dissatisfaction was greater than that for ratings of pain relief as unsatisfactory. It is also noted that women’s views of pain and of pain relief are not the same thing. In 11 of the 21 trials reported in the review, discrepancies were noted between the ratings of pain compared with ratings of pain relief. Synthesis of evidence from all the reviewed papers led to the conclusion that four factors exist:

  • personal expectations
  • the amount of support from caregivers
  • the quality of the caregiver–patient relationship
  • the involvement in decision making.

These factors appear to be so important that they override the influence of age, socio-economic status, ethnicity, childbirth preparation, the physical birth environment, pain, immobility, medical interventions and continuity of care when women evaluate their childbirth experience. The author concluded that the influences of pain, pain relief and intrapartum interventions on subsequent satisfaction are neither as obvious, nor as powerful, as the influences of the attitudes and behaviours of the caregivers.

One RCT was identified that investigated nulliparous women’s satisfaction with childbirth and intrapartum pain relief when labouring at term.122 [EL = 1+] The study conducted in Australia compared epidural and non-epidural analgesia findings and, therefore, are from within the context of a trial. Women were ‘surveyed’ (presumably by questionnaire, but this is not explicit) approximately 24 hours postnatally and again 6 months postpartum by a mailed questionnaire. Women recruited into the study were randomised into one of two groups – the EPI group who were encouraged to have an epidural as their primary pain relief (n = 493), and the CMS group who received one-to-one continuous midwifery support throughout labour and were encouraged to avoid epidural analgesia but instead use Entonox, intramuscular (IM) pethidine and non-pharmacological pain relief (n = 499). There was a high crossover rate within the study: 61.3% women crossed over from the CMS group to the EPI group (n = 306) and 27.8% crossed over from the EPI group to the CMS group (n = 137). Analysis was undertaken on an intention-to-treat basis. Women allocated to the EPI group were significantly more satisfied with their intrapartum pain relief, and the reported pain intensity post-administration was significantly lower for this group. Both groups reported similar and high levels of satisfaction with a degree of midwifery support during labour (median [interquartile range], P value obtained using Wilcoxon rank sum test) (CMS 95 mm [IQR 88 to 100] versus EPI 96 mm [IQR 90 to 100], P = 0.24); participation in intrapartum decision making (CMS 5 [IQR 4 to 5] versus EPI 5 [IQR 4 to 5], P = 0.35); achievement of labour expectations (CMS 3 [IQR 2 to 4] versus EPI 3 [IQR 2 to 4], P = 0.32) and achievement of birth expectations (CMS 2 [IQR 2 to 5] versus, EPI 2 [IQR 2 to 5], P = 0.54). Despite the difference in satisfaction with pain relief and levels of pain experienced between the two groups, reports of the overall labour experience and overall birth experience were similar for both groups (labour: CMS 4 [IQR 3 to 4] versus EPI 4 [IQR 3 to 4], P = 0.74, birth: CMS 4 [IQR 4 to 5] versus EPI 4 [IQR 3 to 5], P = 0.60). Findings obtained from the 6 month follow-up questionnaire (n = 642; response rate = 64.7%) showed that women in the CMS group were significantly less likely to plan to use an epidural in a subsequent labour (OR 0.64 [95% CI 0.47 to 0.89]). Despite the high crossover rates and intention-to-treat analysis, the findings from this study are, perhaps, as would be expected, i.e. improved pain relief associated with epidural use. This may be because women allocated to the CMS group delayed requests for an epidural. Unfortunately, this is not discussed in the paper, thus making interpretation of the findings difficult.

A prospective survey undertaken in Finland (n = 1091) sought women’s expectations for intra-partum pain relief antenatally, measured pain intensity during labour and birth, and followed up women’s satisfaction with pain relief on the third day postnatally.123 [EL = 3] Antenatally, 4% of nulliparous women and 14% of multiparous women felt they would not need any analgesia during labour, with 90% of women overall expressing a wish for intrapartum analgesia. Prior to the administration of any analgesia, 89% of nulliparous women and 84% of multiparous women described their pain during labour as either ‘very severe’ or ‘intolerable’. Twenty percent (n = 213) of women, of whom 14% were nulliparous and 86% were multiparous, received no analgesia during labour. The pain scores of these women did not differ significantly from those women who then went on to receive analgesia. After administration of pain relief, 50% of multiparous women and 19% of nulliparous women still reported pain scores of 8–10 on the BS-11. This difference reflects a higher degree of usage of epidural analgesia among the nulliparous women. Eighteen percent of women rated their pain relief as poor, 37% rated it as moderate, and 45% as good. Surprisingly, views of pain relief were not related to parity. Half of all women complained of inadequate pain relief during labour which, in multiparous women, was significantly associated with the second stage of labour. Overall, 95% of women stated that they were satisfied with their care during childbirth. Ratings of overall satisfaction were not related to parity, level of pain experienced or pain relief received. Findings reflect a lack of effective pain relief, particularly for those women who, for whatever reason, do not choose an epidural. Dissatisfaction with childbirth was very low, and was associated with instrumental births, but not with usage of analgesia. Despite an apparent low level of effectiveness of pain relief, most women expressed satisfaction with care during labour. This may reflect low expectations of pain relief in this population and again demonstrates the complexity of the relationships between reported pain, pain relief, satisfaction with pain relief and the experience of childbirth.

One European multicentre study was reviewed which examined nulliparous women’s expectations and experiences of intrapartum analgesia.124 [EL = 3] The study involved over 100 women from each of five countries (Italy, UK, Belgium, Finland and Portugal; total n = 611). All women were interviewed during the last month of pregnancy and again approximately 24 hours postnatally. Expectations of pain, pain relief and satisfaction were assessed using a 10 cm VAS. Findings showed that women who expected higher levels of pain were more likely to be satisfied with analgesia (Spearman’s rho = 0.15, P = 0.001). Women who experienced higher levels of pain following administration of analgesia were less satisfied with pain relief (Spearman’s rho = −0.66, P < 0.0001). Maternal satisfaction with the overall childbirth experience was positively correlated with pain expectations (Spearman’s rho = 0.23, P < 0.001) and pain before analgesia (Spearman’s rho = 0.16, P < 0.001), and negatively correlated with pain after analgesia (Spearman’s rho = −0.30, P < 0.001). The most satisfied women were those who expected more pain, were satisfied with analgesia received and had good pain relief after analgesia. Pain did not correlate with women’s educational level, age or social class. Generally, women’s satisfaction with analgesia and the birth experience was high. It should be noted that all hospitals involved in the study were tertiary centres with above average epidural rates. Other components of the birth experience, e.g. involvement in decision making, friendliness and expertise of staff, were not investigated in this study.

Evidence statement

A woman’s experience of birth vary enormously and is influenced by many factors including her expectations, degree of preparation, the complexity of her labour, and the severity of the pain she experiences.

The attitude and behaviour of the caregiver is consistently seen to be the most obvious and powerful influence on women’s satisfaction. Women are more satisfied with pain relief when their expectations of pain and how they choose to manage it are met.

GDG interpretation of the evidence (advice to clinicians regarding non-epidural pain relief)

This section offers advice to clinicians caring for women in labour regarding non-epidural pain relief, based on the GDG’s work and deliberations.

It is important to remember that relatively simple things can make a difference.

Women appreciate having someone whom they know and trust with them in labour, although there is no high-level evidence on the benefits of this.

Women should be able to play music of their own choice and drink and eat a light diet if they want to during labour.

They can choose to walk, move around, find comfortable positions, sit, stand up, or lie down on their sides. However, if they lie on their backs, they are likely to feel the pain more intensely.

We have prioritised the options for analgesia on the basis of the strength of the evidence of their effectiveness:

  • The evidence shows that immersion in water provides effective pain relief, so encouraging the woman to get into a warm bath or birthing pool will help reduce the pain of the first stage of labour, and mean they are less likely to need an epidural. As far as we know, this does not adversely affect maternal or neonatal outcomes. Using a bath or a birthing pool for pain relief does not mean that the woman has to remain in it for birth unless she wants to. Women can get out of the water at any time if they do not like it or want to try another method of analgesia.
  • Entonox has the advantage that it acts very quickly and rapidly passes out of the system without affecting the baby and it can be used anywhere – even in the bath. It takes the edge off the pain and helps many women. Some women feel dizzy or light-headed when using it but the advantage of Entonox is that if the woman does not like it, it can be stopped and the side effects will also stop.
  • Women who choose to use breathing and relaxation techniques or massage, by their birth partner, should be supported. The little evidence available shows that they may significantly reduce the pain and they do help many women in labour and do not adversely affect either maternal or neonatal outcomes.
  • Women who choose to use acupuncture or hypnosis should be able to, although they are not provided by the maternity unit. The little evidence available shows that they may reduce the pain of labour and do not appear to adversely affect either maternal or neonatal outcomes.
  • Opioids such as pethidine or diamorphine are widely used and the evidence available shows they provide poor analgesia and can make women feel nauseous and drowsy. As pethidine crosses the placenta, it may make the baby sleepy. This means that the baby may suffer respiratory depression at birth and is sleepy and reluctant to feed for several days after birth. Pethidine should always be administered with an anti-emetic. Women can still use the bath or birthing pool as long as they are not drowsy and have not had pethidine in the previous 2 hours.
  • There is no evidence on the effectiveness of birth balls for reducing the pain of labour. They may, however, help women find a comfortable position.
  • We also know that using a transcutaneous electronic nerve stimulation (TENS) machine does not provide any pain relief once the woman is in established labour, and is therefore not recommended at this stage. There are no trials of its use in latent labour when some women choose to use it.
  • All women use some kind of pain-relieving strategies during labour, and many will use several different ones. What is important is that they are able to communicate with you to ensure that as far as possible, they feel in control and confident and that both of you remain flexible about what is wanted.

Recommendation on women’s views and experiences of pain and pain relief

Healthcare professionals should consider how their own values and beliefs inform their attitude to coping with pain in labour and ensure their care supports the woman’s choice.

5.3. Pain-relieving strategies


The evidence regarding pain-relieving strategies is described below and incorporates a wide range of strategies used by many women over the centuries, to help them cope with labour, that do not require professional oversight.

See also Section 4.3.1 on support in labour.

5.3.1. Breathing and relaxation

Description of included studies

One controlled trial of breathing and relaxation techniques was described in a systematic review of complementary therapies used during labour (n = 54 women, but 20 were lost to follow-up).125 [EL = 1−] Women were randomised into an experimental group who received ‘respiratory autogenic training’ (progressive muscle relaxation and focused slow breathing) or a control group who attended a ‘traditional psychoprophylactic course’ (no details are given about the content of this course, but it may also have included a form of relaxation training).

Review findings

Although a significant reduction in reported intrapartum pain was noted for women in the experimental group, this was only found after adjusting for women who were very anxious during pregnancy. Postnatal reports of labour pain and labour experience did not differ significantly between the two groups.

Evidence statement

There is a lack of evidence that breathing and relaxation techniques reduce measured pain in labour or affect any other outcome.

Recommendation on breathing and relaxation

Women who choose to use breathing and relaxation techniques in labour should be supported in their choice.

5.3.2. Touch and massage

Description of included studies

Two systematic reviews were identified which included evaluation of the use of massage or therapeutic touch for pain relief during labour.85,125 [both EL = 1+] Each review included two controlled trials, with a total of three studies included overall: two RCTs and one prospective cohort study. The two RCTs reviewed were fairly small (n = 24 and n = 60) and conducted in the USA and Taiwan, respectively.

Review findings

Differences between the trials prohibit pooling of the data. In both trials the woman’s partner was shown how to carry out massage and this was then performed for set periods of time throughout the first stage of labour (20–30 minutes/hour). In the larger trial, the control group received a ‘casual’ contact with the researcher for the same periods of time, while in the smaller study the control group received ‘usual care’ including guidance on breathing and relaxation techniques. In the larger study it is not clear whether the nurse carrying out the pain assessment was blinded, while in the smaller trial, blinding of the nurse assessor was carried out. Pain was also assessed by the women themselves. Both trials showed a significant reduction in labour pain as reported by the nurse observers and the women. No mention was made of other analgesia used during labour, for women in either group.

In the smaller study, a significant reduction in intrapartum stress and anxiety was reported by both the women and the blinded observer. There was also a significant improvement in maternal mood (self-rated using a depression scale) both during labour and postnatally.

A prospective cohort study, conducted in the USA, examined the effect of therapeutic touch during labour (n = 90). Women in the experimental group received touch from the midwife (e.g. handholding) for a period of 5–10 seconds after each verbal expression of anxiety. The study was carried out during a 30 minute intervention period at the end of the first stage of labour (8–10 cm dilatation). The control group received ‘usual care’. Despite the seemingly short duration of the intervention, maternal anxiety (as measured by blood pressure, verbal expressions of anxiety and anxiety scores reported by mother in the early postnatal period) were found to be reduced significantly (P < 0.05) in the experimental group, compared with the control group.

Evidence statement

The limited available evidence suggests that massage and reassuring touch reduces a woman’s measured pain and expressed anxieties during labour. There is no high-level evidence that birth outcomes are influenced by massage.

Recommendation on touch and massage

Women who choose to use massage techniques in labour that have been taught to birth partners should be supported in their choice.

5.3.3. Labouring in water


The Winterton report recommended that all maternity units should provide women with the option to labour and give birth in water.95 However, the number of women actually using water during labour is not well reported. A survey between April 1994 and March 1996 identified 0.6% of births in England and Wales occurring in water, 9% of which were home births.126 There would appear to be a wide variation in the use of water during birth, with one birth centre reporting up to 80% of women using water during labour and up to 79% giving birth in water.127

Previous guideline

Water birth was reviewed in the NICE Caesarean Section guideline.6 The guideline reviewed one systematic review, one RCT and some other observational studies and recommended that women should be informed that immersion in water during labour has not been shown to influence the likelihood of CS, although it may affect other outcomes.

Description of included studies

There was one systematic review and one RCT identified for inclusion in the review. The systematic review included eight trials.128 [EL = 1+] Out of the eight trials, six examined labouring in water in the first stage, one examined labouring in water in the second stage, and one investigated the timing of the use of water in the first stage of labour. An additional RCT examined effectiveness of use of water in the first stage compared with augmentation.129 [EL = 1−]

There was no relevant study identified that addressed hygiene measures for water birth.

Review findings

Use of water versus other methods

Women’s outcomes

Meta-analysis of findings from four trials reported in the systematic review128 [EL = 1+] showed that the use of water in the first stage of labour reduces the use of epidural/spinal analgesia/anaesthesia (OR 0.84 [95% CI 0.71 to 0.99]). One trial reported significantly reduced reported pain for those women who laboured in water compared with those not labouring in water (OR 0.23 [95% CI 0.08 to 0.63]).

Meta-analysis of four trials in the review showed no evidence of differences on duration of the first and second stages of labour between women who laboured in water and those who did not. Six trials reported on instrumental birth rates and CS. Findings from a meta-analysis of these trials showed that overall there was no evidence of any difference: instrumental vaginal birth rate (OR for use of water 0.83 [95% CI 0.66 to 1.05]) and CS rate (OR for use of water 1.33 [95% CI 0.92 to 1.91]).

There was no evidence of differences on perineal trauma with labouring in water: episiotomy (OR 0.89 [95% CI 0.68 to 1.15]), second-degree tears (OR 0.90 [95% CI 0.66 to 1.23]) or third/fourth-degree tears (OR 1.38 [95% CI 0.85 to 2.24]).128 [EL = 1+]

Newborn outcomes

Five trials reported on Apgar scores at 5 minutes and there was no difference in the number of babies with a score of less than 7 at 5 minutes (OR 1.59 [95% CI 0.63 to 4.01]). Two trials reported admissions to the neonatal unit and found no evidence of difference (OR 1.05 [95% CI 0.68 to 1.61]). Infection rates were reported in four trials and were found to be very low (6/629 versus 3/633; OR 2.01 [95% CI 0.50 to 8.07]).

Timing of use of water

One trial in the systematic review compared early versus late immersion during the first stage of labour, and found significantly higher epidural analgesia rates in the early group (42/100 versus 19/100; OR 3.09 [95% CI 1.63 to 5.84]) and an increased use of augmentation of labour (57/100 versus 30/100; OR 3.09 [95% CI 1.73 to 5.54]).130

Augmentation versus use of water

One trial compared augmentation versus immersion in water during the first stage of labour.129 [EL = 1−] It showed that use of water reduced rate of augmentation (RR 0.74, P = 0.001) and increased some aspects of satisfaction (freedom of movement MD 1.46, P = 0.001; privacy MD 1.18, P = 0.03; satisfaction with the care MD 1.07, P = 0.49). There were more babies admitted to neonatal units with use of water (admission to neonatal unit 6/49 (water), 0/50 (air), P = 0.01), but there is no evidence of a difference on cord arterial pH or infection rate (cord arterial pH 7.26 (water), 7.25 (air), P = 0.97; infection 8/49 (water), 9/50 (air), P = 0.78).

Evidence statement

Labouring in water reduces pain and the use of regional analgesia. There is evidence of no significant differences regarding adverse outcomes when comparing labours with and without the use of water. There is insufficient evidence on timing of use of water in labour.

There is no good-quality evidence regarding hygiene measures for water birth.

Recommendations on labouring in water

The opportunity to labour in water is recommended for pain relief.

For women labouring in water, the temperature of the woman and the water should be monitored hourly to ensure that the woman is comfortable and not becoming pyrexial. The temperature of the water should not be above 37.5 °C.

Any bath or birthing pool should be kept clean using a protocol agreed with the microbiology department and, in the case of birthing pools, in accordance with the manufacturer’s guidelines.

5.3.4. Birth balls

Overview of available evidence

No studies were identified which examined the use of birth balls during labour.

Evidence statement

There is no evidence of any effect of birth balls on birth experience or clinical outcomes.

5.3.5. Injected water papules

Description of included studies

Two systematic reviews were identified, both of which reviewed the same four RCTs examining the effectiveness of cutaneous water injections.85,125 [EL = 1+]

Review findings

The four included trials were of fair to good quality, with sample sizes ranging from 35 to 272. Women in labour reporting back pain or severe back pain were entered into the trials. Three trials described adequate randomisation, three were double-blinded placebo-controlled trials, and three trials were analysed on an intention-to-treat basis. In all cases there were some missing data due to women giving birth before the end point of the trial [range 4% to 30%]. Differences between the trials mean pooling of data is not possible. In all four trials back pain was significantly reduced for 45 to 90 minutes following the intradermal injections of sterile water, as measured by a VAS. In the one trial that compared subcutaneous and intradermal water injections, both were found to be similarly effective compared with the control of subcutaneous saline injections. Despite the pain relief reported, there was no significant difference between experimental and control groups in three of the trials, regarding subsequent use of analgesia. In one trial, use of subsequent analgesia was higher in the experimental group than in the control group where women received massage, baths and were encouraged to mobilise. In this trial, women in the control group were more likely than women in the experimental group to say that they would choose the same pain relief option for a subsequent labour. In the other three trials, women who had received cutaneous water injections were more likely to say they would choose the same option for a future labour. No trial reported the effects of repeated injections.

One of the main disadvantages of this method of pain relief is the intense stinging pain that many women report during the administration of the intradermal injections. An RCT was conducted in Sweden to compare the perceived pain during administration of intradermal versus subcutaneous injections of sterile water.131 [EL = 1+] The work involved 100 healthy women (not pregnant/in labour) in a blind, controlled trial with a crossover design. Perceived pain was measured using a VAS. The findings showed that intradermal injections were reported as being much more painful than subcutaneous injections (mean 60.8 mm versus 41.3 mm, P < 0.001). It is not known, however, whether this finding would apply to women in labour.

Evidence statement

There is a lack of evidence of the benefit of injected water papules on birth experience or clinical outcomes.

Recommendation on injected water papules

The use of injected water papules is not recommended.

5.3.6. Complementary and alternative therapies

Previous guideline

The Caesarean Section guideline reviewed the effectiveness and safety of complementary and alternative therapies for women during labour.6 The guideline included a systematic review comprising seven trials and five observational studies. In the guideline, it was recommended that women should be informed that the effects of complementary therapies used during labour (such as acupuncture, aromatherapy, hypnosis, herbal products, nutritional supplements, homeopathic medicines, and Chinese medicines) on the likelihood of CS have not been properly evaluated, and further research is needed before such interventions can be recommended.

Acupressure and acupuncture

Description of included studies

Four reasonable quality RCTs were identified.132–135 A Korean trial (intervention n = 36; control n = 39) compared SP6 acupressure to controls that received touch at the same point.132 A second trial, conducted in Norway (intervention n = 106; control n = 92), compared a group of women who received acupuncture with a group who did not receive acupuncture or a placebo.135 A third study, also conducted in Norway, compared acupuncture with false acupuncture (intervention n = 106; control n = 102).134 A Swedish study involving 90 women (intervention n = 46; control n = 44) was also identified. The control did not receive any form of placebo.133 While the trial that investigated effectiveness of acupressure in labour reported separately, a new meta-analysis was conducted using these three trials on acupuncture, as they are considered to have reasonable homogeneity. [EL = 1+]

Review findings

There was evidence of reduction in pain score after SP6 acupressure compared with SP6 touch (WMD −1.20 [95% CI −2.04 to −0.36]), but no evidence of difference in use of pharmacological pain relief (RR 0.54 [95% CI 0.20 to 1.43]).

Meta-analysis of the RCTs showed that acupuncture significantly reduced the use of pharmacological pain relief (two trials RR 0.74 [95% CI 0.63 to 0.86]), epidural analgesia (two trials RR 0.45 [95% CI 0.29 to 0.69]) and the need for augmentation of labour with oxytocin (two trials RR 0.58 [95% CI 0.40 to 0.86]). There was no evidence of differences in pain score after acupuncture (one trial MD −0.20 [95% CI −0.80 to 0.40]) or rate of spontaneous vaginal birth (three trials RR 1.03 [95% CI 0.97 to 1.09]). Outcomes such as maternal satisfaction and maternal and neonatal complications were not investigated.


Description of included studies

A systematic review, published in 2004, involving five RCTs and 14 comparative studies was identified, but only the evidence from the RCTs has been included here.136 All the RCTs were conducted in either the UK or the USA. [EL = 1+]

Review findings

Meta-analysis of the RCTs showed that hypnosis significantly reduced the use of pharmacological pain relief (three trials RR 0.51 [95% CI 0.28 to 0.98]) and of the need for labour augmentation (two trials RR 0.31 [95% CI 0.18 to 0.52]). No other outcomes were considered.


Description of included studies

A systematic review involving one RCT in New Zealand was identified.137 The study population comprised 22 multiparous women. Women in the intervention group received essential oil of ginger or essential oil of lemongrass in the bath, and they were required to bathe for at least 1 hour. [EL = 1+]

Review findings

There was no evidence of a difference in the use of pharmacological pain relief (RR 2.50 [95% CI 0.31 to 20.45]), rates of spontaneous vaginal birth (RR 0.93 [95% CI 0.67 to 1.28]), instrumental birth (RR 0.83 [95% CI 0.06 to 11.70]) or CS (RR 2.54 [95% CI 0.11 to 56.25]). There were no other outcomes investigated.


Description of included studies

One RCT published in 2003 involving 110 women in labour (intervention n = 55; control n = 55) provides the evidence for this subsection.138 Women in the intervention group listened to soft music without lyrics for 3 hours, whereas women in the control group did not listen to music. The trial was conducted in Thailand. [EL = 1+]

Review findings

The trial compared scores made using two VASs, and showed a significant reduction in both the sensation of, and distress from, pain (sensation of pain (pre and 3 hourly post tests undertaken three times): F(1107) = 18.69, P < 0.01, effect size = 0.15; distress of pain (undertaken as above): F(1107) = 14.87, P < 0.001, effect size = 0.12). There were no other outcomes investigated.


Description of included studies

Again, one RCT that was included in the systematic review was included.137 [EL = 1+] This was conducted in the UK and published in 1965. The study population comprised 25 women in labour. Women in the intervention group received audio-analgesia which consisted of ‘sea noise’ white sound set at 120 decibels, and the control group received sea noise at a maximum of 90 decibels.

Review findings

The trial reported maternal satisfaction about care received, which showed no evidence of a difference (RR 2.00 [95% CI 0.82 to 4.89]). There were no other outcomes available.

Evidence statement

There is some evidence from small studies regarding the use of acupuncture, acupressure and hypnosis for the management of pain in labour. There is a lack of evidence on other outcomes.

Acupuncture seems to be associated with a reduction in the use of pharmacological pain relief and augmentation, but with no reduction in pain scores.

Hypnosis seems to be associated with a reduction in the use of pharmacological pain relief and augmentation. There is a lack of evidence on pain scores.

There is a lack of high-level evidence that music, aromatherapy or audio-analgesia influence women’s pain in labour or any other outcome.

Recommendations on complementary therapies

Acupuncture, acupressure and hypnosis should not be provided, but women who wish to use these techniques should not be prevented from doing so.

The playing of music of the woman’s choice in the labour ward should be supported.

Research recommendation on non-invasive techniques in labour

A combination of randomised trials and qualitative research should investigate the effect of a package of care, involving the use of non-invasive techniques throughout labour and birth, on women’s birth experiences. This should include studies that explore which aspects of the package of care affect both women’s experience and maternal and neonatal outcomes.

5.4. Non-pharmacological analgesia


This section covers TENS, which once again does not require professional oversight.

Clinical question

Is there evidence that the type, frequency and mode of administration of the following pharmacological and non-pharmacological pain relief and regional analgesia influence outcomes?

  • pharmacological pain relief: Entonox®, PCAs, pethidine, diamorphine and meptazinol (Meptid®)
  • non-pharmacological pain relief: TENS.

5.4.1. Transcutaneous electrical nerve stimulation (TENS)

Description of included studies

One systematic review conducted in 1997 was identified.139 (n = 877: TENS n = 436; controls (sham TENS or no treatment) n = 441). The systematic review included ten RCTs, among which three RCTs compared TENS with no TENS, seven RCTs compared TENS with sham TENS and one RCT compared both. Only one RCT achieved an adequate level of blinding. [EL = 1+]

Review findings

Pain outcome measures were reported in ten RCTs. There was no consistency in the method of measuring, but no study recorded any difference in pain intensity or pain relief scores between TENS and controls. The need for additional analgesic interventions was reported in eight RCTs. There was no evidence of difference for this need (combined RR 0.88 [95% CI 0.72 to 1.07]). There were no reports of adverse events in the ten RCTs.

Evidence statement

There is high-level evidence that TENS is not an effective analgesic in established labour. There is no high-level evidence on the analgesic effect of TENS in the latent phase of labour.

Recommendation on TENS

Transcutaneous electrical nerve stimulation (TENS) should not be offered to women in established labour.

5.5. Inhalational analgesia


This form of analgesia has been available since 1962 and approved for midwives to administer since 1970. It involves the woman inhaling through a mask or mouthpiece. It has the advantages of rapid action, it is non-accumulative and does not pass across the placenta to affect the baby.

5.5.1. Nitrous oxide

Description of included studies

One systematic review published in 2002 was identified.140 The study included eight controlled studies and eight observational studies. [EL = 2+] While most studies included use of a 50% nitrous oxide concentration, nine involved comparisons of varying concentrations ranging from 30% to 80%. Owing to the inconsistency of the included methods, the results are summarised descriptively.

Review findings

Analgesic efficacy was adequately reported in 11 studies. Although there was no clear, quantitative, objective evidence, seven studies described significant analgesia with nitrous oxide and two studies reported that women chose to continue using nitrous oxide even after the study period was over.

The effect of nitrous oxide on uterine contractions was reported in one study, and no alteration was observed. Another study found no effect on the progress of labour. Nausea and vomiting was reported as ranging from 5% to 36% with nitrous oxide but there were no proper controls in eight studies. Loss of consciousness was reported in two RCTs, but this was not statistically significant.

Apgar scores were reported in four studies, and there was no evidence of any differences. One study also showed no difference in early neurobehavioural scale.

Evidence statement

There is a moderate level of evidence to support the use of nitrous oxide in labour. Nitrous oxide seems to relieve some pain but can make women feel nauseous and light-headed. There is no evidence of harm to the baby.

Recommendation on nitrous oxide

Entonox (a 50: 50 mixture of oxygen and nitrous oxide) should be available in all birth settings as it may reduce pain in labour, but women should be informed that it may make them feel nauseous and light-headed.

5.6. Intravenous and intramuscular use of opioids for labour


Pethidine is widely used as an analgesic during labour. Its ease of administration and the fact that the Central Midwives’ Board approved it in 1950 has probably contributed to its widespread use.

Pethidine did not undergo RCTs prior to its introduction into clinical practice in the UK, and its perceived analgesic efficacy could in part be due to its sedative effects.

5.6.1. Intramuscular use of opioids

Intramuscular (IM) opioids versus placebo

Pethidine versus placebo

Description of included studies

Two double-blind RCTs compared IM pethidine with an IM placebo. The first trial (n = 224) was reported in a systematic review.141,142 [EL = 1+] A second RCT involving 50 women was conducted in Hong Kong.143 [EL = 1+]

Review findings

An RCT, reported in a systematic review, found significantly more women were dissatisfied with pain relief in the placebo group compared with the group of women who received pethidine when assessed during labour (83% versus 71%, P = 0.04) and after giving birth (54% versus 25%, P = 0.00004). It should be noted that the number of women dissatisfied with pain relief was high in both groups. Similarly, significantly more caregivers were dissatisfied with the placebo. No other outcomes were investigated.

Results from a second RCT conducted in Hong Kong support these findings.143 [EL = 1+] A significant reduction in VAS pain score 30 minutes post-administration was found for women in the group who received pethidine (n = 25) compared with those who received the placebo (n = 25) (pethidine: median change −11 mm; placebo: median change +4 mm, P = 0.009). At 30 minutes the median VAS score was significantly lower in the pethidine group compared with the control group (54 versus 78 mm, P = 0.01). Eight women in the group who received pethidine required no further analgesia compared with one in the control group (P = 0.01). Thirty minutes after administration, women were also asked to rate on a 5-point Likert scale how satisfied they were with the pain relief received. Scores were significantly higher for women in the pethidine group, although neither had very high scores (median 2 for pethidine group and 1 for control group). Eight percent of women in the pethidine group were totally dissatisfied with the pain relief received compared with 60% in the control group.

IM opioids versus IM opioids: different opioids

Description of included studies and review findings

Tramadol versus pethidine

Two systematic reviews include three RCTs of tramadol 100 mg versus pethidine 50–100 mg for analgesia in labour.141 142 [EL = 1+] The trial sizes ranged from 40 to 60 women, and in total involved 144 women.

Two trials reported women’s satisfaction with pain relief 1–2 hours post-administration and both found no significant difference between the two groups (women not satisfied with pain relief: 15/50 versus 13/49; OR 1.18 [95% CI 0.49 to 2.84]). The third trial reported VAS scores following administration of analgesia, which was significantly lower in the group of women who had received pethidine compared with those who received tramadol (mean 66.10 mm [SD 18.34 mm] versus 52.91 mm [SD 22.23 mm]; WMD 13.20 mm [95% CI 0.37 to 26.03 mm]). All trials included measures of nausea/vomiting during labour and meta-analysis of the findings showed no significant difference between the two drugs (6/74 versus 9/70; OR 0.63 [95% CI 0.21 to 1.84]). Similarly, meta-analysis of the findings showed no significant difference in drowsiness/sleepiness between tramadol and pethidine (16/74 versus 22/70; OR 0.61 [95% CI 0.29 to 1.29]). There were no significant differences found for mode of birth. Neonatal outcome was not evaluated.

A more recent RCT conducted in Turkey (n = 59) reported greater pain relief with pethidine (100 mg) compared with tramadol (100 mg), although neither provided good analgesia.144 [EL = 1+] On a 5-point Likert scale of pain intensity, the median score at 1 hour post-administration was 4 for pethidine and 5 for tramadol (P < 0.05). Incidence of nausea and fatigue 1 hour following drug administration was also significantly higher in women who were given tramadol (nausea: 1/29 versus 9/30, P = 0.004; fatigue: 15/29 versus 23/30, P = 0.045). The findings of this study are a little difficult to interpret as it is not clear which statistical tests were applied to any given comparison. The incidence of neonatal respiratory depression was high in this study, occurring in three of the babies born in the pethidine group and seven in the tramadol group. It was reported that all recovered with ‘supplementary oxygen therapy in the ICU’. No opiate antagonists were given. Mean Apgar scores at 1 minute were 7.76 (SD 1.06) and 7.13 (SD 1.38), and at 5 minutes 9.28 (SD 0.65) and 9.17 (SD 0.91) in the pethidine and tramadol groups, respectively.

Meptazinol versus pethidine

Evidence for this section is drawn from the two systematic reviews identified in the subsection above and includes the same seven trials that compare pethidine with tramadol.141,142 [both EL = 1+] In six of the trials, 100 mg meptazinol (Meptid®) was compared with a similar dose of pethidine. In one trial the comparison was between 75 mg meptazinol and 50 mg pethidine. The trials involved a total of 1906 women, with trials ranging in size from 10 to 1035 women.

A variety of outcome measures were used to assess pain relief, e.g. lack of satisfaction with pain relief 1–2 hours post-administration, VAS (0–100), need for additional pain relief during labour and use of epidural analgesia. In these studies, analgesia was found to be similar for the two drugs, with no significant differences between the various outcome measures. Three trials investigated nausea and vomiting. In two of these trials, there were no significant differences but the largest trial (n = 1035) showed that pethidine resulted in significantly less nausea and vomiting (184/498 versus 141/507; OR 1.52 [95% CI 1.17 to 1.98]). Meta-analysis of the three trials retains this significant difference owing to the dominant effect of the large trial (OR 1.37 [95% CI 1.09 to 1.72]). The large trial was the only study to investigate drowsiness/sleepiness and found this to be significantly higher in women who had received pethidine compared with those who received meptazinol (202/522 versus 147/513; OR 0.64 [95% CI 0.49 to 0.83]). No significant differences were found regarding mode of birth, fetal distress, Apgar scores, neonatal death or admission to a neonatal unit. Only the large trial reported on naloxone administration as an outcome measure, where a tendency towards higher incidence of naloxone administration was noted for babies born to women in the pethidine group (231/496 versus 198/479; OR 0.81 [95% CI 0.63 to 1.04]). This high incidence of naloxone administration is not commented upon by the authors of the review, although it should be noted that naloxone use is much less frequent in current UK practice following recommendations made by the UK Resuscitation Council.

Diamorphine versus pethidine

One UK RCT compared IM diamorphine (n = 65) with IM pethidine (n = 68).145 [EL = 1+] Nulliparous women were randomised to receive either IM pethidine 150 mg or diamorphine 7.5 mg. Multiparous women were randomised to receive either 100 mg pethidine IM or 5 mg diamorphine IM. All participants received the anti-emetic prochlorperazine at the same time as the trial drugs. Two measures of pain relief favoured diamorphine compared with pethidine: VAS score (0–100) 1–2 hours after administration (58 versus 67; WMD −9.00 [95% CI −10.21 to −7.79], P < 0.0001) and women not satisfied with pain relief 1–2 hours post-administration (35 versus 56; RR 0.63 [95% CI 0.43 to 0.94], P = 0.02). Vomiting during labour was also significantly reduced in the group of women who received diamorphine (11 versus 28; RR 0.39 [95% CI 0.17 to 0.86], P = 0.02). No significant difference was found for sleepiness or drowsiness during labour, mode of birth, 5 minute Apgar scores or neonatal death/admission to neonatal intensive care unit (NICU). Time from first drug dose to birth was shorter for women assigned to the pethidine group (4.5 versus 4.9 hours, WMD 0.40 hours [95% CI 0.26 to 0.54 hours]). This represents a difference of 24 minutes, which is not likely to be significant clinically.

Pentazocine versus pethidine

Six double-blind RCTs compared 40–60 mg pentazocine with 100 mg pethidine. Trial sizes ranged from 60 to 180 women, including 678 women in total, and are summarised in two systematic reviews.141,142 [EL = 1+] Based on a meta-analysis of findings from all six studies, no significant difference was found between the two groups regarding pain relief (measured as women not satisfied with pain relief 1–2 hours after administration): OR 0.99 [95% CI 0.70 to 1.39]. Significantly more women in the pentazocine group required further analgesia (OR 1.95 [95% CI 1.31 to 2.89], data from five studies). There was a trend towards fewer women suffering nausea and vomiting during labour in the pentazocine group, although the numbers involved were small and did not reach statistical significance (OR 0.56 [95% CI 0.30 to 1.07]). No significant differences were noted for drowsiness/sleepiness in labour. Few trials reported on other outcomes and, where they did, the numbers involved were small and differences not statistically significant.

IM opioids versus IM opioids: same opioid, different doses

Description of included studies

Two trials conducted in the 1970s compared a higher and lower dose of pethidine. Both are reported in the two systematic reviews outlined above.141,142 A total of 173 women were involved in the two studies. One trial reported in both systematic reviews compared tramadol 50 mg (n = 30) with 100 mg (n = 30).141,142

Review findings

Pethidine 40–50 mg versus pethidine 80–100 mg

Each study used a different outcome for the assessment of pain relief. In the larger study, women’s satisfaction with pain relief 1–2 hours post-administration was recorded. A high proportion of women in both groups were not satisfied with the pain relief received; 42/55 in the lower dose group and 37/57 in the higher dose group (OR 1.73 [95% CI 0.77 to 3.88]). The smaller study (n = 20 in each group) reported numerical pain scores 2 hours after drug administration. Again there was no difference between the two groups (mean numerical pain score): lower dose 1.70 (SD 0.63); higher dose 1.35 (SD 0.45); OR 0.35 [95% CI 0.01 to 0.69]. Both studies reported the need for additional analgesia (other than epidural), which was significantly higher for women in the lower dose group (28/88 versus 10/85; OR 3.74 [95% CI 1.75 to 8.00]). The use of epidural analgesia was not reported, perhaps because these studies were carried out in the 1970s when the use of epidural analgesia was not widespread. The incidence of nausea and vomiting was also investigated by both studies and, although found to be higher for the higher dose in both, this did not reach statistical significance (9/88 versus 17/85; OR 0.46 [95% CI 0.20 to 1.06]. Drowsiness and sleepiness were also more commonly reported by women in the higher dose group, although again this increase did not reach statistical significance (11/68 versus 19/65; OR 0.48 [95% CI 0.21 to 1.07], one study). No other maternal outcomes were reported. Neonatal outcomes were only investigated by the smaller study, where one baby in the higher dose group required resuscitation and one required naloxone, compared with none in the lower dose group.

Tramadol 50 mg versus 100 mg

Findings from this trial showed that more women in the lower dose group were not satisfied with pain relief 1–2 hours after administration (27/30 versus 7/30; OR 14.44 [95% CI 5.24 to 39.74]). Side effects were rare and slightly more prevalent in the higher dose group, but these differences did not reach statistical significance (nausea or vomiting: 1/30 versus 3/30, OR 0.35 [95% CI.005 to 2.61]; drowsiness or sleepiness: 2/30 versus 3/30, OR 0.65 [95% CI 0.11 to 4.00]). There was one instrumental birth and one caesarean birth in each group. No other outcomes were considered.

5.6.2. Intravenous use of opioids

Intravenous (IV) opioids versus placebo

IV pethidine versus IV placebo

Description of included studies

Two RCTs were reviewed that compared IV pethidine with an IV placebo (saline). The first was a double-blind RCT undertaken primarily in order to investigate the effects of pethidine on labour dystocia, looking at analgesic efficacy as a secondary outcome.146 [EL = 1+] A second RCT, carried out in Thailand, examined the efficacy and side effects of IV pethidine.147 [EL = 1+]

Review findings

In an RCT involving women in delayed active labour (4–6 cm cervical dilatation, delay diagnosed by attending obstetrician), the women were randomly assigned to receive 100 mg pethidine IV (administered in 50 ml saline over 15 minutes) (n = 205) or IV placebo (n = 202).146 [EL = 1+] Pain was assessed using a VAS 15, 30 and 60 minutes after administration. Pain scores, at all times, were significantly lower for women receiving the pethidine (severe pain score (7–10 on VAS): at 15 minutes RR 0.87 [95% CI 0.78 to 0.96]; at 30 minutes RR 0.75 [95% CI 0.66 to 0.84]; at 60 minutes RR 0.74 [95% CI 0.66 to 0.84]; during second stage: RR 0.77 [95% CI 0.69 to 0.86]). However, more than 66% of the women rated their pain scores as severe throughout the first hour following the administration of pethidine. The incidence of side effects was significantly higher in the women who received pethidine (any adverse effect: RR 1.91 [95% CI 1.44 to 2.53]; nausea: RR 1.60 [95% CI 1.05 to 2.43]; vomiting RR 1.97 [95% CI 1.09 to 3.55]; dizziness RR 4.68 [95% CI 2.59 to 8.46]). The need for augmentation with oxytocin was also significantly higher in the intervention group (RR 2.24 [95% CI 1.13 to 4.43]). Neonatal outcomes were also found to be significantly worse following the administration of pethidine, namely: Apgar < 7 at 1 minute: RR 4.11 [95% CI 1.72 to 9.80]; umbilical cord arterial pH< 7.20: RR 1.55 [95% CI 1.13 to 2.14]; umbilical cord arterial pH< 7.10: RR 3.94 [95% CI 1.76 to 8.82]. There were no significant differences in Apgar scores at 5 minutes: Apgar < 7 at 5 minutes: RR 11.82 [95% CI 0.66 to 210.25].

In a second RCT, women in established labour (3–5 cm cervical dilatation) requesting analgesia were randomly allocated to received IV pethidine (n = 42) (women < 75 kg received 50 mg, women > 75 kg received 75 mg) or IV saline (n = 42) (1.0 or 1.5 ml).147 [EL = 1+] Women who had nausea and/or vomiting were also given 25 mg promethazine. VAS scores were reported by women 15, 30 and 60 minutes post-administration. These scores were then categorised prior to statistical analysis (0 = no pain; 1–3 = mild pain; 4–7 = moderate pain; 8–10 = severe pain). An observer recorded the woman’s vital signs, the fetal heart rate (FHR) and rated level of sedation (on a 5-point Likert scale) at the same intervals. No significant differences were found between the intervention and control group regarding blood pressure (BP), pulse or respiratory rate, or the FHR (described as mean differences, no statistical analysis reported). No significant differences were found between the two groups for median pain scores at each time interval. The means of the pain increment scores for each time interval (i.e. 0–15 minutes, 15–30 minutes, etc.) were significantly higher for the control group throughout the study period. It is questionable, however, whether it is meaningful to calculate and compare means of categorical scores derived from a 0–10 scale. Side effects were more frequent in the intervention group: nausea/vomiting: n = 15 versus n = 2; dizziness: n = 11 versus n = 0. The authors reported no significant differences for mode of birth, Apgar scores or administration of naloxone, but no figures were given. Women’s views of pain relief were sought within 24 hours of giving birth. While significantly more women in the intervention group gave positive reports of the effectiveness of pain relief, this figure was only 23.80% compared with 7.10% in the control group.

IV opioids: dose-finding

IV morphine

Description of included studies

A dose-finding study, conducted in Sweden, investigated the analgesic efficacy of IV morphine during the first stage of labour.148 [EL = 3]

Review findings

IV morphine was given to 17 women (11 nulliparae) in active labour (three contractions every 10 minutes lasting at least 60 seconds and a cervical dilatation of at least 4 cm) and requesting analgesia. Amniotomy was performed if membranes had not ruptured spontaneously. All women were given repeated doses of IV morphine (0.05 mg/kg) after every third contraction, until a total dose of 0.20 mg/kg was reached. Pain intensity and level of sedation were measured using a 10 cm VAS scale. Women were also asked to indicate on a schematic diagram where the pain was located. Pain assessments were performed immediately after the first three contractions following each administration of morphine. Morphine was found to significantly reduce reported pain intensity (initial pain intensity versus pain intensity following four doses of morphine: mean = 85 mm [range 53 to 100 mm] to 70.0 mm [range 46 to 99 mm], z = 2.49, P = 0.01, Wilcoxon test). However, this decrease translates to a reduction from ‘unbearable’ to ‘severe’ pain rather than a clinically significant reduction in pain. The number of women experiencing back pain was significantly reduced from 13/14 to 4/14 (P = 0.01) but in 14/17 women there was no reduction in abdominal pain after morphine administration. Following morphine administration, 14/17 women requested and received epidural analgesia. The sedative effects of IV morphine were marked: VAS before versus after morphine administration 0 mm [range 0 to 0 mm] versus 78 mm [range 56.1 to 99.5], P < 0.05. The authors also reported that several women who received the maximum dose of morphine were asleep between contractions, and three could not be given all the dose increments of morphine owing to its severe sedative effects. No difference in neonatal outcome was reported (Apgar scores at 1 and 5 minutes).

IV opioids versus IM opioids

IV pethidine versus IM pethidine

Description of included studies

One Canadian RCT was identified that compared IV pethidine (n = 19) with IM pethidine (n = 20).149 [EL = 1+]

Review findings

IM pethidine was administered in 50–100 mg doses every 2 hours as required, up to a maximum dose of 200 mg. The IV group of women received a 25 mg bolus then a background infusion rate of 60 mg/hour, with an additional 25 mg bolus available at hourly intervals if required. The main outcome measure was pain intensity during labour, which was measured using a 10 cm VAS when the analgesia was administered and every 30 minutes thereafter. Other outcome measures included pulse rate, BP, respiratory rate, side effects of medication, levels of sedation (5-point Likert scale), mode of birth and a second or third day postnatal assessment of satisfaction with pain relief. The baby’s Apgar scores, vital signs and any required resuscitation interventions were also recorded. No significant differences were found between groups for maternal physiological measurements. The women who received IV pethidine had significantly lower pain scores from times 1.5 hours to 4.0 hours. However, women in the IM group received significantly less pethidine (mean = 82 mg) compared with the IV group (mean = 121 mg). Four women in the IV group received one additional bolus of 25 mg pethidine and one woman received two additional boluses. Eight women in the IM group also used Entonox compared with one in the IV group. Subgroup analysis of findings from women in the IV group who received 100–150 mg pethidine (mean dose 127 mg) (n = 10) still showed significantly lower pain scores when compared with women who received 100 mg pethidine IM. No other statistically significant differences were found regarding side effects, infant outcomes or women’s satisfaction 2–3 days postnatally.

IV opioids versus IV opioids

Butorphanol versus pethidine versus butorphanol + pethidine

Description of included studies

A recent US RCT compared 1 mg butorphanol, 50 mg pethidine or both drugs in combination (0.5 mg butorphanol + 25 mg pethidine).150 [EL = 1−] Fifteen women were randomly allocated to each group. Unfortunately, owing to the loss of an undisclosed number of women post-randomisation (including exclusion of women who requested an epidural within seven contractions of IV drug administration), there is a potentially high level of bias within the trial.

Review findings

Level of sedation, pain intensity and nausea were assessed using a 0–10 verbal scale, just before drug administration and between the sixth and seventh contraction post-administration. Women were also asked to choose words from a pain affective magnitude check list to describe the pain of the previous two contractions. All three treatments provided significant, but only moderate, pain relief (verbal scale scores before and after administration (mean): butorphanol: 7.2 (SD 0.6) versus 5.5 (SD 0.8), P < 0.05; pethidine: 7.4 (SD 0.4) versus 5.2 (SD 0.5), P < 0.05; butorphanol + pethidine: 7.4 (SD 0.4) versus 4.7 (SD 0.8), P < 0.05). No significant difference was found between groups regarding degree of pain relief. Unfortunately, the study did not report on the number of women who requested or received additional pain relief (the study ended with the seventh uterine contraction after administration of the study drug). Sedation increased after all drug treatments to a similar degree. Nausea was unaffected by drug treatment. (Exact figures are not reported but the findings are represented graphically.) FHR abnormalities were not significantly different between treatment groups (n = 5, 3, 5 butorphanol, pethidine, combination, respectively). Only two babies had Apgar scores of below 8 at 1 minute (one score of 6 in the butorphanol group and one score of 7 in the pethidine group). All babies had Apgar scores of 8 or above at 5 minutes.

IV patient-controlled analgesia (PCA): different opioids

Description of included studies and review findings

IV PCA: remifentanil versus pethidine

Two small UK RCTs provided the evidence for analgesic efficacy of PCA remifentanil compared with PCA pethidine.151 [EL = 1+] 152 [EL = 1−]

In a recent RCT women received either remifentanil 40 micrograms with a 2 minute lockout (n = 20) or pethidine 15 mg with a 10 minute lockout (n = 20).151 [EL = 1+] Baseline assessments were carried out for pain intensity (10 cm VAS), sedation score (5-point Likert scale), vital signs, nausea and anxiety. These measurements were repeated every 30 minutes following the administration of analgesia along with assessments of women’s satisfaction with analgesia (10-point VAS). Continuous pulse oximetry was also carried out, plus 1 hour of continuous FHR monitoring following the commencement of PCA. One protocol violation was noted for a woman in the pethidine group and her data removed from the analysis (i.e. not an intention-to-treat analysis). Eighteen women in the remifentanil group continued to use the PCA up to, and during, birth compared with 14 women in the pethidine group (NS). Almost all women in both groups used Entonox as well as IV PCA. No significant differences were noted for pain intensity scores between the two groups (overall mean (SD) remifentanil: 6.4 cm (1.5 cm); pethidine: 6.9 cm (1.7 cm)). There were also no significant differences noted for levels of nausea, sedation, anxiety or time spent with oxygen saturation < 94% or < 90%. Satisfaction scores at 60 minutes were significantly higher for remifentanil than pethidine (median): 8.0 [IQR 7.5 to 9.0] versus 6.0 [IQR 4.5 to 7.5], P = 0.029). No significant differences were noted for classification of FHR tracings, Apgar scores or cord blood pH. Babies in the pethidine group had significantly lower Neurologic Adaptive Capacity Scores 30 minutes after birth, but there was no difference after 120 minutes.

An earlier small-scale double-blind RCT conducted at the same UK hospital also compared PCA remifentanil and PCA pethidine, although with slightly different doses.152 [EL = 1−] Nine women were randomised to receive an IV bolus of remifentanil 0.5 micrograms/kg with a lockout period of 2 minutes and eight women were randomised to receive a bolus of 10 mg pethidine with a lockout period of 5 minutes. A 10 cm VAS was used to assess pain, nausea and itching immediately prior to administration of analgesia, at hourly intervals post-administration throughout labour and again 30 minutes after giving birth. Women’s vital signs were also recorded along with 1 and 5 minute Apgar scores. At the start of the study, more women in the remifentanil group were receiving oxytocinon compared with women in the pethidine group (6/9 versus 2/8). Despite this, there was no significant difference in the initial baseline mean VAS score for pain (pethidine 47 mm; remifentanil 48 mm). The mean VAS score for pain throughout labour was reported as being significantly lower in the remifentanil group (actual value not given, although hourly mean scores were represented graphically). The post-birth VAS score was also reported to be significantly lower for women in the remifentanil group (again actual value not stated). No significant differences were found for nausea or itching between the two groups. No episodes of maternal hypotension, bradycardia or respiratory rate < 12 were recorded. Median Apgar scores at 1 and 5 minutes were found to be significantly lower in babies born to mothers who had received pethidine (median at 1 minute: remifentanil: 9 [range 9 to 9]; pethidine: 5.5 [range 5 to 8], P = 0.01; at 5 minutes: remifentanil: 10 [range 9 to 10]; pethidine: 7.5 [range 6 to 9], P = 0.04). One baby in the pethidine group was admitted to the neonatal unit. The trial was terminated early owing to concerns over the neonatal effects noted in the pethidine group.

IV PCA: fentanyl versus alfentanil

A small double-blind RCT conducted in Canada compared fentanyl with alfentanil, both administered as PCA.153 [EL = 1−] Women in the fentanyl group (n = 11) received a loading dose of 50 micrograms IV. The PCA pump was then programmed to deliver a dose of 10 micrograms with a lockout of 5 minutes. A background infusion of 20 micrograms/hour was maintained. Women randomised to receive alfentanil (n = 12) were given a loading dose of 500 micrograms IV. The PCA pump was programmed to deliver a dose of 100 micrograms with a background infusion of 200 micrograms/hour. Hourly measurements were made of the drug dose received, total dose, sedation score and side effects. VAS pain scores were recorded every 30 minutes. Neonatal effects were assessed by Apgar scores, umbilical venous and arterial blood gases and neurobehavioural scores recorded at 4 and 24 hours. Two women were withdrawn from the data analysis owing to failure to observe the study protocol (these are not reported in the figures above). The two study groups were similar regarding demographic and obstetric details. No significant differences were found between the two groups for VAS pain scores from 1 to 3 cm cervical dilatation (mean [SD]: fentanyl: 61.0 mm [19.6 mm]; alfentanil: 67.3 mm [29.2 mm]) or 4 to 6 cm cervical dilatation: fentanyl: 54.9 mm [24.9 mm]; alfentanil: 67.7 mm [20.2 mm]). However, the mean VAS pain scores at 7 to 10 cm cervical dilatation were significantly higher in the alfentanil group compared with the fentanyl group (64.6 mm [12.2 mm] versus 85.7 mm [13.9 mm], P < 0.01). No significant differences were observed for VAS scores for sedation, incidence of nausea or incidence of pruritus. Five of the 12 women receiving alfentanil described the pain relief as inadequate compared with one of the nine in the fentanyl group (NS). There were no significant differences in neonatal outcome with regard to Apgar scores, neurobehavioural scores, umbilical venous pH or naloxone requirement.

5.6.3. Patient-controlled administration for IV and IM use of opioids in labour

IV PCA opioids versus IM opioids

Description of included studies

One RCT was identified that compared IM diamorphine with IV PCA diamorphine for analgesia in labour.154 [EL = 1+] A second small unblinded RCT conducted in the UK compared remifentanil via PCA (n = 18) with 100 mg pethidine IM (+ anti-emetic) (n = 18) (n = 13 primigravid women in each group).155 [EL = 1−]

Review findings

IV PCA diamorphine versus IM diamorphine

This trial, carried out in Scotland in 2000–2002, assigned women to receive either 5 mg diamorphine IM (multigravid women) or 7.5 mg diamorphine IM (primigravid women), or a loading dose of 1.2 mg diamorphine IV with a PCA pump set to deliver 0.15 mg diamorphine per dose with a 5 minute lockout period (maximum dose 1.8 mg/hour) (IM group n = 177; IV PCA group n = 179). Primary outcomes were analgesia requirements during labour and women’s satisfaction with pain relief. Women’s perceptions of pain in labour, side effects and clinical outcomes for the women and babies were also recorded. Pain intensity during labour was measured using a verbal descriptor with pain at four levels and a 10 cm VAS. Pain scores were repeated hourly, between contractions, throughout labour. Findings for primigravid women and multigravid women are reported separately.

In primigravid women, those in the PCA group used significantly less analgesia than those in the IM group (IM mean 3.2 mg/hour; PCA mean 1.7 mg/hour; difference 1.5 mg/hour [95% CI 1.1 to 1.9 mg/hour], P < 0.001). Women in the PCA group were more likely to opt for an epidural and less likely to remain in the trial until the baby was born, although these differences did not reach statistical significance. Most women (over 80% in both groups) used additional analgesia, e.g. Entonox or TENS). Findings for multigravid women were similar. Again women in the PCA group used significantly less diamorphine compared with women in the IM group (IM mean 3.1 mg/hour; PCA mean 1.6 mg/hour; difference 1.6 mg/hour [95% CI 1.1 to 2.0 mg/hour], P < 0.001). Significantly fewer multigravid women completed their labour using IV PCA diamorphine compared with IM diamorphine (61% versus 79%, RR 0.77 [95% CI 0.61 to 0.97] but the need for an epidural was similar between the two groups, and much lower than in primigravid women (15%). Satisfaction with intrapartum pain relief measured 6 weeks postnatally was lower for women in the PCA group. Primigravid women allocated to the PCA group were significantly more likely to state that they were very dissatisfied with their use of diamorphine compared with women in the IM group (PCA 35% versus IM 7%, RR 5.08 [95% CI 2.22 to 11.61]). Only 34 % of primigravid women in the PCA group reported that they would use diamorphine again compared with 61% of the IM group (RR 0.56 [95% CI 0.40 to 0.79]). Findings for multigravid women were similar with significantly more women saying they were very dissatisfied with PCA diamorphine and significantly fewer in the PCA group stating that they would use it again. In addition, 44% of multigravid women in the PCA group felt they had received pain relief too late in labour, compared with 19% of IM users (RR 2.32 [95% CI 1.21 to 4.49]). The mean VAS score for primigravid women in the IM group was significantly lower than that for the PCA group (6.7 versus 5.3, difference 1.4 [95% CI 0.8 to 2.0]). There was no difference in mean maximum VAS scores. No significant differences were found for multigravid women’s reported pain intensity during labour. Clinical outcomes were similar for women and babies in both groups. The authors explained the relatively poor outcomes for PCA diamorphine by stating that women and midwives appeared to lack confidence in the PCA and its ability to relieve intrapartum pain. Most women allocated to the PCA group used only a small proportion of the diamorphine potentially available to them, and quite quickly moved on to other forms of analgesia.

IV PCA remifentanil versus IM pethidine

An unblinded RCT conducted in the UK compared remifentanil via PCA (20 micrograms bolus over 20 seconds, 3 minute lockout, no background transfusion) (n = 18) with 100 mg pethidine (+ anti-emetic) (n = 18) (n = 13 primigravid women in each group).155 [EL = 1−] Pain was assessed using a 10 cm VAS. Sedation and anxiety were assessed using a similar scale. Degree of nausea and vital signs were also recorded. All measurements were made prior to administration of analgesia and every 30 minutes thereafter. All women were monitored using continuous pulse oximetry. Pain scores at 60 minutes post-administration and maximum pain score during the first 2 hours post-administration were significantly lower in the PCA remifentanil group (median scores at 1 hour: 72 versus 48, P = 0.0004; maximum scores over 2 hours: 82.5 versus 66.5, P = 0.009). Women’s and midwives’ assessment of ‘overall effective analgesia’ were both significantly higher in the remifentanil group. For two women receiving pethidine and seven receiving remifentanil, haemoglobin saturations of ≤ 94% were recorded. The minimum saturation did not differ significantly between the two groups. There was no significant difference in the minimum recorded ventilatory rates between women in the two groups. There was no significant difference in numbers of women experiencing nausea and vomiting between the two groups (pethidine n = 10, remifentanil n = 5, P = 0.06). Significantly fewer women in the remifentanil group had a spontaneous vaginal birth (11/18 versus 16/17, P = 0.04). The authors reported no difference in Apgar scores between the two groups; however, this was based on data from the subgroup of women who did not receive an epidural.

Evidence statement

Parenteral opioids have a limited effect on pain in labour irrespective of the agent, route or method of administration. Tramadol, meptazinol and pentazocine are not widely used in the UK and the evidence to date shows no advantage over pethidine. There is limited evidence that diamorphine (IM) provides more effective analgesia than the other opioids studied, with the fewest side effects for the woman.

There is a lack of evidence on the optimum dose or route of administration, as well as the effect of opioids on infant behaviour in the longer term, particularly feeding.

Recommendations on intravenous/intramuscular opioids

Pethidine, diamorphine or other opioids should be available in all birth settings. Women should be informed that these will provide limited pain relief during labour and may have significant side effects for both the woman (drowsiness, nausea and vomiting) and her baby (short-term respiratory depression and drowsiness which may last several days).

Women should be informed that pethidine, diamorphine or other opioids may interfere with breastfeeding.

If an intravenous or intramuscular opioid is used, it should be administered with an antiemetic.

Women should not enter water (a birthing pool or bath) within 2 hours of opioid administration or if they feel drowsy.

Research recommendation on intravenous/intramuscular opioids

An RCT to compare the effect of pethidine [IM] and diamorphine [IM], and to explore optimum doses. Outcomes should encompass analgesic effect, and short- and long-term neonatal outcomes (including breastfeeding).

Copyright © 2007, 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 Intrapartum Care
Intrapartum Care: Care of Healthy Women and Their Babies During Childbirth.
NICE Clinical Guidelines, No. 55.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2007 Sep.


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...