Standard- or high-dose oxytocin for nulliparous women with confirmed delay in labour: quantitative and qualitative results from a pilot randomised controlled trial


  • Correction added on 4 July 2013, after first online publication: author affiliations for S McNicol and J Willars were incorrect, these have now been corrected.



Evidence suggests that a high dose of oxytocin for nulliparous women at 37–42 weeks of gestation with confirmed delay in labour increases spontaneous vaginal birth. We undertook a pilot study to test the feasibility of this treatment.


Pilot double-blind randomised controlled trial.


Three teaching hospitals in the UK.


A total of 94 consenting nulliparous women at term with confirmed delay in labour were recruited, and 18 were interviewed.


Women were assigned to either a standard (2 mU/min, increasing every 30 minutes to 32 mU/minute) or a high-dose regimen (4 mU/minute, increasing every 30 minutes to 64 mU/minutes) oxytocin by computer-generated randomisation. Simple descriptive statistics were used, as the sample size was insufficient to evaluate clinical outcomes. The constant comparative method was used to analyse the interviews.

Main outcomes measures

The main outcome measures: number of women eligible; maternal and neonatal birth; safety; maternal psychological outcomes and experiences; health-related quality of life outcomes using validated tools and data on health service resource use; incidence of suspected delay of labour (cervical dilatation of <2 cm after 4 hours, once labour is established); and incidence of confirmed delay of labour (progress of <1 cm on repeat vaginal examination after a period of 2 hours).


We successfully developed systems to recruit eligible women in labour and to collect data. Rates of spontaneous vaginal birth (10/47 versus 12/47, RR 1.2, 95% CI 0.6–2.5) and caesarean section (15/47 versus 17/47, RR 1.1, 95% CI 0.6–2.0) were increased, and rates of instrumental birth were reduced (21/47 versus 17/47, RR 0.8, 95% CI 0.5–1.3). No evidence of increased harm for either mother or baby was found. The incidences of suspected delay (14%) and confirmed delay (11%) in labour were less than anticipated. Of those who did not go on to have delayed labour confirmed, all except one woman gave birth vaginally.


A pilot trial assessing the efficacy of high-dose oxytocin was feasible, but uncertainty remains, highlighting the need for a large definitive trial. The implementation of national guidance of suspected and confirmed delay in labour is likely to reduce intervention.


The current standard regimen of oxytocin for women whose labour requires augmentation for delay starts with a dose of 1–2 mU/minute, and increases over intervals of 30 minutes or more to a maximum of 32 mU/minute. It was developed for the Royal College of Obstetricians and Gynaecologists' (RCOG) Induction of Labour Guideline,[1] but has never been evaluated by clinical trials.

For women whose labour requires augmentation, a review of current oxytocin regimens was undertaken for the National Institute for Clinical Excellence (NICE) Intrapartum Care Guideline.[2] This found that the use of oxytocin in these relatively low doses was associated with a shorter labour, but without any change in caesarean or vaginal birth rates. It is plausible that this lack of effect of oxytocin arises from an inadequate dose or from a delay in achieving effective doses.

Two recent systematic reviews have suggested that a higher-dose regimen of oxytocin may reduce the risk of caesarean section and increase spontaneous vaginal birth for women who require augmentation for delay in labour.[3, 4] Although both reviews reported increased hyperstimulation in the high-dose groups, neither maternal nor neonatal morbidity were increased. The authors of both reviews highlight the uncertainty around the evidence, the paucity and quality of the existing studies, and insufficient data on neonatal outcomes and maternal birth experience, making this a key priority for further research, as identified by the NICE Intrapartum Care Guideline,[2] and by the RCOG Intrapartum Clinical Study Group (

Evidence from previous trials that have recruited women in labour demonstrates the challenges of conducting studies under these circumstances. While the provision of written information is an integral part of the formal recruitment process to any trial, the extent to which this serves the intended function of ensuring those approached have a full understanding of the trial they are being asked to join is not always clear.[5] Some studies report that participants would have liked more information,[6] and even when considerable care has been taken in the design of written information, including the involvement of consumer representatives, participants can still understand and interpret the trial in ways that are not entirely in accordance with the aims the investigators had when the put the material together.[7]

Women are likely to feel emotional and vulnerable during labour, and may rely heavily on the interpersonal relationships that they have established with those caring for them when making a decision about participation, rather than the standard written information they receive.[8] Although this is also the case for studies in other areas,[9] these tensions are likely to be felt even more keenly when the woman's unborn child is involved,[10] and so the training of clinicians for taking consent is of vital importance. This pilot provided an opportunity to explore in practice the recent advice from RCOG on obtaining consent for research in labour,[11] which suggests that for conditions with the risk of occurrence from one in ten to one in 100 (e.g. delay in labour or fetal blood sampling), a trial outline is provided to potentially eligible women antenatally, with access to additional information if requested. Full trial information should then be provided to women in labour who become eligible.


We conducted a pilot trial of a standard versus a higher-dose regimen of oxytocin in nulliparous women where delay in the first stage of labour was diagnosed strictly using the 2007 NICE definitions. The pilot took place between November 2010 and May 2011 in three maternity units: Birmingham Women's Hospital (BWH; 7400 births annually); Liverpool Women's Hospital (LWH; 8500 births annually); and Royal Victoria Infirmary (RVI; 5300 births annually), Newcastle. Ethical approval was obtained from the National Research Ethics Service in the UK. The pilot planned to develop and evaluate:

  1. numbers of women eligible (nulliparous women at term with delay in labour, as defined by the NICE Intrapartum Care Guideline,[2] and with membranes ruptured);
  2. numbers who agree to be randomised;
  3. women's understanding and experiences of the study, the consent process, and their decision to participate;
  4. support required in units to ensure successful recruitment;
  5. data collection tools and systems, including validated tools to assess pain, maternal support, and control in labour,[12] and mother–infant bonding,[13] as it is plausible that these outcomes could be effected by high-dose oxytocin;
  6. maternal questionnaires (at 6 weeks postpartum) to capture hospital and community health services usage and EuroQol EQ-5D™ data.[14]

Inclusion and exclusion criteria

Eligible women included:

  1. consenting nulliparous women with a singleton pregnancy at term (37–42 weeks of gestation);
  2. confirmed delay in spontaneous labour, as defined by the NICE Intrapartum Care Guideline,[2] and with ruptured membranes (as described below).

Following the confirmation of labour being established (i.e. regular painful contractions and progressive cervical dilation, from 4 cm), delay was suspected when cervical dilatation of <2 cm in 4 hours occurred. On repeat vaginal examination 2 hours later, delay was confirmed when progress of <1 cm in 2 hours was found. This definition also takes into account the descent and rotation of the fetal head, and the strength, duration, and frequency of contractions. Should delay be confirmed, obstetric review is advised and oxytocin augmentation is discussed.

Exclusion criteria included women undergoing induction, women under 16 years of age, women with a body mass index (BMI) >40, a multiple pregnancy, existing maternal or fetal disease or concern, gestational diabetes, previous uterine surgery, vaginal bleeding of clinical significance in this pregnancy, or a contraindication to oxytocin therapy.


For the pilot trial we adopted the model advised by RCOG,[11] and all potentially eligible women over the period of the study received information at about 34–36 weeks of gestation. Once delay was suspected (defined as cervical dilation of <2 cm in 4 hours), the study was discussed and information provided both in full and in summary forms. Once eligibity had been confirmed, consent was obtained by appropriately trained clinicians.


Women were randomised using an automated telephone randomisation service to either a standard (2 mU/min, increasing every 30 minutes to 32 mU/minute) or a high-dose (4 mU/minute, increasing every 30 minutes to 64 mU/minute) regimen of oxytocin. Both women and clinicians were blinded to regimen allocation, achieved by the use of identical vials produced by a Medicines and Healthcare products Regulatory Agency Investigational Medicinal Product-approved manufacturer (Bilcare). A computer-generated program ensured balanced allocation to the two arms, and allocation was stratified by maternity unit.

Clinical outcome measures

The primary outcome was the rate of spontaneous vaginal birth (SVB). The size of the pilot study (94 women) did not allow for reliable assessment of the effect of the intervention on clinical outcomes. Potential neonatal components of any composite primary outcome were explored during the pilot study.

Safety outcomes

As there are known associations with oxytocin, we collected data regarding: uterine tachysystole (defined as more than five uterine contractions in 10 minutes for longer than 20 minutes); uterine hyperstimulation (defined as more than five uterine contractions in 10 minutes for longer than 20 minutes, with suspicious or pathological fetal heart rate); and maternal positive fluid balance, potentially leading to urinary retention or pulmonary oedema, possible fetal hypoxia requiring fetal blood sampling (FBS), and neonatal hyperbilirubinaemia requiring phototherapy. Data recording of pre-specified adverse outcomes (tachysystole, hyperstimulation, FBS with hyperstimulation being present, and pathological CTG leading to immediate birth without FBS) were recorded on bespoke data-collection forms during labour, and were then extracted by the chief investigator (S.K.).

Less seriously, oxytocin may also occasionally cause nausea, vomiting, haemorrhage, or cardiac arrhythmias. In a few cases, skin rashes and anaphylactoid reactions associated with dyspnoea, hypotension, or shock have been reported in the summary of product characteristics (

We also defined events that were serious, and that required additional attention (provided by the Oversight Group). These events were: maternal anaphylaxis; maternal pulmonary oedema; uterine rupture/hysterectomy; massive postpartum haemorrhage (>2 l); maternal admission to a high dependency unit or intensive therapy unit (HDU/ITU); neonatal arterial cord pH of below 7.05 at birth; admission for neonatal intensive care; neonatal fits or seizures; neonatal encephalopathy; and maternal or neonatal death.

Maternal secondary outcomes included: analgesic use during labour (type and dose); pain, scored using a numerical analogue pain-rating scale; length of first, second, and third stages of labour; time to birth (SVB, instrumental birth, and caesarean section) from randomisation; mode of birth (caesarean section or instrumental birth); reason for and time to caesarean section; total oxytocin dose; time to maximum oxytocin rate and maximum oxytocin dose reached; degree of perineal trauma; incidence of other possible morbidity (postpartum haemorrhage, shoulder dystocia, chorionamnioitis); and length of stay in hospital.

Neonatal outcomes included: Apgar scores at 1 and 5 minutes; arterial cord blood gases, when collected; breastfeeding rates; admission to neonatal unit (NNU); and length of stay in hospital.

Psychological outcomes were collected using a questionnaire 2 weeks after birth containing validated tools to assess maternal support and control in labour,[12] and mother–infant bonding.[13]

A qualitative approach was adopted to explore women's understanding and experiences of the study, the consent process, and their decision to participate. All women who took part in the pilot study were invited to take part in an interview by means of an invitation sent 2 weeks after the birth alongside follow-up questionnaires. Although purposive sampling was planned (to include participants from both study arms, those who had a caesarean section, an instrumental birth, or spontaneous vaginal birth, and those who did or did not have an epidural during labour), a lower than anticipated recruitment rate to the qualitative part of the study meant that we interviewed all women who replied to say that they were willing.

Health economic data were collected using a questionnaire at 6 weeks after birth. The questionnaire collected two types of economic data to inform a future trial-based economic evaluation. Firstly, women's health-related quality of life at 6 weeks was described using the EurQol EQ-5D measure.[14] The EQ-5D is the generic, multi-attribute, preference-based measure of health-related quality of life preferred by NICE for cost-effectiveness comparative purposes. Secondly, the 6–week questionnaire also collected a comprehensive profile of hospital and community health and social service use during the period between the initial hospital discharge of mother and baby and 6 weeks post-randomisation. This included the mother's and baby's use of hospital inpatient, daycare, and outpatient services, community health and social services (by location), and prescribed medications. It also included a validated tool to assess mother–infant bonding.[13]

Collection of data

A randomisation form and data-collection forms were developed and refined to collect maternal and perinatal outcome data, both during labour and from notes afterwards. Data regarding pain in labour was collected using a numerical analogue scale completed every 4 hours in labour.

Safety reporting processes and forms were developed and refined during the pilot study to ensure appropriate, accurate, and timely reporting, and a review of any adverse events.

Questionnaires were designed and evidence-based methods were employed to maximise the response rates. These included the use of a £5 high street voucher at the reminder stage of the second questionnaire.

For the interviews, a semi-structured topic guide with broad topic areas was developed from a literature review, discussions within the project team, and with input from our consumer representative (CM). Although this was used to guide the interviews, the emphasis was on encouraging women to discuss their own perspectives freely. Written consent was taken at the interview itself. The interviews were carried out by an experienced qualitative interviewer (J.W.) and, with permission, all of the interviews were audio-recorded and transcribed verbatim.

Those collecting and entering outcome data and interviewing women were blinded to allocation.


The pilot High Or Low Dose Syntocinon (r) Study (HOLDS) was not powered to detect a treatment effect, and as such the focus of the analysis was on the provision of information for the subsequent main trial. Basic descriptive statistics are given, and where treatment effects are calculated, they are presented with 95% confidence intervals (95% CIs). It was not expected a priori that any formal hypothesis testing would take place.

For continuous secondary outcomes, means and standard deviations were calculated. Ranges of plausible treatment effects were calculated using a Student's t–test. For dichotomous outcomes, percentages for each arm, relative risks (RRs), and 95% CIs were produced. Where possible, the median time to the event is given for each arm.

For maternal psychological and health economic outcomes simple descriptive analyses were used. Responses to the five-dimension descriptive system from the EQ–5D measure were converted into health utility scores using the national tariff set for the UK (York A1).[15]

Qualitative data from the interviews were analysed using the constant comparative method,[16] using nvivo. Transcripts were read in detail and open codes were initially applied line by line to the data. The open codes were then incrementally grouped into organising categories or themes. These categories were modified and checked constantly as further open codes were incorporated as analysis proceeded. The categories and their specifications (the coding scheme) were then programmed into nvivo. The coding scheme was used to process the data set systematically by assigning each section of text to a category, according to the category specifications.


Clinical outcomes

Audit data from electronic systems at two of the maternity units (LWH and RVI), and from a case note audit in the third unit (BWH), were used to determine the numbers of women with confirmed delay requiring oxytocin. The incidence of confirmed delay in spontaneously labouring nulliparous women giving birth at >37 of weeks gestation to a singleton from the three units was approximately 11%, and this was similar across all units (Table 1).

Table 1. Numbers of women with confirmed delay in labour
Maternity unitTotal birthsSpontaneously labouring nulliparous births at >37 weeks of gestationConfirmed delay in labour (%)Recruited to HOLDS (%)
RVI2855 (5 months)1366167 (12)20 (12)
LWH4719 (7 months)1937206 (11)29 (14)
BWH1950 (3 months)63465 (10)20 (31)
Total95243934438 (11)69 (16)

The determination of the numbers of women with suspected delay was more challenging, as this was not routinely collected. Prospective audit data from BWH provided this data. This audit was undertaken over a 3–month period (December 2010–February 2011), and ascertained that out of 1950 total births, 634 were to spontaneously labouring nulliparous women at >37 weeks of gestation. Of those, 88 (14%) had suspected delay and 65 (10%) had confirmed delay. Nine women who were included in the audit had delay confirmed with no 2–hour window of suspected delay: the reasons for this are unclear.

This evidence suggests that 32 of the 88 women who had suspected delay made progress in the subsequent 2 hours, and did not have their delay confirmed. Their mode of birth was one caesarean section, 11 instrumental births, and 20 SVGs. All babies had an Apgar score of over seven at 5 minutes; of those who had an umbilical cord pH taken (n = 15), none had a pH < 7.05. All babies were discharged home with their mothers.

In total, 94 women (of a proposed 100) were consented and recruited to the pilot study between November 2010 and May 2011 from the three maternity units (BWH, 45; LWH, 29; RVI, 20). Of these, 47 were allocated the standard dose and 47 were allocated the high-dose regimen (Figure 1). The baseline characteristics of the participants were broadly similar at randomisation (Table S1).

Figure 1.

Consort diagram.

Table 2 shows mode of birth, with SVB marginally increased in the high-dose arm (10/47 versus 12/47, RR 1.2, 95% CI 0.6–2.5). The dose of oxytocin given is shown in Table 3, and shows women allocated the high-dose regimen received a smaller total volume of oxytocin, within a shorter time to the maximum dose, as well as an increased total dose of oxytocin. The first stage of labour was also reduced in the high-dose arm (Table S2).

Table 2. Mode of birth by treatment arm
 StandardHighRR (95% CI)
Number randomised4747 
Mode of birth
Spontaneous vaginal birth (SVB) n (%)10 (21)12 (26)1.2 (0.6–2.5)
Instrumental birth n (%)21 (45)17 (36)0.8 (0.5–1.3)
Abnormal CTG16 8 
Delay in second stage 5 8 
Poor maternal effort 0 0 
Abnormal CTG and delay in second stage 0 1 
Total caesarean section n (%)15 (32)17 (36)1.1 (0.6–2.0)
Caesarean section n (%)13 (28)15 (32)1.2 (0.6–2.2)
Caesarean section following failed instrumental birth n (%) 2 (4) 2 (4)1.0 (0.1–6.8)
Not known (withdrew) n (%) 1 (2) 1 (2) 
Reason for caesarean section
Fetal distress 812 
First stage 3 6 
Second stage 4 6 
Unclear 1 0 
Failure to progress 7 5 
Delay in first stage 5 4 
Delay in second stage 2 1 
Manual rotation attempted n (%) 4 (9) 3 (7)0.8 (0.2–3.2)
Table 3. Oxytocin given by treatment arm
 Standard dose of oxytocinHigh dose of oxytocinDifference/RR (95% CI)
Total oxytocin volume (ml) mean (SD)13.4 (11.9)9.2 (8.5)–4.2 (–8.4, 0.1)
Total oxytocin dose (iu) mean (SD)2.7 (2.4)3.7 (3.4)1 (–0.2, 2.2)
Max. oxytocin rate (ml/hour) mean (SD)3.4 (2.2)2.3 (1.5)–1.1 (–1.8, –0.3)
Max. oxytocin rate (iu/hour) mean (SD)0.7 (0.4)0.9 (0.6)0.3 (0.0, 0.5)
Time to max. oxytocin dose (mins) median (IQR)120.0 (60, 180)75 (30, 150)–30.0 (–71.4, 11.4)

The classification of urgency of caesarean section was undertaken using national categories.[17] An immediate threat to life of either the mother or the fetus was the most common category of urgency of caesarean section in the standard arm with 6/15 (40%), whereas 4/17 (24%) cases were reported in the high-dose arm. Maternal/fetal comprise not immediately threatening to life was reported as being the categorisation of caesarean section for 5/15 (33%) women in the standard-dose arm and 8/17 (47%) women in the high-dose arm. Mothers needing early birth but without compromise was reported for 4/15 (27%) women in the standard arm and 5/17 (29%) women in the high-dose arm.

There was no evidence of an increase in pre-specified maternal adverse outcomes in the high-dose arm (Table 4). One mother in the standard-dose arm was admitted to an HDU, and one baby in the high-dose arm was admitted to an NNU: the reasons were deemed to be unrelated to the dose of oxytocin by the Oversight Group (data not shown).

Table 4. Pre-specified maternal adverse outcomes
MaternalStandard doseHigh dose
Any episode of uterine tachysystole 811
No action11
Reduced oxytocin68
Stopped oxytocin12
Any episode of hyperstimulation 56
No FBS/labour continued34
No FBS/immediate birth01
Caesarean section01
FBS > 7.25 (normal)21
FBS 7.20–7.25 (borderline)
Labour continued
FBS < 7.20 (abnormal)/immediate birth00
Caesarean section00
Any FBS taken, without hyperstimulation being present 1710
FBS > 7.25 (normal)169
FBS 7.20–7.25 (borderline)
Labour continued
FBS < 7.20 (abnormal)/immediate birth11
Caesarean section11
Pathological cardiotocogram, leading to immediate birth without FBS 159
If ‘Yes’ Instrumental105
Caesarean section54
(Caesarean section, had earlier FBS)22
Total FBSs undertaken
Normal >7.252312
Borderline 7.20–7.2512
Abnormal <7.2021

The reasons for the admission of the eight babies to the NNU were reviewed by the Oversight Group, and were found to be unrelated to oxytocin administration, other than two were admitted for phototherapy for jaundice (the other two babies treated for jaundice were not admitted to the NNU).

Selected maternal and neonatal secondary outcomes show no evidence of any difference between the groups (Tables S2 and S3), but overall a relatively high number of women had a postpartum haemorrhage (500–2000 ml, 43/94, 46%).

Maternal psychological outcomes

Maternal perceptions of support and control 2 weeks after birth, and mother–infant bonding collected at both 2 and 6 weeks after birth by questionnaire, showed no difference between the standard and high-dose arms (Tables S4 and S5).

Health economic outcomes

The response rate for the questionnaire at 6 weeks after birth was 58% (n = 56). The mean (SD) health utility score was 0.93 (0.11) in the high-dose oxytocin group, and was 0.88 (0.14) in the standard-dose oxytocin group (where health utilities are expressed on a cardinal scale, with 0.0 representing death and 1.0 representing full health). Only one returned questionnaire contained missing EQ–5D data. Methods for converting health utility data into quality-adjusted life year (QALY) profiles, within the context of a future trial-based economic evaluation, were established.

The questionnaires also collected a comprehensive profile of hospital and community health and social service use during the period between initial hospital discharge of mother and baby and 6 weeks post-randomisation, which included the mother's and baby's use of hospital inpatient, daycare, and outpatient services, community health and social services (by location), and prescribed medications. The most common form of resource use documented in the postal questionnaires was contact with general practitioners: 45 mothers (23 in the standard arm and 22 in the high-dose arm) and 49 babies (25 in the standard arm and 24 in the high-dose arm) had contact with their general practitioners following the initial hospital discharge. A total of 31 mothers (19 in the standard arm and 12 in the high-dose arm) and 12 babies (eight in the standard arm and four in the high-dose arm) had contact with other community health or social care professionals following the initial hospital discharge. Hospital readmissions were rare, with four babies being admitted (three in the standard arm and one in the high-dose arm).

Qualitative outcomes

Interviews took place within 6 weeks of the birth. Although unable to sample purposively, as planned, the sample was diverse, and included: women from the standard (eight) and high (ten) dose arms of the study; women who had a caesarean section (nine), instrumental (seven), or SVB (two); women who had an epidural (12), a spinal for birth (four), or no regional anesthesia (two); and women from all three pilot study sites (BWH, 12; LWH, 3; RVI, 3). Analysis showed that most of the categories of analysis were highly saturated, but without further theoretical sampling it has not been possible to determine whether saturation was reached.

Analysis of the characteristics of the interviewed women compared with those not interviewed showed labour characteristics were similar, but with fewer responses from women of non-European ethnicity and from those who smoked (see Table S6).

Antenatal information

Most women (14/18) said that they first heard of the trial when their labour became delayed – they did not recall receiving the antenatal information. Only four interviewees remembered receiving any information before they went into labour, and typically reported doing little more than ‘glancing over it’, because it did not seem immediately relevant to them.

I think I had, but it'd not registered really, because you don't think, oh hopefully that'll not happen to me, do you know what I mean? So you don't really take any notice. Interviewee 14

Understanding of the trial

Most interviewees (14/18) had at least a basic understanding of the trial. The following is a fairly typical response when asked what they understood the study to be about.

And the midwife just explained to me that basically they were doing an experiment and giving low or high dose, and just to check which one would work the best and if that would make any difference really. I think she did say that they were trying it with women who had been in labour for a while. And that's it really. Interviewee 15

A few women could recall more detailed information about what was involved, such as the name of the drug, the fact that the high dose was double the standard dose, and the fact that it was a double-blind trial. Those who discussed blinding appeared to understand the concept, but a few were unsure as to why it was necessary, and were curious about which dose they had been allocated. There were a few examples of fairly minor misunderstandings: for example, one interviewee thought the trial involved two different drugs rather than different doses of the same drug. Just two interviewees said that they could not recall being told what the trial involved, or said they did not really understand at the time, but trusted the midwives and other staff.

Ability to give informed consent

Although affected to varying degrees by the pain that they were experiencing, the effects of pain relief, and tiredness, the majority of interviewees (14/18) felt that they were able to give informed consent to participate in the HOLDS trial.

Asked ‘Do you feel as though you knew what you were consenting to at the time?’

Yeah, yes, because I don't think I would have consented if I wasn't sure, even though I was at the situation [of being tired, etc.], I still wouldn't have said ‘oh yeah, yeah, whatever’, if I didn't know because at the end of the day it's my baby. Interviewee 17

In many cases, the woman's partner or mother read the information, discussed whether to participate, and helped the woman to make a decision; however, two women felt that they went along with the trial without really questioning what was happening.

I wasn't sort of compos mentis, if you know what I mean, I just was signing everything, and it was at the same time I was signing this, that they gave me the one for theatre, so to be honest I could have been signing anything. I was just signing anything… Interviewee 1

They did not appear to be unduly concerned about this, however, and later in their interviews stated that they were happy they had participated in the study.

There were three interviewees who said they felt pressured in some way to participate, or felt that they had no choice, but it is not always clear-cut whether this related specifically to the trial or to having oxytocin.

I think I was probably told that I needed the syntocinon to help speed things up and would I like to do the trial? I don't think I was sort of – and it's one of those things like obviously when you're told, you're not sort of being asked for your permission, you're being told it's – well you are, but you're not…So this is what you need to have and you say ‘right, OK’. Interviewee 13

Perceptions of the information provided

Most women were happy with the information that had been provided about the HOLDS study. They said it was easy to understand, ‘not full of jargon’, and provided all the information they needed. In terms of quantity, there was a general consensus that not much more information could be dealt with.

Verbal information from health professionals caring for them during labour was felt to be a very helpful addition to the written information, and was often easier to take in.

No I think they did give me some leaflets, but I didn't read it and they talked to me verbally, they sat down with me and spoke, which is better, I think, because yes, you know, they give me, I think I might have had the leaflets or whatever but… Interviewee 10

Reasons for participating

Women gave a number of reasons for their decision to participate in the trial. Three women indicated that they did not think it would increase the level of pain that they were experiencing, and one believed it might decrease the pain. Other comments related to the fact that they thought it would speed up their labour or make it easier, whereas one woman was keen to increase her chance of an SVB. For some (five), previous involvement in research, either as a participant or professionally, was a factor, whereas others (five) spoke about wanting to help women who would be in a similar position in the future, or wanting to help researchers or hospital staff.

Although participants identified several possible benefits that may come from their participation, it was clear that several participants could see no possible harm that may come from their taking part in the study, either to themselves or to their unborn child. These beliefs appeared to be rooted in the fact that they were going to be receiving oxytocin anyway, regardless of whether or not they chose to participate. They did not appear aware of any possible risk of receiving a higher dose, despite this being set out in the information leaflet.

Well at the time they were going to start [oxytocin] anyway, regardless of whether I was taking part in the study it would have been, and I think to me it's neither here nor there whether I was getting [high or standard dose], because surely the reason you're trialling it is because you don't know which one works more effectively. So, you know, it won't hurt me to try. Interviewee 4


Main findings

In this pilot study we have established that a definitive trial is feasible by demonstrating the numbers of women potentially eligible, and who agree to join, that recruitment in labour is acceptable to women, together with establishing the recruitment processes, and data collection methods and tools.

Evidence from this study suggests that the implementation of NICE guidance for the definition of confirmed delay, with the 2–hour window of suspected delay without intervention, reduces the number of spontaneously labouring women who require oxytocin augmentation from approximately 40–11%.[18] This markedly reduces the numbers of women requiring intervention, and suggests that previous definitions of delay have encouraged intervention in women who were actually labouring normally. This information must inform any definitive trial.

Women recruited to HOLDS who had confirmed delay, as defined by NICE, had relatively high morbidity, with increased rates of operative birth (caesarean section, 34%; instrumental birth, 40%). These rates are increased when compared with women whose labours became delayed using the standard definition,[18] which suggested a caesarean section rate of 9.1%, and an instrumental birth rate of 21%. This suggests that the 2007 NICE definition of delay in the first stage of labour identifies women with truly delayed labour, who are therefore more likely to end up with more intervention. High-quality evidence for this group would clearly be advantageous.

Analysis of the interview data showed that most interviewees had at least a basic understanding of the trial, with some being able to recall more detailed information. The majority of interviewees felt able to give informed consent. Very often the woman's partner or mother was important in helping them make a decision, and commonly read the information sheet either with them or on their behalf. Most women were happy with the content, tone, and length of the written information received, although the provision of verbal information was a valued addition. This emphasis on the importance of verbal information in these circumstances further highlights the need for training of the clinicians taking consent. Recent guidelines regarding the training of clinicians provides a template of what should be included for clinicians to satisfy both regulatory and clinical requirements.

Despite the fact that only four women remembered receiving the information about the study in the antenatal period, it is reassuring that the majority of women interviewed had at least a basic understanding of the study that they had joined, and felt that they were able to give informed consent. Although this might appear to suggest that there is little value in informing women about research studies in the antenatal period, guidance from both the RCOG and consumer groups recommend this approach.[11, 19] We were keen not to unduly worry pregnant women about a complication that they may or may not experience, and it may be that by making clear to women antenatally that we had no reason to think that they would become delayed in labour we unintentionally made the study appear somewhat irrelevant to them.

Participants' understanding of the risk involved in joining any study is complex. Inherently there will be evidence of equipoise before any trial is commenced, and so involvement does not guarantee benefit, and there is some element of risk involved in participation. In many trials the option of participation offers an alternative (and potentially better) treatment, which may only be available through participation. This study used an established drug treatment (oxytocin), and its use in a higher dose arm meant there was some increased element of risk involved that participants did not always appear to recognise. The understanding of the risk by participants has been identified by other researchers as being problematic,[7] and we should continue to strive for clarity.

Strengths and weaknesses

Evidence from the interviews conducted in this study supports the rationale for approaching women when they have suspected delay in labour, and suggests that it is acceptable to women that consent is obtained once delay is confirmed in labour; however, we only explored women's experiences following a diagnosis of delay in labour, when the majority of women had an epidural in situ. It is therefore plausible that women's experiences may be different in other intrapartum situations, or if the outcome for mother and baby had not been favourable. We acknowledge that the results are also limited by the lack of non-European participants, and we are concerned about the few women who felt pressurised to participate, although it is unclear whether this was related to the trial. Researchers in this area should continue to include the views of participants as pivotal to their work, with the aim of improving our practice related to clinical trials.


There is evidence of uncertainty as to the optimum dose of oxytocin for women with delayed labour.[3, 4] This topic has been identified as a research priority ([2] This pilot study has shown that such a trial is feasible and acceptable to women. Until a definitive trial is funded, the uncertainty as to the optimum treatment of delayed labour for women and babies remains.

Disclosure of interests

None of the authors have any conflicts of interest.

Contribution to authorship

SK was the chief investigator for the HOLDS pilot study, and conceived and led the study. Expertise was provided in the following areas: obstetrics, SW, TJ, and JW; statistics, VC and AH; qualitative methodology, NA; service user input, CM; health economics, SP and JW undertook the interviews. SM and NA undertook the qualitative data analysis. All authors contributed to the study design, developed the protocol, and ran the pilot. SK drafted the article, which all authors commented on and agreed the final version.

Details of ethics approval

Ethical approval was granted on 23 June 2010 by the National Research Ethics Service Committee Leicester, Northampton and Rutland 1 (ref. no. 10/H0406/30). The trial was sponsored by the University of Birmingham with Birmingham Women's NHS Foundation Trust.


Funding was given by the National Institute for Health Research (NIHR) Research for Patient Benefit (PB-PG- 0407-13193). SK was funded by the NIHR through the Collaborations for Leadership in Applied Health Research and Care for Birmingham and Black Country (CLAHRC-BBC) programme. This report presents independent research funded by the NIHR.The views expressed are those of the authors, and not necessarily those of the UK's National health Service (NHS), the NIHR, or the Department of Health.


We would like to thank the women who joined the study. The Oversight Group comprised Professor D. Tuffnell (Bradford), Dr J. Sanders (Cardiff), and L. Middleton (Birmingham Clinical Trials Unit). We would like to thank the HOLDS midwives (Gemma Barnfield, BWH; Lyndsey Donnison, LWH; and Claire Leader and Nicola Miller, RVI), as well as the obstetricians and midwives who recruited the women. We also thank the HOLDS Trial Co-ordinator Deb Bird and Birmingham Clinical Trials Unit for their support, especially Jane Daniels and Nick Hilken. We thank Dr Andy Ewer for his advice regarding the neonatal outcomes, and Professor DR Jones, Professor of Medical Statistics at University of Leicester, who was involved in the original application.