NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Caesarean Section. London: RCOG Press; 2011 Nov. (NICE Clinical Guidelines, No. 132.)

Cover of Caesarean Section

Caesarean Section.

Show details

6Factors affecting likelihood of caesarean section during intrapartum care

6.1. Place of birth

Planned home birth

One systematic review that includes one small randomised controlled trial (RCT) comparing planned home birth to planned hospital birth was identified (n = 11). The RCT included operative delivery but not specifically caesarean section (CS). No difference was reported for any of the outcomes measured however this was a small RCT and has limited power to detect a difference.171 [evidence level 1b]

A systematic review of observational studies evaluating the safety planned home births (in countries with good health resources) versus planned hospital births identified six cohort studies (n = 24,092)172 [evidence level 2b] Outcome measures included perinatal and maternal mortality, Apgar scores and incidence of maternal lacerations. The review also reported other outcomes including CS rates. No difference was detected in perinatal mortality in any of the individual studies, nor in the pooled data. In the home birth group, both low 5 minute Apgar and maternal lacerations were less frequent in all studies. The odds of CS were lower in the planned home birth group in five studies (reported crude odds ratio [OR] of CS in studies: 0.04; 0.09; 0.31; 0.05; 0.27). No maternal deaths occurred but the studies are underpowered to evaluate this outcome.172 [evidence level 2a]

A subsequent cohort study in Canada (n = 2176) reported on CS rates and maternal and perinatal morbidity between 3 groups, women who had a planned home birth, women who were attended by a physician in hospital and women who were attended by a midwife in hospital. They reported that less women in the home birth group had a CS, compared with women in the physician-attended hospital group (adjusted OR 0.3, 0.22 to 0.43) and compared with the midwife attended hospital group (adjusted OR 0.66, 0.44 to 0.99). Odds ratios were adjusted for maternal age, lone parent status, income quintile, substance use and parity. No difference was detected between the groups for maternal or perinatal morbidity.173 [evidence level 2a]

A large prospective case controlled UK study of 5971 planned home births and 4724 planned hospital births reported that planning a home birth halved the chance of having a CS (unadjusted OR 0.49, 95% confidence interval [CI] 0.39 to 0.62).174 [evidence level 2b]

‘Midwifery-led unit’ or ‘birthing centre’

Current convention in the UK is that the term “midwifery-led units” refers to units that are near to or adjacent to a hospital maternity facility and that “birthing centres” are stand alone units. However, this convention is not standardised in the in the literature. The centres are intended for “low risk” women. The care is midwife led with minimal medical intervention, sometimes described in the literature as ‘home like’. Case series have reported reduced CS or operative delivery in ‘midwifery-led units’ or’ birthing centres’.175180 [evidence level 4]

A systematic review that included six RCTs (n = 8677) compared clinical outcomes between women delivering in a midwife led unit or in a hospital.181 [evidence level 1a] The RCTs were conducted in Stockholm182, Australia183, United Kingdom184186 and Canada.187 The centre in each of the RCTs was situated close to the conventional labour ward within the same hospital setting. The RCTs all describe the environment as ‘home like’ and that the care was aimed at women retaining control and choice with minimal medical intervention. Three of the studies do not describe the study environment any further.182,183,185 Three of the studies describe the furnishings in detail (for example “furnished to appear like a normal household bedroom”)184,186,187 and one RCT also mentions specifically interventions that were avoided such as enemas, perineal shaving, intravenous infusion and electronic fetal monitoring.187 [evidence level 1b]

All RCTs (n = 8646) reported on CS rates, a further 39 outcomes are also reported. No difference was detected in CS rates between ‘midwifery-led unit’ and conventional birth settings (risk ratio [RR] 0.85, 0.72 to 1.00). The review has a 90% power to detect a difference of at least 2% in CS rates if such a difference exists. No difference in instrumental vaginal deliveries was detected (OR 0.87, 0.74 to 1.01). Birth in a ‘midwife-led unit’/’birth centre’ was associated with lower rates of intrapartum analgesia (OR 0.82, 0.72 to 0.93); less augmented labour (OR 0.72, 0.64 to 0.81); and fewer women ‘less than completely satisfied with care’ (OR 0.62, 0.55 to 0.70).181 [evidence level 1a]

A further UK RCT (n = 2578) comparing care ‘midwifery-led unit’ or in a conventional labour ward did not evaluate mode of delivery but assessed maternal satisfaction using a postal questionnaire. No difference was detected in rates of satisfaction between the groups. Women who had their babies in the ‘midwifery-led unit’/’birthing centre saw fewer medical staff, were more likely to report having had a choice as to moving around during childbirth and alternative positions for birth and more likely to have made their own decisions regarding analgesia.188 [evidence level 1b]

We did not identify any RCTs that compared birthing centres which are stand alone to conventional maternity facilities. However we did identify a case series following women admitted for labour and delivery at 84 ‘free standing’ birthing centres in the United States (n = 11,814). The overall rate of CS was 4.4%. The rate of transfer to other maternity facilities before birth was 11.9%. Other morbidity outcomes reported include 5-minute Apgar of less than 7 occurred in less than 0.5% of births.178 [evidence level 4]

An Australian postnatal survey of women's views about their birth experience (n = 395) reports that women who had given birth at home or at a ‘midwifery-led unit’ were more likely to feel that the birth place affected the bonding process and less likely to see birth as a medical condition compared to women who gave birth in a conventional labour ward. Women who gave birth at home were older, more educated, more likely to be multiparous and better informed about childbirth compared to the women who gave birth in the ‘midwife-led unit’ or in the conventional labour ward. Adjusting for these differences, place of birth correlated with women's satisfaction with healthcare providers.189 [evidence level 3]


NumberResearch recommendation
RR 19RCTs comparing planned birth in a stand-alone birthing centre to birth in conventional maternity facilities or midwifery led units.
RR 20Qualitative research is needed to explore women's opinions on place of birth and the impact of place of birth on their birth experiences.

Delayed admission to labour ward

A systematic review included one RCT (n = 209) compared a labour assessment program in a separate unit within the hospital and delayed admission to labour ward until labour is in the active phase, with direct admission to the labour ward.190,191 The RCT did not detect a difference in CS rates between the two groups (OR 0.70, 95% CI 0.27 to 1.79). At least two thousand women would be needed in each group to detect a 3% difference in CS therefore this RCT is underpowered to detect this difference in CS rates. There were differences in other outcomes such as length of time spent in the labour ward, analgesia requirements, oxytocic use and maternal satisfaction, measured using sense of control (see evidence table). [evidence level 1b]

An observational study (n = 3220) reported a reduced likelihood of CS with increased cervical dilatation at the time of presentation in labour The CS rates for nulliparous women presenting at 0–3 cm was 10% compared with 4% for those presenting at 4–10 cm (p = 0.001). This was consistent for nulliparous and parous women.192 [evidence level 2b]


NumberResearch recommendation
RR 21Further RCTs are needed to determine the effect of ‘delayed admission in labour’ on the likelihood of CS.

6.2. Factors reducing the likelihood of CS

One-to-one support

One-to-one support in labour had been evaluated in recently published systematic review;193 this current review replaces the previous review on this subject by the same authors.194 [evidence level 1a]. The first review included 14 RCTs (n = 5000), the new review includes 15 RCTs (n = 12,791) the newly included study is a multi centre RCT (n = 6915 women) conducted in Canada and the US (13 centres). The trial evaluated the effectiveness of continuous labour support by a specially trained nurse/midwives to usual care. Each hospital in the RCT had a CS rate of at least 15%. The main outcome measure was CS rate. The study did not detect a difference in CS rate between the two groups. The use of continuous electronic fetal monitoring higher in the usual care group (79%) compared to those in the continuous support group (75%, p < 0.001). All comparisons of women's likes and dislikes, and their future preference for amount of nursing support, favoured the continuous labour support group.195 [evidence level 1a]

The new systematic review (15 RCTs, n = 12,791 women) evaluates the effects of one-to-one support on women and their babies. In addition the new review also considers whether the effects of continuous support are influenced by routine practices and policies in the birth environment that may affect a woman's autonomy, freedom of movement and ability to cope with labour; whether the caregiver is a member of staff and whether the continuous support begins early or late in labour.193 [evidence level 1a] The RCTs in the review included support persons that varied in terms of their experience, qualifications and relationship to the women in childbirth. In eight RCTs the support was provided by a member of hospital staff. The remaining 7 RCTs included women from the community (“doulas”), with or without prior training, a childbirth educator, or a close female relative. Half of the RCTs were conducted in developed countries, where hospital policy permitted women to be accompanied by their husband/partners or other family members during labour. The remaining RCTs were conducted in developing countries in settings in which only the support person allocated by the study was allowed to accompany the woman during labour. No RCT evaluated the effects of husbands or partners as providers of support.

The results of the review reported that women who had continuous one-to-one support during labour were less likely to have a CS (15 trials, n = 12,791, RR 0.90, 95% CI 0.82 to 0.99). The effects of continuous support on CS appeared to be stronger in settings which did not permit the presence of additional support people (chi squared = 4.46, p < 0.05) and when epidural was not routinely available (chi squared 4.97, p < 0.05). The routine use of EFM did not affect the impact of one-to-one support on CS rates. The reduction in CS was influenced by who was giving the support and the reduction was only seen in the RCTs where the support was not provided by members of staff (RR 0.74, 95% CI 0.61 to 0.9). The difference between different sub-groups of non medical providers of support was not statistically significant. The impact of timing of onset of continuous support was of borderline statistical significance (chi squared = 5.93, p = 0.05) favouring support that began before active labour. Thirty other outcomes were considered in the review, but are not reported here.193 [evidence level 1a]


42Women should be informed that continuous support during labour from women with or without prior training reduces the likelihood of CS. [A] [2004]

Pregnancy after 41 weeks

A systematic review of 26 RCTs compared induction of labour with expectant management after 41 weeks. Offering routine induction after 41 weeks reduced perinatal death (19 RCTs, n = 7925. Peto OR 0.20, 95% CI 0.06 to 0.70) and the rate of CS (9 RCTs, n = 5954 Peto OR 0.87, 95% CI 0.77 to 0.99).196 [evidence level 1a]

It is estimated that by 41 weeks 74% of women have given birth, this increases to 82% by 42 weeks. The risk of stillbirth increases from 1 per 3000 ongoing pregnancies at 37 weeks to 3 per 3000 ongoing pregnancies at 42 weeks to 6 per 3000 with ongoing pregnancies at 43 weeks. A similar increase in neonatal mortality is also reported.197 [evidence level 2a]


43Women with an uncomplicated pregnancy should be offered induction of labour beyond 41 weeks because this reduces the risk of perinatal mortality and the likelihood of CS. [A] [2004]


Progress in labour can be assessed using the clinical parameters of descent of the presenting part and dilatation of the cervix. No study has evaluated tests based on maternal and fetal outcomes. The partogram is derived from a curve describing normal labour (Friedman's curve). The original Freidman's curve was developed using observational data from 100 American primigravid women at term in spontaneous labour (included 98 singleton cephalic, 1 breech presentation and 1 multiple pregnancy). Twenty two percent of the women received caudal anaesthesia and 10 percent received oxytocin augmentation. Cervical dilatation was determined using rectal examination predominantly at 10, 30 or 60 minute intervals. Curves of dilatation versus time were produced and resulted in a sigmoid curve of progress of labour with average progress during the active phase of 1.1cm per hour and average length of labour of 12 hours for nulliparous women and 6 hours for multiparous women.198 [evidence level 3] More recent observational studies from the USA (n = 2511) measured the length of labour in women who had not received oxytocin or epidurals and report average length of labour for nulliparous women to be 19.4 hours and 13.7 hours for multiparous women. This is longer than the originally described normal labour curve.199 [evidence level 3]

On a partogram cervical dilatation and descent of the presenting part are plotted graphically against time. The partogram was initially proposed as a screening tool for use in poorly resourced countries to identify women who needed referral to hospital. The partogram includes two lines, an alert line and an action line. The alert line is set at a rate of 1cm per hour (derived from Friedman's curve). The action line is drawn 4 hours to the right of the alert line. If the progress of labour crossed the action line women were referred to hospital for either augmentation of labour or CS.200,201 [evidence level 3]

Three RCTs have evaluated the use of partograms in the management of labour. The first RCT compared using a partogram with a four hour action line to not using a partogram in the management of labour. This was a cluster randomised trial where the unit of randomisation was a maternity hospital. Four pairs of hospitals participated. Each hospital had a practice of active management of labour including oxytocin use. The effect of the partogram was analysed in a before and after design which compared labour outcome data on 10,049 women who delivered before implementation of the partogram (4 hour action line) with data on 9130 women who delivered after implementation. This RCT did not report CS rates but did report rates of spontaneous vaginal birth. The number of spontaneous cephalic births was increased after implementation of the partogram (83% vs. 86.3%, p < 0.001). There was a decrease in the proportion of women with labours of more than 18 hours (551 versus 249, p < 0.001), labours augmented by oxytocin (p = 0.041) and the number of intrapartum stillbirths (0.5% vs. 0.31%, p = 0.024). There was no change in the overall duration of labour or other neonatal indices. Similar patterns were noted for multiparous and primiparous women.202 [evidence level 1b]

The second RCT (n = 928 women) compared partograms with different action lines (either 2, 3 or 4 hours to the right of the alert line set at 1 cm per hour). The primary outcomes were CS rate and maternal satisfaction. CS rate was lowest when labour was managed using a partogram with a 4 hour action line. Women in the 2 hour arm were most satisfied with their labour experience. No difference was found in the secondary outcomes of neonatal and maternal morbidity.203 [evidence level 1b]

The third RCT conducted in South Africa (n = 694) compared management using a single alert line partogram offering oxytocin if the alert line was crossed (with 2 hour vaginal examinations) to management using a 4 hour action line. CS was a primary outcome. Women in the intervention group were less likely to have a CS (RR 0.68, 95% CL 0.50 to 0.93).204 [evidence level 1b]

Meta-analysis of the 2 RCTs that included comparison of the two hour action line with a four hour action line partogram showed no difference in CS rate between the use of 2 or 4 hour action lines (RR 0.93, 95% CI 0.48 to 1.78).203,204 [evidence level 1b] The use of a 4 hour partogram reduces the number of vaginal examinations that women would undergo during labour.


44A partogram with a 4-hour action line should be used to monitor progress of labour of women in spontaneous labour with an uncomplicated singleton pregnancy at term, because it reduces the likelihood of CS. [A] [2004]
NumberResearch recommendation
RR 22RCT evidence is needed to determine the impact of partograms based on different curves of labour on CS rates and morbidity outcomes.

Decision making for unplanned CS

Second opinion has been proposed as an intervention to decrease CS rates. Second opinion refers to a doctor needing the agreement of another usually more senior second opinion before a decision for CS can be made. A large multi centred RCT in five South American countries has recently been completed however the results have not been reported.

Using the NSCSA data the proportion of CS cases with consultant involvement varied between maternity units, although in the majority of CS, the consultant was the most senior obstetrician involved in the decision (see table).

In maternity units where consultant obstetricians were frequently involved either in the decision for CS or present in theatre for “emergency” CS the crude and adjusted CS rates (having taken into account case mix differences) were lower (see Tables 6.1 and 6.2).

Table 6.1. Proportion of CS with consultant involvement in maternity units.

Table 6.1

Proportion of CS with consultant involvement in maternity units.

Table 6.2. Relationship between proportion of CS where there was consultant involvement and CS rates.

Table 6.2

Relationship between proportion of CS where there was consultant involvement and CS rates.


45Consultant obstetricians should be involved in the decision making for CS, because this reduces the likelihood of CS. [C] [2004]

Electronic fetal monitoring and fetal blood sampling

Systematic reviews of 9 RCTs (conducted between 1976–1993, n = 18,561 women) have compared the use of electronic fetal monitoring (EFM) during labour to intermittent auscultation. No difference is detected in perinatal mortality (RR 0.89, 95% CI 0.60 to 1.33). The use of EFM during intrapartum care results in increased CS rates (RR 1.4, 95% CI 1.23 to 1.61) This increase is less marked if fetal blood sampling (FBS) is used (RR 1.27, 95% CI 1.08 to 1.51 for EFM with FBS, compared with RR 1.41, 95% CI 1.23 to 1.61 for EFM without FBS).205,206 It is therefore recommended that where delivery is contemplated because of an abnormal fetal heart rate pattern, in cases of suspected fetal acidosis, FBS should be undertaken in the absence of technical difficulties or any contraindications. Contraindications to FBS include maternal infection (such as HIV, hepatitis viruses or herpes simplex virus); fetal bleeding disorders such as haemophilia and prematurity (less than 34 weeks). Where there is clear evidence of acute fetal compromise, e.g. prolonged decelerations (longer than 3 minutes), FBS should not be undertaken and the baby should be delivered urgently.2

The NSCSA measured practice against this audit standard for CS.4 Overall an abnormal CTG was noted in 69% of singleton cephalic pregnancies delivered by CS for presumed fetal compromise. If the CTG was noted to be severely abnormal or cervical dilatation was less than 4cm these cases were not included (50%). Overall practice concorded with the audit standard in 44% of cases. However there was marked variation in practice. Five percent of maternity units met the standard in all cases (100%), in 9% the standard was not reached for any case. Units and regions which used FBS more frequently before CS had lower CS rates. Overall, cases in which this recommendation was not met contributed 4.6% to the overall CS rate or about 1% of all births.4


46Electronic fetal monitoring is associated with an increased likelihood of CS. When CS is contemplated because of an abnormal fetal heart rate pattern, in cases of suspected fetal acidosis, fetal blood sampling should be offered if it is technically possible and there are no contraindications. [B] [2004]

6.3. No influence on likelihood of CS

The following interventions during intrapartum care have not been shown to influence the likelihood of CS. These interventions may have other effects (beneficial or harmful) which are outside the scope of this guideline and are not considered here.

Walking in labour

Two RCTs have evaluated the effect of walking in labour to usual care, one conducted in the UK (n = 68)207 [evidence level 1b] and the other conducted in the USA (n = 1067)208 [evidence level 1b]. No difference was detected in the CS rates between women who walked around during labour and those who did not (RR 0.71, 95% CI 0.43 to 1.20). Most of the weight of the pooled RR in the meta-analysis comes from the larger RCT. Therefore it is not surprising that the US RCT did not detect a difference in CS rates between groups (RR 0.73, 95% CI 0.43 to 1.24). The study has 80% power to detect a difference of at least 4% in CS rate, therefore if walking in labour has an impact on CS rates it is likely to be less than 4%. The RCT did not detect a difference in other outcomes including length of the first stage of labour and need for analgesia. The results were similar for parous and multiparous women.208 [evidence level 1b]

Position in the second stage of labour

A systematic review209 of 18 RCTs evaluated the effect of different positions for the second stage of labour. No difference was detected between any upright or lateral position during second stage on CS rates compared to supine or lithotomy positions (12 RCTs; n = 2250; RR 0.87, 95% CI 0.52 to 1.45). Use of any upright or lateral position, compared with supine or lithotomy positions, was associated with the following: reduced duration of second stage of labour (12 RCTs. Weighted mean difference: 5.4 minutes, 95% CI 3.9, 6.9 minutes); a reduction in assisted deliveries (17 RCTs. OR 0.82, 95% CI 0.69 to 0.98); a reduction in episiotomies (11 RCTs: OR 0.73, 95% CI 0.64 to 0.84); an increase in second degree perineal tears (10 RCTs: OR 1.30, 95% CI 1.09 to 1.54); increased estimated risk of blood loss greater than 500 ml (10 RCTs: OR 1.76, 95% CI 1.34 to 3.32); reduced reporting of severe pain during second stage of labour (1 RCT: OR 0.59, 95% CI 0.41 to 0.83) and fewer abnormal fetal heart rate patterns (1 RCT: OR 0.31, 95% CI 0.11 to 0.91).209 [evidence level 1a]

Immersion in water in labour

Water births and the use of immersion in water during labour comprise 0.6% of births in the UK.210 [evidence level 3] A systematic review211 [evidence level 1a] that included three RCTs (n = 988) compared immersion in water during labour (no births occurred in the water) to conventional care. Another RCT (n = 1237) on this topic has been published since this review.212 [evidence level 1b] The CS rate in the intervention arm of these RCTs ranged from 1.8% to 8.9%, in the control group it ranged from 0% to 7.9%. A new meta-analysis of the findings from these 4 RCTs (n = 2225) did not detect a difference in CS rates between the two groups (RR 1.31, 95% CI 0.89 to 1.93) [evidence level 1a]. Overall these studies have a 90% power to detect a difference of at least 4% in CS rates between the two groups therefore if water birth has an effect on CS rate it is likely to be less than 4%.

One of the above RCTs interviewed a subset of women about their use and satisfaction with care in labour. Women most liked the presence of a support person and immersion in water.213

A national cohort study using regional UK survey data compared the perinatal mortality and morbidity of 4032 births either in water (or following labours in water) to births not in water. They report no difference in perinatal mortality (RR 0.9, 99% CI 1.2 to 3.6). There were two cases of water aspiration which required admission to NICU.214 [evidence level 3] A prospective observational study in Switzerland of 7508 births of which 2014 were water births showed no increased risk for women or their babies. The study reported: lower episiotomy rates, higher rates of intact perineum, lower blood loss and lower use of pain killers in women who had a waterbirth.215 [evidence level 3]

A number of position papers have provided guidelines for water births in the absence of adequate evidence, and have suggested the continued reporting of adverse events.216,217 [evidence level 4]

Analgesia during labour

There has been an increase in the use of epidural analgesia in labour and there has been concern that this may have contributed to an increase in CS. Observational data provides conflicting results.218227 [evidence level 3]

Two systematic reviews have included RCTs of women in spontaneous labour who requested analgesia and were randomised to receive either epidural analgesia or usual analgesia (such as intravenous or intramuscular pethidine). The first review of 10 RCTs (n = 2369) did not detect a difference in CS rates between the two groups (OR 1.5, 95% CI 0.81 to 2.76).228 [evidence level 1a]. A subsequent review includes 11 RCTs (n = 3157, it includes 6 RCTs from the previous review, 2 new RCTs229,230 and 2 RCTs not included in the first review231234). It also did not detect a difference in CS rates (OR 1.30, 95% CI 0.93 to 1.83).235 [evidence level 1a]

We did not identify any RCTs that had compared parenteral analgesia (intravenous or intramuscular opiate derived analgesia) to placebo or complementary therapies on mode of birth and risk of CS.

Raspberry leaf during labour

An RCT (n = 192) was conducted that looked at the use of raspberry leaf, given in tablet form during labour. No difference was detected in length of labour or mode of birth, including “emergency” CS236 [evidence level 1b]. Earlier descriptive studies of raspberry leaf used in labour excluded women who had a CS from their analysis.237 [evidence level 3]


47Women should be informed that the following interventions during intrapartum care have not been shown to influence the likelihood of CS, although they may affect other outcomes that are outside the scope of this guideline:
  • walking in labour
  • non-supine position during the second stage of labour
  • immersion in water during labour
  • epidural analgesia during labour
  • the use of raspberry leaves. [A] [2004]
NumberResearch recommendation
RR 23RCT evidence is required to evaluate the effect of parenteral analgesia (intramuscular and intravenous morphine based analgesia) used during childbirth on the likelihood of CS.

Complementary therapies during labour

Complementary therapies used during pregnancy include acupuncture, aromatherapy, hypnosis, Chinese medicines, herbal products and nutritional supplements, homeopathic medicines and raspberry leaf (discussed previously). We have only considered their use during labour in this guideline. The antenatal use of complementary therapies is included in the NICE Antenatal Care Guideline.1

We identified a systematic review of complementary therapies for pain management in labour which includes seven RCTs (n = 366) using different modalities of pain management238 [evidence level 1a]. CS rates were considered as secondary outcomes in two of the included studies: one RCT using acupuncture (n = 90), one aromatherapy RCT (n = 22), neither showed any difference in CS rates however the trials were underpowered to evaluate this outcome. Two RCTs (n = 125) have compared the use of hypnosis to usual analgesia. CS was not reported. However women in the hypnosis group were more likely to have a spontaneous vaginal birth (RR 1.38, 95% CI 1.13 to 2.47).238 [evidence level 1b]

A large survey (n = 8058) of women views on the effect of using aromatherapy during labour. Effect was measured using a Likert scale. About half of the women reported aromatherapy was helpful, a minority (14%) found it unhelpful.239 [evidence level 4]

The suggested benefits of Chinese medicines in labour include prevention of nausea and vomiting, heartburn and fatigue. We did not identify any RCTs on their use in labour. We identified a cohort study on the use of Chinese medicines during pregnancy which reported no effect on mode of birth.240 [evidence level 2b]

Surveys from the USA and Australia suggest that there is widespread use of herbal products and nutritional supplements during pregnancy, 12% of women in Australia241 [evidence level 4] and 7% in the USA.242 [evidence level 3] A UK survey of midwives estimated that 34% of midwives offer some form of complementary medicine to women during pregnancy or childbirth.243 [evidence level 4] The majority of this use is antenatal with only certain herbal products used during labour or to induce labour. We did not identify any RCTs on the use of herbs during labour but a number of expert opinion papers offer advice and suggested guidelines for their use. Using information from midwives surveys they recommend caution with the use of blue cohosh (due to reports of dizziness, fainting, nausea and meconium stained liquor as well as case reports of neonatal heart failure); black cohosh and castor oil to induce labour.244 [evidence level 4] There have not been reported complications with either evening primrose oil or raspberry leaf.245,246 [evidence level 4]


48Women should be informed that the effects on the likelihood of CS of complementary therapies used during labour (such as acupuncture, aromatherapy, hypnosis, herbal products, nutritional supplements, homeopathic medicines, and Chinese medicines) have not been properly evaluated and further research is needed before such interventions can be recommended. [D] [2004]
NumberResearch recommendation
RR 24RCTs are needed to establish the safety and efficacy of complementary therapies used during labour.

6.4. ‘Failure to progress’ in labour and CS

In the NSCSA, “failure to progress” in labour (FTP) was the primary indication for CS in 35% (n = 4896) of women with term cephalic pregnancies and no uterine scar. For 17% (n = 811) of these women cervical dilatation at the time of CS was less than 4 cm. While 74% of these women had their labour augmented (65% were given oxytocin and amniotomy, 7% amniotomy only, 2% oxytocin only), 24% (n = 193) had no augmentation of labour before CS. The majority (98%) of women with cervical dilatation of at least 4 cm at the time of CS had either amniotomy or oxytocin or both. Twenty-five percent (n = 1231) of CS for FTP were done at a cervical dilatation of 10 cm, 28% (n = 345) of these women did not have oxytocin before CS. These cases in which labour augmentation with oxytocin was not used contributed 3.2% to the overall CS rate.4 [evidence level 3]

We searched for research that evaluated the impact of packages of interventions such as active management of labour and interventions such as routine amniotomy or oxytocin infusion used together or alone are included.

Active management of labour

Active management of labour refers to a labour ward protocol that includes routine amniotomy and early augmentation with oxytocin as well as strict criteria for the diagnosis of labour, abnormal progress in labour and fetal compromise. It also includes the continual presence of a midwife or support person during labour and peer review of assisted deliveries. Observational studies by the initiators of active management reported lower CS rates, reduction in the number of women having prolonged labour, better neonatal outcomes and improved maternal satisfaction.247 Subsequent observational studies did not replicate these findings.248,249 It has remained an area of controversy.250 [evidence level 3]

A systematic review of 10 RCTs (n = 5111) evaluated the effects of a package intervention of early augmentation of labour with amniotomy and oxytocin in nulliparous women compared to usual care (‘care at the discretion of the individual doctor/midwife attending the woman in labour’). Overall there was no reduction in the likelihood of CS with early amniotomy and early oxytocin infusion (OR 0.9, 95% CI 0.7 to 1.1). Subgroup analysis of the therapy RCTs (recruited women in whom a delay in progress was diagnosed) (3 RCTs, n = 109) and prevention RCTs (7 RCTs, n = 5002) were undertaken. No difference in CS rate was apparent in these subgroups. However the therapy subgroup is too small and is therefore underpowered to evaluate this outcome.251 [evidence level 1a] None of the RCTs had maternal satisfaction as an outcome measure.251 [evidence level 1a]

A recently published RCT from South Africa (n = 694) compared using a single line partogram, two-hourly vaginal examinations and use of oxytocin if the partogram line was crossed in nulliparous women to usual management (4 hour vaginal examinations). CS rates in the intervention group were lower (RR 0.68 95% CI 0.50 to 0.93) Analysis is by intention to treat but it was noted that there was a high proportion of protocol violations in both groups (about 30%)204 [evidence level 1b]. It was not possible to include this RCT with the earlier RCTs as the descriptions of management of labour were not consistent.


Most RCTs identified incorporate the use of oxytocin into active management of labour. However we identified one RCT (n = 60) that looked at the effect of oxytocin without other components of active management of labour in women in whom there was a delay in labour progress. Women whose cervical dilatation was less than 0.5 cm per hour were randomised to one of three groups: group one – oxytocin was deferred for 8 hours; group two – low-dose oxytocin infusion (2mu/minute) or group three – high-dose oxytocin (7mu/minute). The CS rates between the three groups were not statistically different (45%, 35% and 26% respectively χ2 1.6346 2df). There were no differences between the groups in terms of neonatal outcomes.252 [evidence level 1b] This RCT is underpowered to assess these outcomes.

Observational data from the original active birth management study suggested benefit of the early use of high dose oxytocin infusions.247 [evidence level 3] Subsequent observational studies that looked at the use of oxytocin alone in labour suggested that it decreased the CS rates253 and did not result in increased neonatal morbidity.254,255 [evidence level 3]


A systematic review of nine RCTs looked at the impact of early routine amniotomy.256 CS rate was reported in 8 of the included RCTs (n = 4008). No difference in CS rates was found between early routine amniotomy and no routine amniotomy (OR 1.26, 95% CI 0.96 to 1.66). Amniotomy was associated with a reduction in labour duration of between 60 and 120 minutes, reduction in the likelihood of 5 minute Apgar of less than 7 (OR 0.54, 95% CI .0.30 to 0.96) and a decrease in the use of oxytocin (OR 0.79, 95% CI 0.67 to 0.92). Groups were similar with respect to other neonatal indicators.256 [evidence level 1a]

Operative delivery in the second stage

Four percent (n = 1203) of all CS were performed for failure to progress in the second stage of labour (in women without a previous CS who had a term singleton cephalic infant). In the majority 55% (n = 661) no other method of delivery had been attempted before CS. In 35% (n = 427) of these occurrences, CS followed a failed attempt at ventouse, in 7% (n = 81) both ventouse and forceps had been attempted prior to CS and in 2% (n = 27) CS followed a failed attempt at forceps delivery. Overall in the UK while CS rates have increased, operative vaginal delivery rates have remained relatively constant (about 10–11%).4,257 [evidence level 3] However there has been a marked reduction in the use of forceps and an increase in the use of ventouse since the early nineties.4,257 [evidence level 3] Within RCTs the use of ventouse is associated with an increase in failure to achieve a vaginal delivery but it is not associated with a concomitant increase in CS rates.258,259 [evidence level 1a]

A cohort study has compared the maternal and neonatal outcomes following either instrumental vaginal delivery or CS in the second stage of labour (n = 393 women, 184 had a vaginal delivery, 209 CS).260 [evidence level 2a] Major haemorrhage (blood loss > 1000 ml) was more common after CS than vaginal delivery (adjusted OR 2.82, 95% CI 1.1 to 7.62). Length of hospital stay was increased with CS. No difference was detected in wound infection, blood transfusion, need for opiate analgesia or rates of breastfeeding. Odds ratios were adjusted for maternal body mass index, pre-eclampsia, maternal diabetes, duration of second stage, station and position of the presenting part, demographic differences and experience of the operator.260 [evidence level 2a] A further study following up the same women after 3 years reported half had achieved a further pregnancy after 3 years. There was no difference the proportion of women who had difficulty conceiving but there was an increase in involuntary infertility of more than 1 year. Of women who choose not to have more children there was no difference in the proportion that stated they “could not go through childbirth again”. Of women who had a further pregnancy those who had had a previous instrumental vaginal birth were more likely to aim for and achieve a vaginal birth again (adjusted OR 15.55, 95% CI 5.25 to 46.04; adjusted OR 9.50, 95% CI 3.48 to 25.97).261 Qualitative research of women views on the impact of operative delivery in the second stage of labour (n = 27) described that women felt unprepared for operative delivery and that antenatal education had not adequately prepared them for this event.262


49The following aspects of intrapartum care have not been shown to influence the likelihood of CS for ‘failure to progress’ and should not be offered for this reason, although they may affect other outcomes which are outside the scope of this guideline:
  • active management of labour
  • early amniotomy. [A] [2004]
NumberResearch Recommendation
RR 25More RCTs are required to determine the effect of oxytocin augmentation as single interventions or as part of a package of interventions (such as “active management of labour”) on the likelihood of CS and other outcomes including women's satisfaction with care.
RR 26Further research on the short and longer term health impacts of CS during the second stage compared to operative vaginal delivery are needed.

Female genital mutilation

Female genital mutilation is defined by the World Health Organization (WHO) as, ‘all procedures that involve partial or total removal of the female external genitalia or other injury to the female genital organs whether for cultural, religious or other non-therapeutic reasons’. An estimated 10,000 to 20,000 girls in the UK are thought to have undergone genital mutilation.263 [evidence level 3]

The association between female genital mutilation and intrapartum complications has been systematically reviewed by the WHO.264 Possible complications include obstructed labour, fetal distress and increased perinatal mortality however the evidence for these are contradictory.264266 [evidence level 3] 267 [evidence level 2a] No RCTs or observational studies have addressed the effect on health outcomes of CS in the management of female genital mutilation. It is outside the scope of this guideline to address the antenatal or intrapartum management of female genital mutilation.

6.5. Eating during labour

The practice of encouraging women to eat and drink during labour in order to maintain their strength for the second stage changed following publication of a case series (n = 66) of aspiration pneumonitis. In this paper Mendelson suggested that mortality due to aspiration pneumonitis could be reduced if women did not eat and drink during labour.268,269 [evidence level 3] This work continues to influence practice both in the UK and elsewhere. In the UK less than 5% (12/268) maternity hospitals have a policy of unrestricted intake during labour,270 this is also usual practice in many other countries.271,272 [evidence level 3] An exception to this is the Netherlands where a survey reported that the majority of obstetricians and midwives had an unrestrictive policy on fluid and food intake. The Netherlands do not have a higher mortality rate due to aspiration pneumonitis than other countries.273 [evidence level 3] A UK survey of women's views about eating in labour reported that 31% of women said that would have liked to have eaten during labour.274 [evidence level 3] Many historical overviews, comments, surveys or non-systematic literature reviews have been written discussing the benefits and harms of eating during labour.275277 [evidence level 3]

One RCT (n = 94) compared offering a low residue diet of toast cereal, crackers and low fat cheese during labour to offering a range of drinks to women during labour (water, tea, coffee, cocoa). Women included in the trial were in spontaneous labour, at term with singleton cephalic presentation and who did not request parenteral opioids (because opioids can delay gastric emptying). Outcome measures used were women's metabolic profile, volume of gastric contents as well as labour outcomes such as length of labour, use of oxytocin and mode of birth.274 [evidence level 1b] Women who had a low residue diet were less likely to have ketosis and had higher plasma glucose at the end of labour than women in the drinks only group. Gastric contents were significantly higher in those eating a low residue diet and these women were more likely to vomit at birth, vomit higher volumes and vomit more solid material. Higher gastric volumes could be of importance if unexpected general anaesthesia was needed. No differences were detected in labour outcomes between the two groups but the study is underpowered to evaluate these outcomes.278 [evidence level 1b] This issue is currently being evaluated in another RCT.279

A further RCT (n = 60) compared drinking an isotonic drink to drinking water only during labour. Metabolic indices and gastric volumes were measured. Isotonic drinks reduced ketosis but did not increase gastric volume. There was no change in labour outcomes but the study was underpowered to assess these outcomes.280 [evidence level 1b]


50Women should be informed that eating a low-residue diet during labour (toast, crackers, low-fat cheese) results in larger gastric volumes, but the effect on the risk of aspiration if anaesthesia is required is uncertain. [A] [2004]
51Women should be informed that having isotonic drinks during labour prevents ketosis without a concomitant increase in gastric volume. [A] [2004]
NumberResearch recommendation
27RCTs that evaluate the effects of eating during labour compared with restricting intake on labour outcomes are needed. Cohort or case control studies on the risk factors for aspiration and other morbidities for women having CS are needed.
Copyright © 2011, 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 []. 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.

Bookshelf ID: NBK115300
PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (2.7M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...