A multicentre, randomised controlled trial of position during the late stages of labour in nulliparous women with an epidural: clinical effectiveness and an economic evaluation (BUMPES)

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 08/22/02. The contractual start date was in January 2010. The draft report began editorial review in January 2016 and was accepted for publication in March 2017. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Debra Bick reports grants from the National Institute for Health Research (NIHR) Health Technology Assessment (HTA) programme, the NIHR Public Health Research programme, the NIHR Programme Grants for Applied Research programme and the Department of Health Policy Research Programme during the conduct of the study. Peter Brocklehurst reports personal fees from Oxford Analytica, grants from the National Institute for Health and Care Excellence, grants and personal fees from the Medical Research Council (MRC) and grants from the NIHR Health Services and Delivery Research programme, NIHR HTA programme and the Wellcome Trust outside the submitted work, and is chairperson of the NIHR HTA Maternal, Neonatal & Child Health Panel (MNCH). Pollyanna Hardy reports grants from the NIHR HTA programme outside the submitted work. Edmund Juszczak reports grants from the NIHR HTA programme and NIHR Efficacy and Mechanism Evaluation programme outside the submitted work, and is a member of the NIHR HTA programme Commissioning Board. Christine MacArthur reports grants from the NIHR outside the submitted work. Phillip Moore reports grants from the NIHR HTA programme during the conduct of the study. Oliver Rivero-Arias reports grants from MRC, NIHR HS&RD, NIHR HTA outside the submitted work and is a panel member of the NIHR Fellowship Awards and the UK National Screening Committee Fetal, Maternal and Child Health Group. Andrew Shennan reports grants from the MRC, the NIHR Research for Patient Benefit programme and the NIHR HTA programme outside the submitted work.

Copyright statement

© Queen’s Printer and Controller of HMSO 2017. This work was produced by Bick et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2017 Queen’s Printer and Controller of HMSO

Scientific background

As the most effective form of pain relief in labour, epidural analgesia is chosen by up to 30% of women in the UK each year,1 and this proportion has remained relatively stable over the last decade. 2 The uptake is greater in nulliparous women, with up to 40% of women having an epidural in large obstetric sites. 3 However, a systematic review of 23 randomised controlled trials (RCTs) that compared epidural analgesia with non-regional or no analgesia in labour found that epidural analgesia was associated with an increased risk of instrumental vaginal delivery (IVD) [risk ratio (RR) 1.42, 95% confidence interval (CI) 1.28 to 1.57]. 4

The trials that made the most contribution to the evidence base were conducted with epidural techniques that caused dense neuraxial blockade. Significant peripheral motor blockade, which can accompany conventional high-dose local anaesthetic epidural analgesia, inhibits mobility or the adoption of upright positions in labour. ‘Low-dose epidurals’, which use low-dose local anaesthetic in combination with opioids (usually fentanyl), were introduced in the early 1990s and are now in widespread use in the UK. This approach has been shown to result in a lower risk of IVD;5 however, the rate of IVD is still higher than that in women with no epidural. 6

Reducing the rate of IVD and increasing the spontaneous vaginal birth (SVB) rate would reduce short- and long-term morbidity for women by reducing the risk of perineal trauma and the effects of surgical repair. The incidence of perineal pain, dyspareunia and incontinence following IVD could also be reduced. 7–12 Although mobile epidurals preserve motor function (allowing greater mobility throughout labour) and can enable women to adopt upright positions, there is debate about whether or not an upright posture in the second stage of labour increases the SVB rate.

It is worth noting that the terms ‘ambulation’ and ‘mobilisation’ are often used interchangeably in the literature about epidural techniques that maintain motor function in the lower limbs. As the posture a woman adopts in labour is in part dependent on the motor power she retains, and this can be compromised by the peripheral motor blockade that accompanies effective epidural pain relief, it is clearly important to draw a distinction between mobilisation, the ability to move one’s legs, change position or move around the bed normally, and ambulation, which refers to the act of walking during labour. The ability to adopt upright postures in labour requires that women retain the capacity to mobilise, and some of these women will be able to ambulate.

A systematic review of the impact on mode of delivery of ambulation or upright positions in the first stage of labour (before full dilatation of the cervix) among women with epidurals found no significant difference between IVD (RR 1.16, 95% CI 0.93 to 1.44) and caesarean section (RR 0.91, 95% CI 0.70 to 1.19). 13 The second stage of labour may represent a period during which the adoption of an upright posture could exert the greatest influence and affect delivery mode by facilitating descent of the fetal head. A Cochrane review of position in the second stage of labour in women without epidurals found a reduction in IVD rate in the upright group (19 trials; RR 0.78, 95% CI 0.68 to 0.90). 14

Effectiveness of an upright position in the second stage of labour for women with epidurals

A Cochrane review of position in the second stage of labour among women with epidural analgesia was published in 2013,15 after the BUMPES trial started. This review included trials that compared upright positions with recumbent positions. The RR of SVB reported in the five included trials, including 879 women, was 1.02 (95% CI 0.81 to 1.28). There was clinical heterogeneity between the trials in relation to the eligibility criteria (some included multiparous women whereas others were restricted to nulliparous women) and in the nature of the interventions. In the upright group, for example, some women were actively encouraged to walk, and others were supported in a sitting position. In the recumbent group, some trials had allocated women to a sitting position and others to a lateral position. The authors concluded that there was no clear evidence about whether or not position in the second stage of labour made a difference to outcomes. 15

Effects on short- and longer-term maternal morbidity

An intervention that increases the rate of SVB by reducing the rate of IVD or caesarean section would also be expected to have an effect on short- and longer-term maternal morbidity. Faecal incontinence is clearly documented as being associated with forceps,11,16 including ongoing symptoms in women who only ever have one forceps delivery. 17 There may also be an increased risk of urinary incontinence, although this may be more closely associated with a longer second stage of labour;18,19 however, women who have a caesarean section have a lower risk of symptoms. 20,21 Other bowel problems such as haemorrhoids19,22,23 and constipation24 are more common after IVD, as are perineal pain and dyspareunia. 24,25 Caesarean section has many adverse sequelae, but, with the exception of faecal incontinence, most of these symptoms are less likely to occur in association with this delivery mode. It is therefore important to investigate positive impacts as well as any possible negative impacts of upright positions in the second stage of labour on maternal health outcomes.

There is increasing interest in obtaining maternity service users’ views of satisfaction with their experience of birth, as an indicator of the quality of their care and to inform organisational and policy changes. 26 Satisfaction is poorly defined and measured, although it is generally agreed that it is a multidimensional concept. 27,28 In a systematic review of factors influencing women’s satisfaction with birth, with a focus on the role of pain and pain relief, four factors (caregiver support, participation in decision-making, personal expectations and caregiver–patient relationship) were identified as important influences. 28 As position in the second stage of labour could influence a woman’s perceptions of the support she receives, her feelings of control and her expectations and experiences of labour and birth, satisfaction is an important consideration. The impact of negative consequences of the position adopted in the second stage of labour on these perspectives should also be identified.

Policy and practice at the time the trial commenced

Up to 30% of women in the UK use epidural analgesia for pain relief at some point in labour,1 with wide variation in the rate of epidural use between units. In a 1997 survey of UK units regarding epidural analgesia for labour, the epidural rate, including ‘low-dose’ epidurals, ranged from 0% to 85%, with an average rate of 24%. Of the 190 units that replied to the survey, 45 (24%) offered ‘low-dose epidurals’. 29 There is variation in the epidural technique employed to provide pain relief in labour and hospital policies governing maternal ambulation with an epidural in situ. A UK survey was conducted via the Obstetric Anaesthetists’ Association in 2008 to characterise national epidural practice and policy, with a response rate from lead clinicians of 80%. 30 It found that 95% of respondent units employed various epidural techniques consistent with the adoption of a range of upright positions, including ambulation, and that less than 50% of women actually did ambulate. Findings from the BUMPES trial are therefore widely generalisable to the majority of the nulliparous population that chooses epidural pain relief. With regard to reported hospital policies, 34% permitted maternal ambulation with low-dose epidural analgesia in situ. 30 Of those units that did not permit ambulation, 37% cited lack of evidence of a beneficial effect as a reason for this policy. This reluctance may reflect the current uncertainty in this field and that in general midwives have less experience of enabling women with epidurals to ambulate in second-stage labour rather than being in bed.

Rationale for a trial comparing upright with lying-down position

The National Institute for Health and Care Excellence (NICE) guidelines on intrapartum care published in 200731 (with no change in the update published in 2014) noted that there is ‘no effect of mobilisation following epidural analgesia on any maternal or neonatal outcomes’, and recommended that ‘women with regional analgesia should be encouraged to move and adopt whatever upright positions they find comfortable throughout labour’ (section 1.5.7, p. 22). This guidance is likely to lead to an increase in the use of upright positions, hence the need to compare upright positions with ‘lying-down’ positions rather than with usual care, given that usual care will increasingly include women assuming an upright position. Good-quality evidence is needed on whether or not upright positions in the second stage of labour in women with epidural analgesia have any beneficial effect on delivery mode and other important outcomes. It is crucial that the policies for the upright and comparison groups are clearly defined and monitored to ensure separation of the two approaches and to provide robust evidence about whether or not adopting an upright position does improve outcomes for women and their babies.

Aim of the BUMPES trial

This was a multicentre RCT in which the primary objective was to evaluate whether or not, in nulliparous women who choose low-dose epidural analgesia, a policy of adopting an upright position throughout the second stage of labour is associated with an increase in the incidence of SVB, compared with a policy of adopting a lying-down position.

This objective was supported and supplemented by the following secondary objectives:

• to evaluate whether or not there are differences between the two policies in important clinical outcomes for women and babies around the time of birth and 1 year post partum

• to evaluate the cost-effectiveness of the two policies for position during second-stage labour from a NHS perspective

• to measure women’s satisfaction with, and experience of, labour and delivery.

Trial design

The BUMPES study was a pragmatic, multicentre, individually randomised controlled trial that had a target recruitment of 3000 nulliparous women who had a low-dose epidural in situ. It was a two-arm parallel-group trial with one arm allocated to adopting an upright position during the second stage of labour and one arm allocated to adopting a lying-down position during the second stage of labour (Figure 1).

FIGURE 1.

Flow chart of participant recruitment.

Participant eligibility

The following inclusion criteria were applied throughout participant recruitment.

Sample population

All women who met the inclusion criteria were considered potentially eligible to participate in the study.

Study setting

Trial recruitment was undertaken in the labour wards of participating NHS maternity hospitals.

Information for women and obtaining informed consent

Information about the trial was provided to all nulliparous women during the antenatal period, after their booking appointment. This process was individualised for each participating centre depending on their routine practice to maximise the number of women offered information well in advance of labour. For example, in some sites, women were provided with information about the trial at their routine anomaly scan appointment (18–22 weeks). All women had the opportunity to ask questions of their midwives or obstetricians at the hospital, or they could contact the trial office. When a woman in a participating centre had an effective epidural established during the first stage of labour, she could then be offered a participant information leaflet on the study. If, after reading this and having the opportunity to ask questions, she was willing to take part in the study, then informed consent was taken. The participant information made it clear that women were free to withdraw from the trial at any time for any reason without prejudice to their future care, and with no obligation to a give a reason for the withdrawal. Written informed consent was obtained by a health professional (e.g. midwife, obstetrician or anaesthetist) with delegated authority from the principal investigator at each site. Consent comprised a dated signature from the woman and a dated signature of the person who obtained informed consent. A copy of the signed informed consent document was given to the woman. In addition, one copy was retained in the woman’s medical notes, one was retained in the study site file and one was sent to the Trial Co-ordinating Centre.

Interventions

Women were allocated to a policy of either upright maternal position (intervention group) or lying-down maternal position (control group).

Intervention group

Women were allocated to a policy of upright maternal position that would maintain the pelvis in as vertical a plane as possible during the second stage of labour, with the intention of continuing this until the birth. Women allocated to the ‘upright’ group were encouraged by their midwife to adopt positions that allowed for as upright a posture as possible. This could include walking, standing, sitting out of bed, supported kneeling or completely upright in an obstetric bed (Figure 2) for as much of the second stage as possible.

FIGURE 2.

Possible positions for women randomised to the upright maternal position. (a) Seated; (b) supported kneeling; (c) seated with extended legs; and (d) completely upright.

Control group

Women were allocated to a policy of a lying-down maternal position that would maintain the pelvis in as horizontal a plane as possible during the second stage of labour, with the intention of continuing this until the birth. Women allocated to the ‘lying-down’ group were encouraged to adopt a lying-down position that would mean lateral positions or lying down in bed for as much of the second stage of labour as possible. The bed could be tilted at up to a maximum of 30 degrees from horizontal (Figure 3).

FIGURE 3.

Possible positions for women randomised to the lying-down maternal position. (a) From in front; and (b) from behind. Note: a truly supine position (i.e. flat on the back) should not be used during labour because of the risk of aortocaval compression from the gravid uterus causing maternal hypotension.

Monitoring of adherence to allocation

In the second stage of labour, women with an effective epidural anaesthetic frequently have no desire to push. After confirmation of the second stage of labour, women were entered into the study. Midwives were encouraged to manage the second stage in two phases: a period of passive second-stage labour, allowing time for descent of the fetal head, followed by an active phase of expulsive pushing.

Training emphasised to the midwives the importance of supporting the woman in her allocated position, especially for the passive stage (which could last up to 2 hours). Positions were recorded on the trial worksheet at 15-minute intervals using a tick box, and midwives recorded ‘reason for change’ if the woman was moved out of her allocated position. As a pragmatic study, it was agreed that there would be expected reasons for changing position, for example fetal distress, fetal blood sampling or maternal discomfort, or to help improve pushing in the active second stage of labour. It was emphasised that midwives were required to record this information.

Randomisation

Participants were randomised to the allocated intervention (allocation ratio 1 : 1) using a web-based central service. To confirm eligibility, investigators were required to confirm the woman’s age and gestational age, that this was the woman’s first birth, that the fetus was a singleton with cephalic presentation and that an effective epidural was in situ, as well as obtaining signed consent. The randomisation software used random permuted blocks of sizes 2, 4, 6, 8 and 10, selected according to the proportions specified by Pascal’s triangle (1 : 4 : 6 : 8 : 10) to ensure that the staff recruiting women to the trial could not reliably predict the next allocation. Because of the large numbers of women recruited in each centre, no stratification by clinical characteristics was planned, although there was stratification by centre. The procedures for randomisation were fully documented, tested prior to the start of the trial, and monitored by the randomisation centre during the trial.

Outcome measures

Sample size

The proposed sample size was 3000 women. At the time of writing the funding application, the assumed rate of the primary outcome of SVB in the control group was 55%. This was derived from data published from the Comparative Obstetric Mobile Epidural Trial (COMET) reflecting SVB rates in nulliparous women with a mobile epidural in the second stage of labour. 5 A total sample size of 3000 women (1500 in each arm) would have 90% power to detect a clinically significant (absolute) difference of 6% in the SVB rate between the two policies (with a 95% CI). The cost of implementing this technology is low; therefore, even modest differences in outcome are likely to be cost-effective. Detecting the smallest and most clinically relevant effect size possible was therefore desirable. A 6% absolute risk difference, which equates to a 10% RR reduction (approximately), was well within the uncertainty of the existing evidence (despite the existing trials’ heterogeneity) and was considered sufficient to change clinical practice.

The proportion of the upright group achieving a SVB was anticipated to be 61% under the null hypothesis. The test statistic used was the two-sided z-test with pooled variance. The significance level of the two-sided test was targeted at 5%. When considering longer-term outcomes, the proposed sample size of 3000 would be sufficient to detect a difference in the prevalence of faecal incontinence of 12% in the control group compared with 8% in the intervention group. The incidence of this outcome has been estimated as 14% among forceps deliveries and 10% among women with a SVB. 11

On collation of the pilot data for an interim analysis presented to the Data Monitoring Committee (DMC) in 2011, it was recognised that the combined primary outcome event rate was lower than anticipated. As of 6 December 2011, the overall SVB rate for BUMPES (combining upright and lying-down groups) was 33.8% [(95% CI 26.1% to 42.1%) based on 49/145 events]. With a reduction in the control group event rate (from an anticipated 55% to between 30% and 40%), keeping the sample size fixed at 3000 would mean that a RR of between 1.13 and 1.19 would be detectable, equivalent to an absolute risk reduction of 5–6%. Although there was not sufficient power to detect a RR as small as the planned 1.11, the absolute risk detectable is similar. The Trial Steering Committee (TSC) agreed that changes to the target sample size were therefore unnecessary.

Governance

Serious adverse event reporting

Serious adverse events were reported to the University College London (UCL) Trial Co-ordinating Office within 48 hours. The Trial Co-ordinating Office notified the chairperson of the DMC and the REC. All SAEs occurring during the trial observed by the investigator or reported by the participant, whether or not attributed to the trial, were reported on the DCB. SAEs considered to be related to the trial by the investigator were followed up until resolution or until the event was considered stable. The local investigator was asked to provide follow-up information when necessary. All related SAEs that could have resulted in a participant’s withdrawal from the trial, or which were present at the end of the trial, were followed up until a satisfactory resolution occurred.

The chief investigator submitted to the REC, once a year throughout the clinical trial, a safety report that included all SAEs.

Data handling, checks, cleaning and processing

All data collection forms (i.e. DCBs, HLC forms, maternal satisfaction and 1-year follow-up forms) and consent forms, once completed and returned to the UCL Comprehensive Clinical Trials Unit (CCTU), were logged as received and date stamped. Data were double entered at the UCL CCTU using the study database, by independent data clerks. Validation routines checked for missing data and inconsistencies on an ongoing basis. This included screening for out-of-range data, with cross-checks for conflicting data within and between data collection forms using computerised logic-checking screens. Any validation errors on the DCBs and HLC forms were queried and documented. Queries were communicated to the appropriate centres by the trial manager. Errors on the maternal satisfaction questionnaire and the 1-year follow-up form were not queried with the woman.

Cost-effectiveness analysis

An economic evaluation was conducted as part of this trial and is reported in detail in Chapter 6.

Patient and public involvement

When the initial investigator group was being assembled to develop the trial, the National Childbirth Trust was approached to suggest a lay member who would be willing to join the group as a co-investigator. Mary Nolan agreed to join the group, and assumed equal membership of the co-investigator group at all planning meetings and trial conduct meetings, and in the drafting of the application, developing the detailed trial protocol and data collection forms, and report and paper writing. Mary took a lead in helping the team to develop participant information leaflets to be used in the antenatal period and at the time of labour, as well as helping to plan dissemination activities and drafting and developing the summary information for the public. During the course of the trial, Mary Nolan left the NCT to take up a position as Professor of Perinatal Education at the University of Worcester, but continued to represent the potential participant’s perspective in all aspects of the trial development, conduct and analysis.

Patient groups for analysis

Losses to the trial post randomisation were defined as any of the following:

• women for whom a valid consent form was not received

• women for whom consent to use their data was withdrawn

• women not in the second stage of labour when randomised and who did not reach the second stage before delivery

• women not in labour or without an epidural in place at the time of randomisation.

The numbers (with percentages of the randomised population) of post-randomisation exclusions are reported by randomised treatment group, and the reasons summarised.

Women could specify whether or not data collected up to the point of withdrawal could be used. If the response was ‘no’, then they were counted as post-randomisation exclusions. If the response was ‘yes’, then they were reported as ‘missing’ for all subsequent outcomes.

For the primary analysis, participants were analysed in the groups into which they were randomly allocated, that is, comparing the outcomes of all women and infants for women allocated to a policy of an upright position with those of women allocated to a policy of lying down, regardless of position recorded at any time during the second stage of labour. Losses to the trial post randomisation are excluded from all analyses, with the exception of the safety-reporting population, which excluded women for whom a valid consent form was not received and women who withdrew and did not consent to use of their data.

The unit of analysis was the woman for all maternal outcomes and the infant for all infant outcomes.

Descriptive analyses

The flow of participants through each stage of the trial is summarised using a Consolidated Standards of Reporting Trials (CONSORT) diagram (see Figure 8). Specifically, for each intervention group we report the numbers of women randomly assigned and women for whom the incorrect allocation was recorded in the eligibility section of the DCB. The number of ineligible women randomised is reported, with reasons for ineligibility. The number of post-randomisation exclusions and women analysed for the primary outcome is also reported. We report numbers for the 1-year follow-up, women lost to follow-up and women who withdrew before 1 year. The total number of eligible women was not collected during the conduct of this study, as it was considered too great a burden for the participating centres and would not be sufficiently reliable.

Numbers (with percentages) for binary and categorical variables and means (and SDs) or medians (with lower and upper quartiles), or geometric means for continuous variables are presented; no tests of statistical significance were performed, nor CIs calculated, for differences between randomised groups on any baseline variable.

The number (with percentages) of losses to follow-up among women selected for the 1-year assessment is reported in the CONSORT flow chart by trial arm, and the reasons reported. Selected demographic and clinical characteristics, the primary outcome and selected short-term outcomes of women and their infants with 1-year data available were compared with those without 1-year data.

Missing data for primary and secondary outcomes, from baseline to the end of follow-up, are summarised for the two trial arms. Not all data were routinely collected by all hospitals; for example body mass index (BMI), cord artery pH and base deficit were sometimes omitted. The DCB allowed midwives to tick ‘data not recorded’. These data are summarised by trial arm and reported separately from data missing or unknown. Missing data for the primary outcome were negligible. If any data items were missing on the DCBs, every effort was made to extract these data from the hospital involved.

Primary effectiveness analyses

Outcomes are summarised by trial arm using counts and percentages for categorical variables, means and SDs for normally distributed continuous variables or medians and interquartile ranges for other continuous variables. In addition, geometric means are presented for durations of stages of labour, as these are inherently highly skewed data.

An adjusted analysis was performed on all comparative analyses adjusting for centre (the stratification factor at randomisation) as a random effect. Binary outcomes were analysed using log-binomial regression models and results presented as adjusted RRs with corresponding CIs. If the model did not converge, then log-Poisson regression models with robust variance estimation were used. 33 If the model was still unstable, then the centre was removed and unadjusted RRs presented. Continuous outcomes were analysed using linear regression models and results presented as adjusted differences in means with associated CIs. Unadjusted Hodges–Lehmann34 median differences (plus CIs) for skewed continuous variables are presented. The estimates are based on a difference between distributions. The Hodges–Lehmann median difference is calculated by forming all possible differences between the first treatment group and the second treatment group, and taking the median of those differences.

In addition, geometric mean ratios (GMRs) are presented for durations of the stages of labour as the distribution of these data is highly skewed. A geometric mean is a measure of central tendency that is based on the product of values (as opposed to an arithmetic mean that sums the values). A ratio of geometric means provides an indication of how large one geometric mean is relative to another.

Comparisons between randomised groups of all primary and secondary outcomes are reported in full for completeness and transparency, that is, there is no selective reporting of outcomes.

In order to take account of the number of comparisons, 95% CIs are presented for the primary outcome and 99% CIs for all other outcomes.

Additional effectiveness analyses

The primary analysis was adjusted further for the primary outcome (pre-specified in the SAP) to investigate the impact of the following known prognostic factors (in addition to centre): age as a continuous variable, ethnicity, diagnosis of delay and onset of labour (induced vs. spontaneous).

To examine whether or not the effect of policy of position during the second stage of labour was consistent across specific subgroups of women, the following prespecified subgroup analyses were undertaken:

• gestational age (37⁺⁰ to 38⁺⁶ weeks; 39⁺⁰ to 40⁺⁶ weeks; and ≥ 41⁺⁰ weeks)

• maternal age (≤ 24 years, 25–29 years, 30–34 years and ≥ 35 years)

• augmentation with oxytocin (Syntocinon^®; Novartis Pharmaceuticals UK Ltd, Frimley/Camberley, UK) in the first stage of labour (yes/no)

• Index of Multiple Deprivation (IMD; population-based quintiles 1–5; derived using the postcode of the woman’s last known address based on the Indices of Multiple Deprivation 201035 and Ordnance Survey Code-Point Open36 February 2013).

For the trial primary outcome, results are presented as forest plots showing the RR plus 95% CI for each subgroup,37 by intervention group, with the p-value for the statistical test of interaction. 38 Centre was included as a stratifying factor in the list of subgroup analyses in the original protocol, as we were expecting to recruit to target using five centres only. Recruitment rates were poor and we expanded the number of recruiting centres to 41. A subgroup analysis on 41 centres was therefore not considered relevant.

A prespecified sensitivity analysis on the 1-year maternal outcomes was carried out on a restricted data set that excluded all women who were pregnant or had another child at the time of completing the follow-up questionnaire.

Protocol violations and deviations

A protocol violation is a failure to comply fully with the final study protocol as approved by the REC and research department, such as a serious non-compliance with the protocol resulting from error, fraud or misconduct, and results in the exclusion of a patient from the analysis for the study. There were no protocol violations.

A protocol deviation is a departure from the final study protocol as approved by the REC, with minor consequences on the integrity of the data. Protocol deviations are those that resulted in exclusion from the analysis reported in Chapter 5 (see Figure 8). There was only one other protocol deviation, and that was unrecognised at the time of randomisation; the woman had intrathecal analgesia.

Recruitment

Trial recruitment was initially planned to be undertaken in the maternity units of four acute NHS trusts in England and one health board in Wales. The BUMPES study design originally outlined in the trial protocol described a single-centre internal pilot study to assess feasibility, develop teaching materials and field-test trial data collection processes. After 9 months of the pilot phase, it was noted that, although the trial infrastructure and data capture were satisfactory, accrual did not meet projected targets, despite accurate predictions of available participants. At the recommendation of the TSC, the trial was initiated at a second pilot site prior to ‘roll-out’, in order to establish if these limitations were site specific or reflected broader barriers to recruitment. It was noted that recruitment across the two pilot centres remained unsatisfactory, with an average of 49% of the overall recruitment target being met over the 6 months since the second pilot site opened to recruitment (Figures 4 and 5).

FIGURE 4.

Recruitment at first pilot centre.

FIGURE 5.

Recruitment at second pilot centre.

Challenges to study recruitment

Recruitment strategies

During the whole period of recruitment, a number of initiatives were launched to improve recruitment. These included:

• Timing of informed consent. During the pilot phase, it was noted that gaining informed consent in the second stage of labour (from full dilatation of the cervix to birth of the baby) was delaying time to randomisation and therefore study entry. As the second stage of labour is a clinically demanding time on labour wards, this could potentially be a barrier to recruitment. Following approval (REC amendment 4, 25 July 2011), consent could be sought and obtained from potential study participants in the first stage of labour, once an effective epidural had been administered. Randomisation had to be delayed until the second stage of labour had been confirmed, but this process alleviated the burden of recruitment for the attending midwife.

• To recruit at more sites. Following a proposal from the CIG and agreement from the TSC, there was approval to recruit a further 36 maternity units, which were opened to recruitment over a 24-month period (a total of 41 hospitals). Additional units that had a good track record of participation in health research in pregnancy were approached, along with hospitals that had already expressed an interest in participating. An initiation visit from the research midwife in the BUMPES team was arranged to fully explain the study to lead midwives, anaesthetists and local R&D departments, and also to evaluate their enthusiasm and the level of support that they would offer the study. Following R&D approvals, dates were arranged for the research midwife to attend the maternity units, and provide training to staff and support them during initial recruitment. This usually took 1 full week, covering day and night shifts, and involved small groups of midwives and anaesthetists. A training manual, posters, recruitment packs and randomisation flow charts, as well as 24-hour contact details, were in place for all centres prior to the start of recruitment. Further training was also provided to many units on request to support recruitment.

• Change to funding model. Initially, BUMPES provided funding to appoint a ‘BUMPES midwife’ at each of the original five maternity units for 2 days per week. Their role was to support training, recruitment, data collection and the day-to-day running of the study. With the involvement of 36 more maternity units, the existing funding model was unsustainable within the trial budget, so this was changed to a ‘payment-per-recruit’ model (£85) for each of the maternity units. This proposal was approved by both the TSC (9 December 2011) and the NIHR HTA programme, and was in place from January 2012.

• Development of local BUMPES champions and Comprehensive Local Research Network (CLRN) support (England). Given the change in the funding model, and the loss of specific BUMPES midwives, a revised model of local support was designed. This involved the introduction of BUMPES champions. Clinical midwives active on each labour ward shift were identified to promote the study, identify potential recruits, facilitate consent and support randomisation. This ensured that, as much as possible, someone was available who was knowledgeable about the trial and able to support recruiting midwives. This was supported in some units with extra CLRN funding and in others by the payment-per-recruit monies.

• DCBs. Following feedback from the units, the DCBs were redesigned. Staff complained that DCBs were too long and the amount of information requested was too much, so that, on a busy labour ward shift, midwives were put off recruiting or completing the booklets. The DCBs were redesigned into parts 1 and 2. Part 1 was reduced to a single-page A3 worksheet and was the only section that the attending midwife during labour needed to complete. This requested information that could not be collected at a later date, for example visual analogue scale score (pain assessment), the date and time when the woman adopted the allocated position, times and positions every 15 minutes and reasons for change. CLRN and the National Institute for Social and Health Research [(NISCHR) Wales] research staff or staff employed using the BUMPES payment per recruit monies were able to complete part 2 at a later date with information from the maternal and neonatal notes.

• Increasing midwifery ownership. A short article to raise awareness was published in the Royal College of Midwives Journal (2012). This was designed to encourage midwifery ownership of the study and the importance of the results, which could potentially have an impact on future midwifery practice and be beneficial to women. A Collaborators’ Study Day for recruiting units was arranged in November 2012 to improve networking, and for sharing ideas and identifying areas of good practice. This was attended by 34 midwives from 18 participating centres and feedback from the day was excellent.

• Promoting BUMPES. Life-sized posters and other promotional items were designed and, when required, received REC approval. This helped to encourage promotion of the study to midwives, women and antenatal educators.

• Recruitment updates and newsletters. Recruitment updates were sent to units monthly. Newsletters were published quarterly and included recruitment targets, the identity of new participating units and answers to frequently asked questions, to help improve awareness and address common errors and queries.

• Incentives. Approved incentives for midwives such as fob watches, notebooks, Post-it^® (3M, Cynthiana, KY, USA) notes, mugs, key rings, pens, tape measures, lanyards and lip gels were purchased and given during training sessions. This helped to identify and promote contact details of the study. Occasional gift vouchers were approved and given to support recruitment as well as seasonal gifts, such as Love Hearts (Swizzels Matlow, New Mills, Stockport, UK) for Valentine’s Day, Easter eggs, summer rock candy, Halloween-themed sweets and an advent calendar in December.

The combination of marketing the trial more actively and participation of the additional centres resulted in a substantial improvement in recruitment. The project management group continued to monitor recruitment closely throughout the trial. An example of the monitoring data reviewed is shown in Table 3 and Figure 6.

FIGURE 6.

Study recruitment details from March 2012 to February 2013.

However, the delays inherent in establishing the participation of a greater number of centres resulted in a request to the NIHR HTA programme for a 12-month extension of the trial. This no-cost extension was granted in September 2013.

Methods

A health economics analysis plan was developed to guide the health economics team during the development of the economic evaluation (see Appendix 8). Overall, the team followed this analysis plan during the conduct of the economic evaluation, but there was one deviation from the plan that affected the final presentation of results of the economic analysis. The original analysis plan stated quality-adjusted life-years (QALYs) as the primary health outcome measure and used a cost–utility analysis to present the results of the economic evaluation. However, quality-of-life data at randomisation were not collected owing to the intrinsic difficulties in collecting this information during labour. In addition, it was not planned to collect data during the early postnatal period as part of the trial. As a result, the calculation of a QALY profile for the trial duration could not be derived. The next alternative could have been to conduct a cost-effectiveness analysis using the primary outcome in the trial or another clinical end point to present the results of the economic analysis. However, given that this was, to our knowledge, the first economic evaluation of position during labour, the results of such an analysis would have been difficult to interpret by decision-makers in the absence of a willingness to pay for health gains for the selected outcome. Therefore, we decided to conduct a cost–consequences evaluation as the primary analysis for the economic evaluation. In a cost–consequences analysis, the different components of costs and benefits of the interventions under evaluation are presented in a disaggregated manner without an attempt to estimate a summary measure [e.g. incremental cost-effectiveness ratio (ICER)]. 41 As a secondary outcome of the economic analysis we conducted a cost-effectiveness analysis using the number of additional cases of SVB as the outcome measure. The latter analysis was included as a benchmark for future economic evaluations studies in obstetrics.

Health outcome measures

Health outcomes for women and their babies before discharge and at the 1-year follow-up were evaluated as potential consequences to include in the economic evaluation. Chapter 5 reported a number of secondary maternal, neonatal and longer-term outcomes in addition to the primary outcome in the trial. Most of these outcomes led to health-care service utilisation, which was accounted for in our analysis, and only a selection of maternal and infant outcomes were therefore included in the cost–consequences analysis. For women, we selected the incidence of SVB, maternal satisfaction with labour as reported immediately after the birth and urinary and faecal incontinence at 1 year’s follow-up. An Apgar score of < 4 at 5 minutes and major morbidity at 1 year’s follow-up were the consequences selected for infants. When reporting the results of difference in estimates between upright and lying-down positions for any of these outcomes, we refer to the appropriate tables in Chapter 5 rather than replicating the tables here.

In addition to the above consequences, maternal HRQoL information using the EuroQol-5 Dimensions, three-level version (EQ-5D-3L) and the Short Form questionnaire-12 items (SF-12) instruments was collected at the 1-year follow-up. The EQ-5D-3L is an increasingly widely used multiattribute generic instrument for measuring HRQoL in cost–utility analyses. 42 It has two components: a descriptive system covering five dimensions (mobility, self-care, usual activity, pain/discomfort and anxiety/depression), each of which has three levels (no problem, some problems and extreme problems), and a ‘feeling thermometer’ using a visual analogue scale. The 243 health states of the EQ-5D-3L can be converted into preference-based utility values using a value set obtained from a British general population sample. 43 The SF-12 is a generic HRQoL instrument derived from its longer counterpart, the Short Form questionnaire-36 items (SF-36),44 designed to measure general health functioning. The SF-12 items measure physical or emotional limitations, physical functioning, pain, general health, vitality, social functioning and mental health problems. The number of levels in each SF-12 question varies depending on the version used, and in this study the SF-12 version 2 was administered. Health states from the SF-12 instrument can be converted into preference-based utilities using the Short Form questionnaire-6 Dimensions (SF-6D) algorithm, which was used in this study. 45

NHS health-care resource use

Detailed information about secondary care usage was collected and included resources consumed during the late stages of labour to hospital discharge, and during the first 12 months after birth as reported at the 1-year follow-up. Chapter 5 reported some data about health-care usage that we also present in this chapter with additional information. Data from trial entry up to postnatal hospital discharge were collected from hospital records and included in the DCB (see Appendix 1). A postal questionnaire was used to collect secondary care information at the 1-year follow-up, and this was sent by the trial management team, which also dealt with reminders and appropriate double-data entry and data cleaning. Information was collected for women and their babies. The different items of resource use collected for each category of secondary care health service are summarised in Table 18.

No intervention-specific costs were assigned to either upright or lying-down position as neither was associated with the use of any additional resources . Given that all randomised women already had epidural analgesia and that any remaining medication after the birth was considered to be waste, epidural-specific costs were excluded from the cost analysis. In addition, any top-up epidural drugs costs in both arms were excluded from the cost analysis as there was no evidence of a difference between groups (see Table 7).

At the study design stage, there was a general concern about including primary care and community care visits as part of the data collection because these tend to be frequent and poorly recalled by new mothers compared with secondary care visits. 50 It was agreed that hospital care constituted the main cost driver for this population and the target source data to collect in the study. Therefore, primary care and community care visit data were not collected. However, urinary and faecal incontinence are important outcomes following birth and may be related to the mode of delivery, and can have long-lasting effects on HRQoL and additional visits to primary care. 17,21 Therefore, it was decided that, if necessary, primary care visits related to these adverse events would be estimated using recent data from the literature if significant differences between treatment arms were observed. Nevertheless, this was not the case (see Table 14), and such visits were not incorporated as part of the categories of resource use in the cost analysis and are presented as part of the health outcomes in the cost–consequences analysis. We also assumed that any costs for specific surgeries were reflected in the length of stay and the unit cost attached to the admission. Therefore, we did not conduct a micro-costing approach for the maternal and infant surgeries performed in different time periods.

Unit cost data collection

Sources and associated estimates of unit costs for the different categories of resource use are presented in Table 18. Unit costs were mainly extracted from national sources, including the Personal Social Services Research Unit51 and the NHS Reference Costs 2013–14,48 and from a recent published cost-effectiveness analysis of alternative planned places of birth in woman at low risk of complications. 47 The unit cost of undertaking fetal blood sampling for the assessment of metabolic acidosis was not available in any of the sources consulted and was provided by the finance department of a large obstetric unit in Oxford. All costs were expressed in 2013/14 pounds sterling inflated to this base using the Hospital and Community Health Service Inflation Index52 where appropriate.

Statistical analysis

Volumes of categories of resource use were multiplied by the corresponding unit cost to estimate the cost per woman or baby in a particular category. This was then averaged across each trial arm to obtain a mean cost per woman or baby of a particular category. We estimated health-care resource use deriving mean estimates and SDs either across all women and their babies in an arm of the trial or only for those consuming the resource category. The latter analysis using standard descriptive statistics was expected to provide information about potential outliers influencing the estimation of mean resource use and costs for a particular category of resource use in each arm of the trial. Categories of resource use and associated costs are presented separately for women and their babies, but the estimation of summary costs for a particular resource use category (e.g. total costs at 12-month follow-up) was calculated adding information from the pair. Health-care resource use between treatment arms was compared using RRs for binomial variables and mean differences for continuous covariates. Costs were compared using mean differences between treatment arms. Recent evidence suggests that both parametric and non-parametric methods accurately estimate the true standard errors (SEs) of means, even when data are highly skewed, and moderate to large (n > 50) sample sizes for continuous variables. 53 Hence, mean resource use and cost differences and associated uncertainty for particular categories of resource use and costs between the two positions during the late stage of labour were estimated using parametric methods. In line with the statistical analysis of the primary outcome, differences between treatment arms were adjusted using a random intercept binomial (for RRs) or linear (for mean differences) model using hospital centre as a random effect. In order to be consistent with the original SAP, a 95% significance level was defined to determine significant differences in health-care resource use for the primary outcome (SVB) whereas a 99% significance level was used for the other categories of resource use. A 95% significance level was used to compare main categories of costs between treatment arms.

Mean differences and associated uncertainty in EQ-5D-3L and SF-6D utilities between the treatment arms were assessed using parametric methods and adjusted using a random intercept linear model for hospital centre.

The number of missing data for secondary care resource use up to hospital discharge was very low in each treatment arm (< 1%), and we present the health-care resource use and costs during this period using a complete-case analysis. The proportion of missing data on both resource use and HRQoL was higher at the 1-year follow-up. In this case, resource use was presented using a complete-case analysis, but in the case of maternal quality-of-life scores and costs we implemented a multiple imputation framework with chained equation. 54 Current guidance on handling missing data in cost-effectiveness analysis was followed to inform such analysis. 55 Only missing utility scores (for both EQ-5D-3L index and SF-6D utilities) and individual cost items were imputed and the distribution of responses for both instruments was reported for data available. We constructed an imputation model that included covariates with complete data on trial entry characteristics (maternal age, gestational age at entry, BMI and baby’s birth weight), HRQoL variables (EQ-5D-3L and SF-6D scores), and all individual categories of cost variables. We used prediction mean-matching, estimated 50 different imputations, and the imputation model was implemented separately by trial allocation. Mean estimates and estimates of SE were combined between imputed data sets using Rubin’s rule56 and adjusted using a random intercept linear model for hospital centre.

Mean differences in the total costs at 12 months’ follow-up and the number of additional cases of SVB between treatment arms were combined into the ICER to determine cost-effectiveness as a secondary outcome in the economic evaluation. The lying-down position was used as the comparator in the ICER calculation. Uncertainty around the ICER was evaluated parametrically using 95% CIs. 57 The five parameters needed (difference in costs, SE of difference in costs, difference in effects, SE of difference in effects and correlations of differences in costs and effects) to estimate parametrically the uncertainty around the ICER were obtained from the multiple imputation analysis.

The statistical analysis was conducted in Stata/SE for Windows version 13.1.

Results

NHS costs

Patterns of missing data for summary cost categories using a complete-case analysis are presented in Figures 11 and 12. It was evident from the plots that in each trial arm the number of missing data from trial entry to hospital discharge was low following a primarily monotonic missing data pattern. A larger number of missing data out of the overall data available was observed at the 12-month follow-up. The charts showed the potential cost information available (indicated by the grey area) to inform the multiple imputation model. The frequency of missing information in each trial arm and for all categories of health-care resource use are reported in Tables 19–21.

FIGURE 11.

Pattern of missing data for main categories of costs for the upright group. Blue shading represents missing data for one or more individuals (horizontal axis) for a particular cost variable (vertical axis); green shading represents observed data.

FIGURE 12.

Pattern of missing data for main categories of costs for the lying-down group. Blue shading represents missing data for one or more individuals (horizontal axis) for a particular cost variable (vertical axis); green shading represents observed data.

Table 24 reports the results of the cost analysis using multiple imputation over the study period for mothers and their babies. A statistically significant mean cost difference (95% CI) of £59 (£6 to £111) between trial arms favouring the lying-down policy was detected in the delivery-related cost category. This translated into a statistically significant mean difference (95% CI) of £78 (£13 to £143) from trial entry to hospital discharge per delivery favouring the lying-down strategy. A significant mean cost difference (95% CI) of £0.002 (£0.001 to £0.004) favouring the lying-down position was observed for the birth-related cost category. This difference disappeared when adding the cost of HLC admissions to the total cost for infants from trial entry to hospital discharge. A summary of the overall costs from trial entry to 12-month follow-up for women and their infants using multiple imputation is also shown at the end of Table 24. The mean (SE) total cost per woman/infant pair at 12-month follow-up (adding all categories of costs) was estimated to be £3207 (£73) and £3252 (£81) in the upright and lying-down positions, respectively, a non-significant mean cost difference (95% CI) of –£42 (–£254 to £169) favouring the upright position.

Maternal health-related quality of life at 12-month follow-up

Summary EQ-5D-3L and SF-6D scores at 12-month follow-up for each trial arm using a multiple imputation analysis is reported in Table 25. Quality-of-life scores were very similar (almost identical) between trial arms for both instruments and therefore no significant mean differences were detected.

The distribution of responses of available data for both instruments is reported in detail in Tables 26 and 27, respectively. It is clear that, using both instruments, the majority of women participating in BUMPES reported a good quality of life at 1 year post delivery.

Cost-effectiveness analysis as a secondary outcome in the economic evaluation

A summary of cost-effectiveness results comparing upright with lying-down positions during the second stage of labour is shown in Table 28. The ICER (95% CI) was estimated to be £722 (–£2968 to £6358) per additional case of SVB.

Discussion

This chapter has described in detail the methods and results of the economic evaluation conducted as part of the BUMPES trial. We have implemented a robust methodology based on more recent guidance to present these results. 58

Our results suggested that women randomised to the lying-down position consumed significantly fewer resources than those randomised to an upright policy during the original hospital stay. Such results were driven by more SVBs in the lying-down treatment arm. Infants incurred similar costs in both arms of the trial during this time. At the 12-month follow-up, no significant overall cost differences were observed between upright and lying-down positions for mothers or their babies. The significantly higher costs incurred from trial entry to discharge by women in the upright group were offset by the slightly, but non-significantly, higher costs incurred during follow-up by the babies of women in the lying-down arm. A possible explanation is that SVB was associated with higher follow-up costs than instrumental delivery or caesarean section. However, additional analysis assessing follow-up costs by mode of delivery between the trial groups suggested that this was not the case (data available from the corresponding author). No evidence of differences was found for the other secondary maternal outcomes, neonatal clinical outcomes or maternal HRQoL. Therefore, the results of the cost–consequences analysis provided robust evidence clearly in favour of lying down, at no risk to women or their babies, and at no extra cost to the NHS.

Future studies can use the results of our cost-effectiveness analysis using the cost per additional case of SVB as an outcome measure as the benchmark for future comparisons in the area. However, we recommend that such comparisons are used with caution owing to well-known problems of comparability to other disease areas and ‘double-counting’ when estimating the ICER. 59,60

To our knowledge, this is the first study reporting an economic evaluation of upright versus lying-down positions during the second stage of labour in nulliparous women with an epidural. The most comprehensive systematic review of economic evaluations of potential interventions during intrapartum care has been recently updated in a NICE clinical guideline. 31 The review aimed to identify all published cost-effectiveness evaluations in the areas within the remit of the guideline that included the second stage of labour, but no published economic evaluations were identified. Therefore, the results reported here should be considered the most up-to-date evidence about the cost-effectiveness of alternative position strategies during the second stage of labour in women with an epidural.

We were not able to collect HRQoL data at trial entry (or late pregnancy) or in the early postnatal period, precluding the estimation of QALYs. Therefore, one of the limitations of the current analysis is the limited comparability of our study with evaluations in other health areas using a cost-per-QALY approach, as recommended by NICE. 58 The use of QALYs has the potential to be a feasible and responsive outcome measure to evaluate the effectiveness of delivery position interventions during labour. A recent study has assessed the impact of mode of delivery on maternal HRQoL postnatally, suggesting that caesarean section is associated with larger quality-of-life decrements, followed by instrumental delivery and then SVB. 61 Collecting quality-of-life information during late pregnancy and in the early postnatal period to use in cost-effectiveness analysis is controversial, because there is limited evidence about the validity of recommended instruments such as the EQ-5D-3L in this context. 62,63 However, it would be possible to estimate QALYs in the BUMPES study using a modelling exercise that synthesises available quality-of-life information during late pregnancy and in the early postpartum period from a literature review with the collected EQ-5D-3L data at 12-month follow-up. Such analysis is of interest to the research team and will be explored in future research. However, we feel that our cost–consequences analysis provides important evidence that can be considered by decision-makers in obstetrics when making recommendations about position during the late stages of labour in women with an epidural.

Longer-term outcomes

The response rate to the 1-year follow-up was 61%. Therefore, there is a possibility that the follow-up results are less than robust because of non-response bias. This poor response, despite many attempts to improve the rate (described in Chapter 8), is a phenomenon increasingly being seen with contemporary clinical trials in this area. We provided a variety of options for women to complete their questionnaire: on paper, via the internet or by telephone. The majority of women still prefer to use paper, but, despite these various offers, the response rate remained poor. Responders and non-responders exhibited clear differences in a variety of demographic and clinical factors that were anticipated. Women who responded were more likely to be older, more likely to be white, less likely to live in a deprived area and less likely to have achieved a SVB. There were, however, no apparent differences between the two randomised groups in their response rates or characteristics, suggesting that there were minimal biases in the comparison between the two groups.

The lack of an impact of the risk of SVB on longer-term outcomes, such as faecal incontinence, is of interest. The observation that IVD is associated with increased risks of faecal incontinence is robust; however, in the BUMPES trial the differences between the randomised groups of women in their risk of SVB and instrumental delivery were relatively small, so, although there are associations between different modes of birth and long-term outcomes, these are likely to be diluted in a trial in which these differences in actual mode of birth are relatively modest (only a 6% absolute difference in the risk of SVB). This is likely to explain the lack of an observed difference in long-term outcomes.

Adherence

Adherence is clearly important when considering an intervention of this nature. The observation that there was greater adherence in the upright group than in the lying-down group suggests that there may be even greater benefit from adopting a lying-down position if adherence could be improved. However, the ability to comply with the allocated intervention may be a function of the intervention itself. For example, women in the lying-down group may be more difficult to monitor when using an external cardiotocograph monitor because of the ability to position the monitor accurately to pick up the fetal heart rate. If this is the case, then non-adherence for clinical reasons is to be expected. It would therefore be difficult to improve adherence in these women if that meant that they could not be effectively and safely monitored. Further work needs to be undertaken on the association between adherence and various characteristics of the nature of the non-adherence, and the risk of the various outcomes, before much more can be said about whether or not improved adherence may improve the benefit of the lying-down position.

Challenges with equipoise

Although the trial was initially expected to recruit from five UK centres, the expansion to 41 centres suggests that the results have good generalisability, in that a large number of centres participated throughout the country. However, the reason for the expansion in the number of centres was because of the concerns expressed by midwives in terms of recruiting to the trial. It was clear that many midwives did not have equipoise about position in the second stage of labour for women with an epidural, and in all of the participating centres it was a relatively small number of midwives who recruited the women in those centres. At no stage in the trial was there any suggestion that women were expressing a preference not to participate in the trial if they were approached. This effect of midwives having particular views about the benefits of position may limit the generalisability if midwives are not prepared to accept the evidence produced by the BUMPES trial. When there are strong views, for example that an upright position is preferable to a lying-down position, if the results of BUMPES fail to be implemented, then the population benefit of a lying-down position in these women will be relatively small. By producing results that contradict current NICE guidance and what is widely accepted, that is, that an upright position improves the chances of achieving a SVB, these results challenge an orthodoxy that may make their implementation more challenging. The results of the trial, however, are clear. If women giving birth to their first babies and their carers are keen to increase the chances of achieving a SVB, then adopting a lying-down position in the second stage of labour, for as much of that second stage of labour as can be achieved, results in a modest but real increase in their chances of doing so. There is no evidence to suggest that this improvement carries a risk either to the women themselves, in relation to perineal trauma, or to their babies, in relation to newborn compromise. Similarly, there is no evidence of any long-term risks to the health and well-being of the women or their babies. Given that there was no evidence of increasing resource consumption with the intervention, the authors believe that it could be straightforward to implement and therefore realise benefits for a substantial number of women having their first babies throughout the UK and other countries.

Monitoring adherence

Another observation from the conduct of this trial is whether or not a trial management group can and should monitor adherence. In general, adherence is often monitored by the DMC, which is the only group that sees any data by allocated intervention. For drug trials, in which monitoring of adherence may be relatively straightforward, this may not be a problem. But for interventions that are less well circumscribed, such as that in BUMPES, it is important that people with a good knowledge of the trial and its interventions are able to monitor adherence to ensure that there is reasonable separation between the arms of the trial. This can be achieved by the DMC; however, given the complexity of monitoring adherence to this intervention, particularly with respect to taking account of clinically acceptable reasons for periods of non-compliance, such as fetal blood sampling, the nature of DMCs (in that they tend to meet annually) means that the DMC is removed from the conduct of the trial, which may be a disadvantage. In a pragmatic trial, there is a clear reluctance to impose strict adherence rules, but it is helpful and useful to identify centres or particular individuals with very poor adherence so that behaviour can be modified. In BUMPES, with 41 centres, it was possible to monitor adherence by centre and identify at least one outlier centre that had particularly poor adherence to one of the allocated arms, which was at odds with all the other participating centres. This difference did not come to light until the trial investigators had received agreement from the DMC that the trial management group should be able to monitor adherence by allocated intervention on an ongoing basis. We believe that there is an important lesson to be learnt here about the nature of monitoring adherence, and agreements should be in place early in the trial on who should undertake this, what parameters should be monitored and how this should be overseen carefully to ensure that the degree of monitoring is reasonable, and that monitoring adherence cannot be used by the trial management group to make any inferences about outcomes.

Rationale for nested study

The return rate of 1-year follow-up questionnaires for BUMPES was lower than expected and the use of incentives to evaluate the impact on questionnaire return rate was discussed by the TSC. It was noted by the TSC that there was a lack of evidence relating to the value of incentives to increase collection of follow-up data in trials. As the evidence base for incentives included a variety of populations, it could be that postnatal women respond differently. A proposal for a nested study within a trial (SWAT) was therefore developed.

Background

Maximising follow-up rates for postal questionnaires for RCTs is an important aspect of a well-designed and well-conducted study. Loss to follow-up can lead to bias and compromise the internal and external validity of the results.

Use of incentives to promote questionnaire return in clinical trials has been researched. Existing systematic reviews suggest that they are effective,65,66 but not all studies have sufficient funds to use them. Promising an incentive once data are returned can reduce the cost burden of this approach. Brueton et al. 66 found that an offer of a monetary incentive was comparable with the addition of a monetary incentive with the questionnaire (pooled RR 1.04, 95% CI 0.91 to 1.19). However, it may be possible to provide further cost-savings if the offer was restricted to the reminder letters only.

We evaluated the effect of promising a monetary incentive at first mail-out versus a promise on reminder letters only, with the incentive being posted out on receipt of a completed follow-up questionnaire.

Objective

To assess the effectiveness on the return rate of the 1-year follow-up postal questionnaires for BUMPES of the promise of a monetary incentive made at the point of sending the questionnaire for the first time compared with a promise made on reminder letters only.

Trial design

Parallel-group RCT nested within BUMPES.

Study setting

All women randomised into the BUMPES study who consented to be contacted at 12 months and who had not yet been sent their 1-year follow-up questionnaire were included.

Participant eligibility

Interventions

Women were randomly allocated to the following two groups.

1. Incentive cover letter: this contained details of the promise of a monetary incentive when the questionnaire was first sent. A £10 gift voucher redeemable at high-street shops was sent to the woman on return of a completed questionnaire. The cover letter included a sentence explaining that the voucher was to thank participants for their time and effort. All reminder letters included a sentence about the incentive.

2. Incentive reminder letters: the standard cover letter did not mention any incentive. All subsequent reminder letters sent if the questionnaire was not returned detailed the promise of an incentive. A £10 gift voucher redeemable at high-street shops was sent to the woman on return of a completed questionnaire.

For both groups, women were contacted electronically and via text messaging if the contact details had been collected. The content of the e-mails and texts sent reflected the group to which the woman was randomised. All women were provided with an option of completing the questionnaire online.

Outcome measure

Data collection

Recording of questionnaire receipt, date received and voucher sent was made using internal trial administration systems. Postal versus online receipt was also recorded.

Sample size

The sample size was predetermined by the numbers of questionnaires remaining to be sent at the point of start of the nested study.

BUMPES started recruiting in October 2010 and finished in January 2014. A total of 3236 women were randomised. It was estimated that approximately 1150 women remained to be followed up at the start date of this study (beginning August 2014). Assuming that approximately 15% of these women would be excluded from receiving the questionnaire due to stillbirth, infant death, or address details unknown or different from the infant, 980 women would be eligible to be randomised in the nested study (approximately 490 per group).

In order to assess the detectable effect size possible with the given sample size, we estimated the control group risk based on current literature. Khadjesari et al. 67 investigated the use of an offer of an incentive [a £10 Amazon (Amazon.com, Inc., Bellevue, WA, USA) gift voucher] versus no offer of an incentive on follow-up rates in an online trial. They found an increase of 9% (95% CI 5% to 12%) when using the offer of an incentive. Kenyon et al. 68 investigated the use of a monetary incentive included in reminder letters versus no incentive and found an improvement in the response rate between the two groups of 11.7% (95% CI 4.7% to 18.6%).

The follow-up questionnaire return rate for BUMPES up to June 2014 was 59%. Assuming that this could increase by at least 5% with use of the offer of an incentive either with an incentive cover letter or with an incentive reminder letter only, a sample size of 980 would be sufficient to demonstrate an increase in questionnaire return rate of 8% from 64% in the incentive reminder letter group to 72% in the incentive cover letter group at a two-sided 5% significance level with 80% power. Figure 13 illustrates the proportion detectable in the incentive cover letter group for control group risk varying between 60% and 70%. The detectable difference lies between 8% and 8.5% for varying control group estimates.

FIGURE 13.

Proportion in incentive cover letter group by proportion in control group. Significance = 0.05; power = 0.80; sample size per group = 490.

Randomisation

Allocation was by computerised random number generation stratified by BUMPES allocation and by centre. Randomisation to incentive cover letter or incentive reminder letter occurred at each woman’s next follow-up point during the conduct of the BUMPES study, with a block size of four. Each BUMPES participant was randomised to incentive cover letter or incentive reminder letter once only.

Blinding

Trial staff were aware of allocation as a result of the nature of the interventions and the practicalities involved in sending the letters and the vouchers.

Statistical analysis

For all analyses, participants were analysed in the groups into which they were randomly allocated, that is, comparing outcomes for women allocated to the incentive cover letter with outcomes for women allocated to the incentive reminder letter, regardless of allocation received.

All analyses were based on all women randomised for whom data were available.

The flow of participants through the trial was summarised using a CONSORT flow diagram (see Figure 14). Specifically, the number of women recruited to the BUMPES main trial and subsequently recruited to the BUMPES SWAT is reported along with reasons for not being included in the SWAT.

Participants in the two randomised groups are described separately with respect to baseline demographics and clinical characteristics, including the primary outcome for the main BUMPES study, and recorded on the BUMPES DCB.

Numbers (with percentages) for binary and categorical variables and means (and SDs) or medians (with lower and upper quartiles) for continuous variables are reported.

The return rate and chase rate before the introduction of the randomised interventions (i.e. before the SWAT started) and at the end of the study (with both SWAT trial arms combined) are presented using numbers and percentages.

The return rate and chase rate by method of completion (online vs. postal) are described by trial arm using numbers and percentages.

An adjusted analysis was performed on the two return rate outcomes adjusting for centre (the stratification factor at randomisation) as a random effect. The analysis was carried out using log-binomial regression models and results are presented as adjusted RRs with 95% CIs.

To examine whether or not the effect of when vouchers were sent was consistent across specific subgroups of women, a subgroup analysis by IMD quintile was prespecified. Results are presented as RR plus 95% CI for each subgroup, by intervention group, with the p-value for the statistical test of interaction.

Stata/SE for Windows (version 13.1) was used for all analyses.

Results

Randomisation to the incentive nested study started on 31 July 2014 and continued until all questionnaires and reminders had been sent (last letter sent 6 March 2015). The total number of women in the SWAT was 1026. Eight women were excluded from the analysis as it was discovered after they had been randomised to the SWAT that they had changed address (Figure 14).

FIGURE 14.

Nested study participant flow diagram. a, If not contactable after 15 months since randomisation, questionnaire was not sent.

Balance between the SWAT trial arms in baseline characteristics and centre of recruitment to BUMPES was good. There were only small imbalances in onset of labour (spontaneous or induced), diagnosis of pre-eclampsia and SVB (the BUMPES primary outcome) (Table 29).

The percentage of questionnaires returned before the SWAT started was considerably lower than the overall percentage returned from participants included in the SWAT (55.6% vs. 73.0%, respectively). This trend is also seen in the percentage returned without any reminder letters being sent [35.3% vs. 46.8%, respectively (Table 30)].

Return rates by postal and online completion between the two SWAT arms are presented in Table 31. A total of 152 questionnaires (20.5% of all questionnaires returned) were completed online, with slightly more being returned online in the reminder letter group than in the cover letter group (18.0% vs. 23.0%).

Figure 15 and Table 32 present the percentages of questionnaires returned, according to how many times a reminder letter was sent, and broken down by postal versus online completion. A higher percentage of questionnaires were returned without chasing after receipt of a cover letter promising an incentive than in the group receiving a standard cover letter (51.5% vs. 42.1%, respectively). However, if a reminder was sent, fewer women returned the questionnaire in the group receiving the promise of an incentive in the cover letter than those receiving a promise in the reminder letter (11.5% vs. 14.6%, respectively, for the first reminder, and 8.9% vs. 11.5%, respectively, for the second reminder).

FIGURE 15.

Percentages of questionnaires returned according to how many times a reminder letter was sent, broken down by postal vs. online completion.

For the primary outcome, the percentage of questionnaires returned overall for those receiving the incentive cover letter was slightly higher than those receiving the incentive reminder letter (74.2% vs. 71.8%, respectively), but this was not statistically significant at the 5% level (adjusted RR 1.03, 95% CI 0.96 to 1.11) suggesting no evidence of an effect resulting from when the incentive is offered (Table 33). However, women who received a cover letter promising an incentive were more likely to return their questionnaire without a reminder letter being required than those who received a standard cover letter (adjusted RR 1.22, 95% CI 1.07 to 1.39). The mean difference in the total cost of the vouchers was £4.56 (95% CI £4.02 to £5.11), with the cost being higher in the group receiving the standard cover letter.

The pre-specified subgroup analysis is presented as a forest plot in Figure 16. There is no evidence of heterogeneity for IMD subgroups for the primary outcome of overall response rate (p = 0.43).

FIGURE 16.

Subgroup analysis for questionnaires returned.

Discussion

In this SWAT, there is no evidence to suggest that the offer of a monetary incentive at first mail-out, compared with only when a reminder letter is sent, makes a substantial difference to the overall return rate of a 1-year follow-up postal questionnaire. Although slightly more questionnaires were returned in the group receiving the offer at first mail-out (an absolute difference of 3.4%), the corresponding RR of 1.03 (95% CI 0.96 to 1.11) was not statistically significant at the 5% level.

The return rate for women included in the SWAT compared with that before the SWAT was introduced showed a marked improvement (absolute difference 17%). Although this is not a randomised comparison, it is consistent with that found by Kenyon et al. 68 in a randomised study investigating the inclusion of a high-street voucher versus no voucher sent with a reminder letter to parents of 7-year-old children, which showed an increase in the return rate of 11.7% (95% CI 4.7% to 18.6%). This study is included in a systematic review that showed that the addition of monetary incentives was more effective than no incentive at increasing response rates to postal questionnaires (RR 1.18, 95% CI 1.09 to 1.28). 66

This SWAT used a £10 high-street gift voucher as a monetary incentive. The mean cost of vouchers per participant was greater in the group receiving the offer at first mail-out (£7.50 vs. £3.00). Coupled with the lack of evidence of a difference in the overall return rate, this would indicate that sending the offer of an incentive with a reminder letter only is a cost-effective approach to improving return rates. However, there is evidence to suggest that the return rate without requiring reminders is higher in the group for whom the incentive is offered in the first mail-out (absolute difference 9.4%). The cost of administering the additional reminder letters was not calculated, but is a serious consideration that would need to be offset against the expected cost of the vouchers and could depend on the resources available as well as the sample size of the study.

There are ethical issues to consider with the approach of only sending an offer of an incentive to those participants who do not return their questionnaire promptly. Consideration should be given to the chance that participants in a study may communicate with each other, and share their experiences regarding whether or not they received an incentive.

This is the first known SWAT to investigate the use of incentives for improving questionnaire return rates in a population of first-time mothers with infants around 1 year old. This study suggests that offering incentives when a reminder is required could be cost-effective depending on the sample size of the study and hence the resources required to administer the reminder letters.

Contributions of authors

Professor Debra Bick (Professor of Evidence Based Midwifery Practice, King’s College London): trial design and oversight, advice on measures of maternal satisfaction and interpretation of the results.

Dr Annette Briley (Consultant Midwife, Guy’s and St Thomas’ NHS Foundation Trust): trial oversight and interpretation of the results.

Professor Peter Brocklehurst (Professor of Women’s Health, UCL): trial design, oversight, analysis and interpretation of the findings. Responsible for the first draft and co-ordination of the production of this report and is its guarantor.

Ms Pollyanna Hardy (Senior Trials Statistician, National Perinatal Epidemiology Unit, University of Oxford): trial design, oversight of production of SAP, analysis and interpretation of the results.

Professor Edmund Juszczak (Associate Professor–Director, National Perinatal Epidemiology Unit, University of Oxford): trial design and oversight, statistical advice and interpretation of the results.

Ms Lynn Lynch (Midwifery Lecturer, Cardiff University): co-ordination of local staff at all of the participating centres, central day-to-day management of the centres and interpretation of the results.

Professor Christine MacArthur (Professor of Maternal and Child Epidemiology, University of Birmingham): trial design, oversight, advice on measures of post-partum well-being and morbidity, and analysis and interpretation of the findings.

Dr Phillip Moore (Consultant Anaesthetist, University Hospital Birmingham NHS Trust): trial design, oversight, advice on anaesthetic issues, analysis and interpretation of the findings.

Professor Mary Nolan (Professor of Perinatal Education, University of Worcester): trial design, oversight, advice on participant information material, analysis and interpretation of the findings.

Professor Oliver Rivero-Arias (Associate Professor/Senior Health Economist, National Perinatal Epidemiology Unit, University of Oxford): oversight of production of health economics analysis plan, analysis and interpretation of the results.

Dr Julia Sanders (Consultant Midwife/Reader in Midwifery, Cardiff University): trial design, oversight, analysis and interpretation of the findings.

Professor Andrew Shennan (Professor of Obstetrics, King’s College London): trial design, oversight, analysis and interpretation of the findings.

Dr Matt Wilson (Consultant in Obstetric Anaesthesia/Senior Lecturer in Anaesthesia, Sheffield Teaching Hospital/University of Sheffield): trial design, oversight, advice on anaesthetic issues, analysis and interpretation of the findings.

Co-investigators group

• Professor Debra Bick, Professor of Evidence Based Midwifery Practice, King’s College London.

• Dr Annette Briley, Consultant Midwife, Guy’s and St Thomas’ NHS Foundation Trust (replaced Geraldine O’Sullivan in 2012).

• Professor Peter Brocklehurst, Professor of Women’s Health at UCL.

• Oya Eddama, Health Economist, National Perinatal Epidemiology Unit, University of Oxford.

• Professor Janesh Gupta, Professor/Honorary Consultant in Obstetrics and Gynaecology, Birmingham University/Birmingham Women’s Foundation NHS Trust.

• Pollyanna Hardy, Senior Trials Statistician, National Perinatal Epidemiology Unit, University of Oxford.

• Professor Edmund Juszczak, Associate Professor–Director, National Perinatal Epidemiology Unit, University of Oxford.

• Lynn Lynch, Senior Research Midwife, Cardiff University.

• Professor Christine MacArthur, Professor of Maternal and Child Epidemiology, University of Birmingham.

• Professor Rona McCandlish, Epidemiologist: Maternal Health, National Perinatal Epidemiology Unit, University of Oxford (until 2012).

• Dr Phillip Moore, Consultant Anaesthetist, University Hospital Birmingham NHS Trust.

• Professor Mary Nolan, Professor of Perinatal Education, University of Worcester.

• Dr Felicity Plaat, Lead Clinician and Consultant Anaesthetist, Queen Charlotte’s and the Hammersmith Hospital/Senior Lecturer, Imperial College London.

• Dr Dean Regier, Senior Health Economist, National Perinatal Epidemiology Unit, University of Oxford (until 2012).

• Dr Julia Sanders, Consultant Midwife/Reader in Midwifery, Cardiff University.

• Professor Andrew Shennan, Professor of Obstetrics, King’s College London.

• Dr Geraldine O’Sullivan, Lead Clinician in Obstetric Anaesthesia, Guy’s and St Thomas’ NHS Foundation Trust (deceased 2012).

• Dr Matt Wilson, Consultant in Obstetric Anaesthesia/Senior Lecturer in Anaesthesia, Sheffield Teaching Hospital/University of Sheffield.

Trial Steering Committee

Independent members:

• Dr Paul Howell, Consultant Anaesthetist, St Bartholomew’s Hospital

• Professor Dame Tina Lavender, Professor in Midwifery, University of Manchester

• Professor Alan Montgomery (Vice-Chairperson), Professor of Medical Statistics and Clinical Trials, Nottingham Clinical Trials Unit

• Professor Stephen Palmer, Professor of Health Economics, University of York

• Ms Justine Pepperell (Consumer Representative)

• Professor Steve Robson (Chairperson), Professor of Fetal Medicine, Medical School, University of Newcastle.

Data Monitoring Committee

Independent members:

• Professor Christine Kettle, Professor of Women’s Health, Academic Unit of Obstetrics and Gynaecology, University Hospital of North Staffordshire and Staffordshire University

• Mr Stephen Walkinshaw (Vice-Chairperson), Consultant in Maternal and Fetal Medicine, Liverpool Women’s NHS Foundation Trust

• Dr Steve Yentis (Chairperson), Consultant Anaesthetist, Chelsea and Westminster Hospital

• Dr Pat Yudkin, Emeritus Reader in Medical Statistics, University of Oxford.

Clinical Trials Unit Staff

Independent midwife assessor for clinical coding, London

Sarah Bryan

Recruiting centres

Centre: Arrowe Park Hospital, Wirral. Principal investigator: Suresh Singaravelu. Recruiting staff: Carly Nulty, Carolyn Bragg, Gerri Griffiths, Helen Burghall, Jane Murphy, Julie Dale, K Mccoy, Lynne Lacy, Rachel Roberts, Sandra Hutcheon, Tanya Wynne, Tracy Green and Vickie Heller.

Centre: Bedford Hospital. Principal investigator: Yaqub Latoo. Recruiting staff: Anita Males, Anni Price, Babs Harris, C Dyer, Carla Ball, Carol Handrahan, Donna James, Elizabeth Carlyle, Emma Clarke, Hayley Smith, Jen Welsh, Jenny Cowland, K Emery, Katie Summers, L Church, Liz Dodd, Lucy Wills, Marion Moore, Melissa Coles, Paula Griffiths, Rachel Pressley, Rebecca Adcock, Ruth Croot, Ruth Steward, S Harris, Sarah Gates, Sarah Johnson-Clarke and Sue Hill.

Centre: Birmingham Women’s Hospital. Principal investigator: Phil Moore. Recruiting staff: A Connolly, Alexandra Bellamy, Anna Zhao, Anya Chruscinska, B Oniono Kuafor, Becky Cullen, Bethany Lunn, Bobby Sharma, C Bishop, C Graves, Charlotte Bowman, Charlotte Davies, Charlotte Hinks, Chloe O’Hara, Claire Bissell, Debbie Baker, Deborah Robinson, Elisha Randell, Elizabeth Ewers, Emily Byrne, Emma Wright, Erica Henry, Fiona Musgrave, G George, Gemma Barnfield, Gemma Wadsworth, Hannah Wood, Harriet Fisher, J Rose, Jane McNair, Jennifer Prescot, Jenny Pledger, Jess Shaw, Jessica Gregory, Joelle Rowland, Juillete Webster, Julia Cheshire, Justine Craig, K Horton, Karen Davies, Karen Elkin, Katie Freitas, Kerry Hudson, Kimberley McMahon, Kiranjit Sehmi, Kirsten Emson, Kirsty Elwell, Laura Mulryan, Lauren Hill, Lauren Webley, Lisa Salmon, Lucy O’Grady, Madeleine Parry, Margarita Bariou, Maureen Joseph, Megan Corbett, Michaela Dzioba, Michaelene Cole, Michelle Bennett, Michelle Neal, Nichol Ross, Nicola Mcenery, Nikki Robbins, Novia Samuels, Orphelia Atkins, Pam Simpson, Paula Trinham, Rachel Singer, Rebecca Cullen, Rebecca Gallimore, Rebecca Leon, Rebecca Mckenzie, Rhea Bond, Ruth Cavey-Wilcox, Samuel Todd, Sarah Blythe, Sarah Ketley, Sasha Hamilton, Sethenia Beckford, Sian Wilkie, Stella Bibb, Stephanie Henry, Teresa Vann, Tracey Bond, V Preece and Victoria McDonagh.

Centre: Bradford Royal Infirmary. Principal investigator: Diane Farrar. Recruiting staff: Alexandra Fozard, Alice Tunney, Alison Chapman, Amanda Wilson-Thompson, Aongola Ngenda, Carrie Owens, Christina Scott, Clare Cummings, Fozia Arshad, Geraldine Atkinson, Gillian Butterfield, Heather Darlow, Helen M Sharp, Jennie Robertshaw, Joanne Mortimer, Joanne Watson, Josephine Hartley, Judith Lowther, Katarzyna Denkiewicz, L Jarockyj, Laura Mckenna, Lisa Thompson, Lydia Brookes, Maryanne Naylor, Nicola Davies, Rachael Jones, Rachel Hemers, Rachel Wild, Rebecca Skelton, S Marriott, S Nicholson, Sania Iqbal, Stacey Ryles, Susie Weekes, Talitha Grandison, Tracey Germaine-Rylance and Victoria Jones.

Centre: Brighton Royal Sussex County Hospital. Principal investigator: Vanessa Fludder. Recruiting staff: Claire Miles, Emma May, Florence Crawley, Hannah Tomms, Igone Sesma, Jayne Denyer, Kate McCambridge, Melanie Wight, Nichola Tuck, Omotoyin Awonuga, Paula Alonso-Gonzalez, Rachael Chatterton, Rachel Cox, Rosie Darling and Victoria Wright.

Centre: Dorset County Hospital, Dorchester. Principal investigator: Christine Grother. Recruiting staff: Allison Hamilton, Bev Robertson, Carly Smith, Christine Grother, Hilary Fletcher, Jane Linger, Jo Hartley, Julie Bonifacio, Julie Younger, Karen Myers, Kathryn Dyer, Louise Shreiber, Nichola Coliandris, Sarah Haigh and Tina Parker.

Centre: Frimley Park Hospital. Principal investigator: Karen Plews. Recruiting staff: Abbey Ford, Alexandra Mayrs, Alison Shilton, Amanda Dowling, Anna Holland, Anna Kemsley, C Green, Cara Gambon, Catherine Bressington, Charlie Thompson, Christina Longman, Cindy Port, Claire Smith-White, Danielle Perkins, Deidre Hussey, Di Wyeth, Fiona Allison, Florence Chauyara, Frances Warner, Gifty Dadzie, Hannah Brown, Helen Hawkins, Irene Tan, J Willard, Jackie White, Jaimie Sutherland, Jannine Bailey, Jayne Moss, Jenny Evans, Jessica Main, Jo Green, Joanna Broomham, K Wren, Karen Plews, Karen Spencer, Karen Wrigley, Kasia Russel, Katie Harrald, Kirsty Fisher, Laura Kirk, Lauren Barnett, Lauren Bartlett, Leah Pueschel, Lisa Rudall, Liz Treen, Louise Wylie, Melanie Taylor, Michelle Chuter, Michelle Hardy, Michelle Mcloughlin, Michelle Nicholls, Naomi Davies, Nicola Wimshurst, Pat Harris, Paula Ball, Penny Schnabel, R Warner, Rachel Bright, Rachel Rouen, Rebecca Beddows, Ruth Beesley, Sabine Everett, Sarah Heath, Saras Bishop, Sham Shelbourn, Sharon Phipps, Shian Fethney, Sophie Adams, Sophie Hutton, Susan Meyjes, Tabitha Stuthridge, Tania Gaffney, Theresa Thomas, Tina Longman, Tracy Hopkins, Vicky Donovan, Vivienne Novis, Wendy Bascal and Zoe Farr.

Centre: Gloucestershire Royal Hospital. Principal investigator: Louis Khor. Recruiting staff: Angela Smith, Angela Stevens, Bryony Bell, C Pearson, Carly Avis, Caroline Broadhurst, Cassy King, Chantel Coleman, Charlotte Barwick, Cody Allen, Ellen Reeves, Fiona Liddle, Hayley Marvill, Jane Barradale-Smith, Jane Soule, Jenna Surman, Jennifer Pratley, Kerrie Lotsu, Lisa Frattolillo, Louise Broadley, Lucy Broad, Michelle Dimery, Michelle Partington, Nikki Dobson, Nina Kellow, Rachael Harris, Rachel Midwinter-Marland, Ridwana Pandor, Sarah Pilcher, Sian Harrington and Tracey Miller.

Centre: Great Western Hospital, Swindon. Principal investigator: Tracey Sargent. Recruiting staff: Amy Pitcher, Angela Bunce, Anita Long, Ashley Heal, Danielle Heywood, Debra Hunt, Denise Selby, Elaine Price, Jade Gordon, Jane McGregor, Jennie Hone, Joanna Coulson, Joanne Lewis, Julia Sewell, Kate Welsh, Kathryn Owen, Kelly Greenslade, Kimberley Tubb, Lisa Nicholson, Nicola George, Rachel Ravati, Rebekah Tollafield, Ruth Davies, Sara Brown, Sophie Stewart, Susan Bint, Tamara Byrne, Tanya Miles, Victoria Norman, Viv Cutler and Zhilla Majadabadi.

Centre: Hillingdon Hospital. Principal investigator: Jane Terry. Recruiting staff: Denise Ahmed, Emma Speirs, Fiona White, IR Howarth, Kirsty Griffith, Kirsty Stark, Licricia Ngahan-Tchaptchet, M Lawlor, Manjit Matharu and Sarah Bell-Ryan.

Centre: James Paget University Hospital, Great Yarmouth. Principal investigator: Mumtaz Rashid. Recruiting staff: Andrea Bedford, Angela Oram, Ann Pye, Caroline Fox, Elsie Gibbs, Emily Boyce, Emily Cole, Faye Hewitt, Helen Cullimore, Jane Ward, Jennifer Thompson, Joan Timewell, Kerry Burwood, Laura Jarrett, Lauren Goodfellow, Lesley Yates, Nora Hassan, Pauline Studley, Sinead Osborne, Sophie Neville, Tracey Porter and Victoria Reeve.

Centre: King’s College Hospital, London. Principal investigator: Cathy Walton. Recruiting staff: A Tully, Agnes Kimbowa, Alice-Amanda Hinton, Amelia Evans, Anna Mazzarelli, Bridget Rance, Bryanna Chenoweth, C Beckmann, Christine Murphy, Clare Patterson, Clemmie Hooper, Dorcas Appah, Dorisilla Adolwa, Emily Stockton, Erica White, Erika Glenny, Esther Annan, Halina Szajna, Iqra Khan, Katherine Clark, Katrice Currie, Kelly Macfadyen, Laura Santos, Mary Bollard, Mary Obud, Michelle Lynch, Modupe Adebayo, Olivia Snowball, Omatalani R Sangare, Rachael Waldron, Rachel Barlow, Rebecca Macleod, Rebecca Manners, Ruth Graham, Ruth Landis, Sadie Holland, Sarah Skivens, Sarah-Ann Evans, Sophie Halton-Nathan, Sophie Steward, Stacey Robinson, Sue Byrne and Susie Urquhart.

Centre: Kingston Hospital. Principal investigator: Arezou Rezvani. Recruiting staff: Alexandra Frost, Alice Cox, Amanda Carey, Amisha Chauhan, Anne-Marie Greaves, Annie Stott, B Hellmich, Bronwen Kenward, C Caton, Charlotte Rose, Chloe Du Parcq, Elka Dimitrova, Fiona Smith, Hui Tam, Jana Durtova, Jennifer Slee, Joanna Pitcher, Laura Grainger, Laura Sowell, Leanne Bateman, Lizzanne Roman, Louise Jones, Lucy Bellinger, Lyndsey Smith, Morwenna Trevan, N Haysum, Nicky Ni, Nicole James-Lowe, Nikki West, Perrine Dhaisne, Rachel Bell, Rachel Rolfe, Rebekah Hoadley, Rosemary Mukasa, Ruth Sentenga, Sam Frewin, Sarah Lowe, Scarlett Beland-Tyce, Sophie Wismer, Suesan Beirouty, Susan Leahy, Susan O’Callaghan, Toni Brown, V Gunawardena and Zandra Rubia-Mendoza.

Centre: Lewisham Hospital, London. Principal investigator: Frances Jones. Recruiting staff: Ellen Madamombe, Jade Johnson, Kay Holford, Mabinty Leigh Sian Turner and Suzannah Sheerin.

Centre: Medway Maritime Hospital, Gillingham. Principal investigator: Dorothea Smith. Recruiting staff: Andrea Curling, Belinda Newman, Deborah Simmons, Debra Nwosa, Dorothy Smith, Helen Jones, Jane Simmons, K Ashwell, Kerry Sturgess, Lovelace Oti, Ludmila Wilson, Lynn Deller, Michelle Hayes, N Jones, Patricia Chaplin, Sarah-Jane Cross, Sharon Small, Tatjana Molotkova, Valerie Andrew and Zoe Wood.

Centre: Nevill Hall Hospital, Abergavenny. Principal investigator: Louise Taylor. Recruiting staff: Amanda Lisle, Andrea Priddle, B Markey, Cath Barwise, Deb Oliver, Kath Barwise, Kerry Owen, Kim McKie, Leanne Ball, Mandy Jones, Pippa Nicholas, Sue Jordan and Wendy Howells-Smith.

Centre: New Cross Hospital, Wolverhampton. Principal investigator: Gowri Simon. Recruiting staff: Becci Leathley, Claire Morgans, Deborah Brettle, Ellen Poniatowska, Emma Clews, Hazel Peden, Hazel Remmet-Booth, Jane Hussellbee, Joanne Ridley, Julia Icke, Karen Evans, Laura Brooks, Louise Wood, Marni Fassett, Pip Grocott, Siva Basra and Tracy Willetts.

Centre: North Manchester General Hospital. Principal investigator: Viv Dickinson. Recruiting staff: Andrea Kerti, C Sinclair, Catherine Holt, Catherine Hughes, Colette Robinson, Collette Riley, Dawn Littler, E Foley, Ellie Cardnell, Emma Baxter, Emma Groom, Emma Park, G Charles, Georgina Cartridge, Helena Spencer, Jean Davis, Jenna Haycocks, Jo Jakubowski, Joanna Ward, Julie Ainsworth, Julie Whitby, Karen Robbins, Kate O’Hagan, Kerri Delaney, Laura Ashton, Laurene Mannix, Lileath Fisher, Louise Blinkhorn, Lyndsay Yates, Lynsay Ingham, Margaret Kerins, Michelle Salt, Pam Whittle, Rachel Tully, Sacha Jackson, Stella Oakes, Tina Affleck, Valerie Julie Walker and Zoe Davies.

Centre: Pinderfields Hospital, Wakefield. Principal investigator: Nicolas Akerman. Recruiting staff: Anna Warburton, Caroline Paterson, Chiew Poskitt, Rachel Stock, Muhammad Faisal Ehsan, Elise McShee, Jacqueline Edge, James Bedford, Jill Greenwood, Julie Wormstone, Karen Simeson, Laura King, Marie Knox, Michelle Mowbray, Penny Barker, Rosalind Morley, Stacey Dunn, Sue Winrow, Tracy Langcake, Ujala Ahmad and Veronica Walker.

Centre: Prince Charles Hospital, Merthyr Tydfil. Principal investigator: Liz Edwards. Recruiting staff: Bev Jones, Catherine Bush, Ceri Hill-Jones, Diane Lewis, Jodie Hodges and Theresa Jones.

Centre: Princess of Wales Hospital, Bridgend. Principal investigator: Sarah Fox. Recruiting staff: Angela Davies, Annette Jones, Christie-Ann Lang, Donna Hall, Elinor Taylor, Elizabeth James, Emily Grainger, Gemma Griffiths, Hannah Lambert, Joy James, Julie Roberts, Kate Richards, Kathryn Greaves, Lauren Yaw, Lynne Grieves, Mari Davies, Megan Cave, Michelle Bassett, Rachel Morgan, Sarah Jones, Sian Middleton and Tracey Bowman.

Centre: Queen Alexandra Hospital, Portsmouth. Principal investigator: Aneeta Sinha. Recruiting staff: Amy Pollard, Andrea Grey, Ann Going, B Edge, Beryl Pullen, Carol Richardson, Carole Longley, Ella Edwards, Ellie Jenkins, Emma Connelly, Emma Kellagher, Fiona Moore, Genevieve O’Docherty, Gill Allen, Isla Campbell, Isobel Murtagh, Jemma Cave, Jill Hall, Jo Jordan, Jo Warwick, Jodie Carolan, Karen Darr, Karen Munks, Karen Wellspring, Katherine O’Mara, Katrina Walker, Kim Leonard, Laura Davis, Linda Lishman, Lucy Galloway, Lulu Russell Smith, Lynda Baker, Lynne Groves, Mandy Whittle, Mary Taylor, Mary-Ann Sheehan, Melanie Say, Mo Turnbull, Naomi Simpson, Penny Bone, Penny Cox, Sally Griffiths, Sarah Burr, Sharon Buttriss, Suzanne LeBrocq, Tracey Hall, Tracey Lasisi, Vanessa Garlish, Wendy Bessant, Wendy Marsh and Zoe Garner.

Centre: Queen Charlotte’s and Chelsea Hospital, London. Principal investigator: Felicity Platt. Recruiting staff: Alice Gautreau, Grace Kember, Igbeka Hayes, Karen McArtney-Roberts, Lisa Rickwood, Lucy Simpson and Suzanne Ridley.

Centre: Queen Elizabeth Hospital, Kings Lynn. Principal investigator: Anoop Surendran. Recruiting staff: Beverley Golding, Caroline Tucker, Catherine Bent, Debby Ramsdale, Donna Allen, E Cervi, Emma Chapman, Helen Parker, Jacinta Baptista, Jean Keen, Jodie Cully, Jodie Jupe, Lisa Gormley, Liz Tyler, Michaela Bouskova, Naomi Seaman, Rachel McCabe, Rosie Hucklesby, S Tennant, S Wingfield, Sarah Russell, Tracey Stafford, Tracy Cooke and Yvonne Fulcher.

Centre: Royal Cornwall Hospital, Truro. Principal investigator: Nila Cota. Recruiting staff: Amy Dunstan, Charlie Fulcher, Dariel Rowe, Eddi Theedham, Jane Parke, Jane Stubbs, Jenny Heron, Jo Bennett, Josie Dodgson, Julie Wallis, Karen Needham, Katherine Holland, Kerry Youngman, Kim Hewlett, Kimberly Fanson, Laura Quinn, Lisa Marchetti, Lizzie Cowan, Lorraine Kennedy-Snaith, M Hobson, M Underwood, NPM Middleton, Samantha Broughton, Sarah Grigg, Suzanne Bryant, Tracey Rowe, Victoria Bassett and Wendy Preen.

Centre: Royal Gwent Hospital, Newport. Principal investigator: Louise Taylor. Recruiting staff: Amie Cook, Beth Comben, Carmen Rubio-Batanas, Chloe Rowsell, Donna Crocombe, Donna Fleming, Elleanor Griffiths, Fay Smith-Warren, Fiona Carter, Helen Bishop, Jane Morgan, Janet Lawson, Jessica Waters, Karen Halford, Lesley-Ann Bushell, Margot Jones, Michelle Haggart, Naomi Martin, Nicola Smith, R Green, Rose Thomas, Sophie Savigar-Jones, Tara Welch and Tracey Griffiths.

Centre: Royal Hallamshire Hospital, Sheffield. Principal investigator: Vicky Wilson. Recruiting staff: Alison Morison, Alison Norris, Amanda Muller, Amy King, Anne Hemingway, Benash Nazmeen, Caroline Dabinett, Carollynn Jones, Carolyn Metcalfe, Cheryl Popovich, Claire Craine, Claire Sayan, Clara Mwatati, Clare Hennessey, Clare Lord, Clare Newton, Dalia Peretz, Deborah Cresswell, Faith Tye, Gill Hunt, Hannah Tebbs, Heather Croft, Helen Frow, Holly Hickman, Jessica Brookes, Jill Parton, Jo Varley, Joy Herdman, Judy Chang, Julia Thackray, Julie Hawksworth, Justine Todd, Kate Fish, Kathleen Farrand-Green, Kay Crowch, Laura Asher, Laura Chadwick, Leanne Rutkowski, Louise Roberts, Lucy Bon, Medeline Mudehwe, Melinda Pagden, P Pokorna, R Nye, Rebecca Bustani, Rebecca Weston, Rio Cooper, Rosie Barker, Sally Dawn, Samantha Young, Sara Calow, Sarah Bell, Sarah Senbeto, Sarah Swift, Steven Mackie, Sylwia Szarwark, Tracy Hobson, Victoria Lee, Victoria Wilkins, Wendy Few, Wendy Murphy and Zoe Riley.

Centre: Royal Sussex County Hospital, Brighton. Principal investigator: Vanessa Fludder. Recruiting staff: Alanna Dunkerton, Emma Peck, Helen Williams, Kate Clark and Rosheen Baker.

Centre: Royal United Hospital, Bath. Principal investigator: Tracey Sargent. Recruiting staff: Angela Fitzpatrick-Nash, Anne Moffatt, Anne White, Annie Collingwood, Ashley Heal, Bridget Dack, Camilla Hawkett, Charley Reschwamm, Christie Harrison, Cindy Stamp, Donna Williams, Ellie Grant, Elly Doyle, Emily Craig, Emma Tanner, Gemma Day, Hannah Cross, Hannah Jewell, Hannah Reid, Helen-Marie Crooks, Hilary Paice, Jane Norris, Jemma Freegard, Jennifer Reid, Jenny Pullen, Jo Waldron, Jo Woodburn, Julia Grant, Karen Doran, Kate Boulton, Katherine Jackson, Kathryn Vosper, Kathy L Holford, Katie Gooding, Kerry Perkins, Kim Miles, Laura Friend, Leah Harrold, Linda Davis, Liz Norton, Martina Grey, Mirella Popescu, Naomi Bonett-Healy, Nora Seager-Wilkendorf, Rachel Brierley, Rachel Coleman, Rebecca Lamb, Rebecca Murdoch, Rebecca Pendry, Rebecca Walsh, Rhian Motean, Rose Jenkins-Hunt, Ruth Branson, Sara Burnard, Sara Driver, Sarah Marks, Sasha Cairns, Sharon Seager, Susan Collins, Tamara Carr-Gomm, Tina Coffey, Tracy Boakes, Wendy Duberry and Wendy Giles-Smart.

Centre: Singleton Hospital, Swansea. Principal investigator: Sarah Fox. Recruiting staff: Amanda Bates, Cath Harris, Danielle Clifton, Ellie Brown, Felicity Curtis, Julie Ellerton, Julie Thomas, Kate Phillips, Kim Hillier, Linda Richards, Lisa Rees, Lucie Warren, Nicky Court, Outi Morris, Rachel Williams, Rebecca James, Rebecca Lewis, Sarah Fox, Sharon Cooling, Sharon Evan, Sian Phillips and Vicki Lennon.

Centre: South Tyneside Hospital, South Shields. Principal investigator: Mrs Shamma Al-Inizi. Recruiting staff: A MacKay, Allison Nicholson, C Greaves, Delia Brennan, Emma Hindes, Helen M Parker, Judith Black, Linda McNamee, Louise Nicholson, Nicola Tindall, Shelley Bowie and Stacey McFarley.

Centre: St George’s Hospital, London. Principal investigator: Asma Khalil. Recruiting staff: Amy Ridout, Angel Segura, Astell Aikines-Aryeetey, Bridget Okereke, Carmen Martin Martinez, Christiana Appeah, Claire Davies, Cristina Perez, Danielle Holbrook, Dede Efueye, Elaine Sheehan, Eleonor Cowlard, Emma Corrigan, Emma Freeman, Emma McCheyne, Erin Hutchings, H Gardner, Iona Hughes, Iryna Santoskkaya-Marsh, Isabel Aylward, Isabelle Cornet, Jessica Welham, Joyce Adu-Amankwah, Judith Mugerwa, Julia Plana Soria, Leila Zahedi, Lorena Santana Cardenosa, N Karali, Natalie James, Ngozi Asika, Ojevwe Owereh, Ophelie Granger, Paula Lavandeira, Paulette Joy Palmer, Raquel Vives Font, Rosie Sands, Roxanne Vidal, Sarah Esegbona-Adeigbe, Silvia Campo, Steve-Samuel King-Inneh, Sylvia Zoldak and Zinab Jalloh-Conteh.

Centre: St Mary’s Hospital, London. Principal investigator: Felicity Platt. Recruiting staff: Birima Darego-Wokoma, Leticia Alvarez, Margarita Lopez-Liesa, Sarah Pilgrem, Suzanne Ridley and Teresa Ribera.

Centre: St Thomas’ Hospital, London. Principal investigator: Annette Briley. Recruiting staff: A Perlepe, A Veitch, Abby Stewart, Abosede Kako-Are, Agatha Okafar, Ailsa Gill, Alhan Javan, Alice Du Preez, Alison Armstrong, Amanda Stephens, Amy Davies, Amy Smith, Ana Elices, Ana Llamas, Anita Meagher, Ann Marie McHugh, Anna Dios, Anna Gaudion, Anna Kenny, Anne Cobell, Annie Bell, Beth King, Birima Darego-Wokoma, Blanca Rodrigo-Ibanez, Bree Cant, Camella Main, Dede Efueye, E Blasse, Edith Onyeneri, Elena Martinez-Zuddin, Eleonora Bruni, Elizabeth Connelly, Elsa Moro, Emilie Grantham, Emily Jelen, Emma Copley, Emma Grey, Emma Grey, Emma Paten, Emmeline Mudford, Erika Duncan, Faye Safari, Florence Awichi, Fran Lawrence, Funso Adegoke, Gail Roberts, Gemma Baillie, Gillian Donaldson, Gloria Brempong, Grace Obwona-Lanana, Hannah Delmar-Addy, Hannah Emerson, Hannah Levy, Hannah Rogers, Hannah Veazey, Harriett Ivey, Hayley Osborne, Henrietta Simire, Holly Hickman, Ida Bradley, Iro Perlepe, Isabelle Arrabel, Jacqueline Mhako, Jane Love, Janet Cooper, Jennifer Tubby, Jess Cavaya, Jess Floyd, Jessica Quaroni, Jessima Cavaya, Jo Hoffmann, Jo McCarthy, Jo Parker, Jordan Anderson, Juliet Falola, Juliette Falolu, Karine Tweddle, Katia Ciccarella, Kaz Herlihy, Kylie Gould, Laura Bridle, Lauren Chandler, Lianne Phipps, Lianne Prior, Lola Shomefun, Louise Higgs, Lydia Gerrie, Madelena Wilders, Maeve O’Connell, Maggie Lee, Mara Bruno, Marcia Trusty, Maria Pipi, Marina Daniele, Marisa Alvarez, Marta Fernandez Diez, Monika Franklin, Moronike Agboola, N Carlin, Namgyal Gonkatsang, Olivet Macfarlane, Olivia Wheeler-Robinson, P Blair, Pauline Jackson, Rachel Grazette, Rosalind Pouteaux, Sarah Driver, Sarah Evison, Sarah Fowlie, Sarah Kensington, Sarah Tanner, Selina Tettey, Sharon Mumford, Sonia Pereina, Sophie Robinson, Stacy Brown, Stefania Andrian, Stella Nanseera, Sue Turner, Sumaira Bashir, Vaishni Moorji, Vic Offredi, Vivienne Gosden, Yemisi Fadoungbo, Yvonne Mcgrath, Zahra Famili, Zainab Jalloh, Zeenath Uddin and Zekiye Degmenlibey.

Centre: Sunderland Royal Hospital. Principal investigator: Kim Hinshaw. Recruiting staff: Amanda Bargh, Carol Forrester, Christine Evans, Claire Liddel, Deb Holmes, Deborah Bonney, Denise Mace, Donna Rodgers, Donna Rogers, Eileen Walton, Hannah White, Janet Rooks, Julie Harris, Julie Taylorson, Karen Hutchinson, Karen L Armstrong, Kate Reedman, Kathryn Evans, Katrina Dowell, Leeann Adey, Linda Adamson, Lisa Wilson, Lyndsey Summerbell, Natalie Graham, Nicola Easton, Pam Cheek, Sheila Ford, Sonia Thompson, Sophie Robson, Stephanie Hepple, Suzanne Stelling and Victoria Young.

Centre: Tameside Hospital, Ashton-under-Lyne. Principal investigator: Gillian Singleton. Recruiting staff: Ann Gibson, Donna Saleh, Felicia Taylor, Gabby Greenwood, Gemma Lumley, Gillian Singleton, Helen Clase, Jackie Tomlinson, Jan Moriarty, Janet Danzi, Karen Rothera, Kate Firth, Kerry Jackson, Lisa Fisher, Louise Nelson, Paula Frazer, Rachel Drain, S Bungaree, Sharon Aldous, Sophie Hook, Teresa Quinn and Tracey Leicester.

Centre: University College Hospital, London. Principal investigator: Belinda Green. Recruiting staff: Abisola Bashua, Amy Tiltman, Ann Esquerdo, Anna White, C Amaning, Christine Haron, Constance Mvududu, Donata Hoesch, Edna Farah Dahir, Eleri Bates, Ellie Sanderson, Emily Nygaard, Fenya Jonas, Hayley Gilroy, Heidi Buhlmann, Ivan Bettinsoli, J Cole, J Evans, Jennifer Lang, Jenny Keys, Jes Surtees, Lucia Fitzsimons, Margarita Akyla, Meghan Jackson, Nellie Sarmiento, Teresa Okemadu and Zahra Khan.

Centre: University Hospital of Wales, Cardiff. Principal investigator: Rachel Collis. Recruiting staff: A Morgan, A Aimee Jones, Aime Symes, Alex Andrews, Alice Fairman, Alice Snell, Alyson Gardiner, Ami Wolstenholme, Amy Garrett, Amy Vaithilingam, Amy Welsh, Angela Amey-jones, Angela Jones, Anika Brodd, Anna Jones, Annie Kitchen, C Swallow, Cara Moore, Cara Moruzzi, Carla Blackshaw, Catherine Downing, Cheryl Cox, Debbie Grey, Debbie Hunt, Debbie Jones, Deborah Powell, Ed Cross, Elaine Patterson, Elin Phillips, Emily James, Emily Shaw, Emma Bull, Felicity Callan, G McElroy, Gloria Lane, Hannah Hills, Hannah Thomas, Helen Lock, Hollie Power, Jane Reid, Jayne Frank, Jenna Parsons, Jenna Terry, Jenny Rickson, Jess Holmes, Joanne Bowen, Jodie Clark, Jude Casey, Julia Morgan, Juliet Grimes, Karen Jennings, Karina Downing, Kate Lynch, Kate Murphy, Kate Siddal, Katherine Fischer, Katherine Williams, Kathryn Smith, Katie O’Bradovic, Katie Stubbs, Kelly Bennett, Kerri Hamblin, Kirsty Jones, Laura Mundy, Laura Rose, Laura Terry, Lauren Quirke, Lieska Hoes, Lindsey Hilldrup, Louise Houghton, Luisa Canale, Lynette Rowlands, Miranda Millett, Misha Harry, Natalie Rees, Nerys Kirtley, Nia Cleal, Nicola Savoury, Nicola Schilling, Patricia Chan, Polly Ferguson, Rachel Bain, Rachel Harry, Rebecca Boselli, Ros Howells, Ruth Leonard, Sally Alqaddo, Sally Meek, Samantha Crouch, Sara Davies, Sarah Heap, Sarah James, Sarah Lucas, Sarah Madley, Sarah Morris, Sarah Spencer, Sherrie Bird, Shirley Goodwin, Sian Jones, Sofia Odugleh, Tamsin Edwards, Tracey Lawrence, Trudy Thomas-Jones, Wendy Hoggan and Zoe Millichap.

Centre: Warrington Hospital. Principal investigator: Rita Arya. Recruiting staff: Alison Quine, Ann Pathmakumar, Ann-Marie Brooke, Ann-Marie Hatton, Cate Fitzpatrick, Cath Kidd, Danielle Stotton, Deborah Fletcher, Debra Clements, Donna Abbott, Eileen Fielding, Einir O’Neill, Elaine Armitage, Hannah Stevens, Hayley Axon, Heather Mee, Helen Ling, Helen Poulton, Jackie Richards, Jayne Wright, Katherine Conquest, Kerry Jones, Lesley Hampson, Linda Hennon, Mags Odell, Marie Wheatley, Mary Cubbon, Mary Hornby, Simone Peters, Susan Evans, Tamsin Hawkins, V Hodson and Vicky Littlewood.

Centre: West Middlesex University Hospital. Principal investigator: Louise Page. Recruiting staff: A Akodu, Adebola Aroboto, Adelaide Adubuffour, Alicia Thomas, Alyson Brown, Amanda Bray, Anna Piasecka, B Collard, B Snee, C Gordon-Jack, Deborah Reid, Eleanor Fraser, F Addow, Fiona Ghalustians, Grace Volo, Hannah Thomas, Ilaria Torre, J Fernandez, Jennifer Ryan, Jessica Howard, Joy Ataderie, Julia Harris, Juliet Joseph, Karen Beck, Karen Lundie, Kelly Mack, Kirsty Dolling, Lisa Takab, M Farrell, Maddie Saunders, Marie Garvey, Marie O’Connell, Marie Oliver, Mercy Batchelor, Neveen Jivan, Nikki Jaques, P Laurence, Pia Tomeldon, Po Ying Li, R Shadna, Renske van Gaans, Revai Chingwa, Risi Akodu, S Harrison, Sally Dauncey, Sally Seaman, Sarah Dixon, Silviya Giffin, Sophie Pike, Tsakani Tshavane, V Henry and W Fambe.

Publications

Bird H. Labour Position Study Launches. The Royal College of Midwives, 14 August 2012. URL: www.rcm.org.uk/news-views-and-analysis/news/labour-position-study-launches (accessed 2 October 2017).

Hardy P, Bell J, Brocklehurst P. Evaluation of the effects of an offer of a monetary incentive on the rate of questionnaire return during follow-up of a clinical trials: a randomised study within a trial. BMC Med Res Methodol 2016;16:82.

The Epidural and Position Trial Collaborative Group. Upright versus lying down position in second stage of labour in nulliparous women with low dose epidural: BUMPES randomised controlled trial. BMJ 2017;359:j4471.

Data sharing statement

Data can be obtained from the corresponding author.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health.

Clinical Investigator Group

• Professor Debra Bick, Professor of Evidence Based Midwifery Practice, King’s College London.

• Dr Annette Briley, Consultant Midwife, Guy’s and St Thomas’ NHS Foundation Trust (replaced Geraldine O’Sullivan in 2012).

• Professor Peter Brocklehurst, Professor of Women’s Health at UCL.

• Oya Eddama, Health Economist, National Perinatal Epidemiology Unit, University of Oxford.

• Professor Janesh Gupta, Professor/Honorary Consultant in Obstetrics and Gynaecology, Birmingham University/Birmingham Women’s Foundation NHS Trust.

• Pollyanna Hardy, Senior Trials Statistician, National Perinatal Epidemiology Unit, University of Oxford.

• Professor Edmund Juszczak, Associate Professor–Director, National Perinatal Epidemiology Unit, University of Oxford.

• Lynn Lynch, Senior Research Midwife, Cardiff University.

• Professor Christine MacArthur, Professor of Maternal and Child Epidemiology, University of Birmingham.

• Professor Rona McCandlish, Epidemiologist: Maternal Health, National Perinatal Epidemiology Unit, University of Oxford (until 2012).

• Dr Phillip Moore, Consultant Anaesthetist, University Hospital Birmingham NHS Trust.

• Professor Mary Nolan, Professor of Perinatal Education, University of Worcester.

• Dr Felicity Plaat, Lead Clinician and Consultant Anaesthetist, Queen Charlotte’s and the Hammersmith Hospital/Senior Lecturer, Imperial College London.

• Dr Dean Regier, Senior Health Economist, National Perinatal Epidemiology Unit, University of Oxford (until 2012).

• Dr Julia Sanders, Consultant Midwife/Reader in Midwifery, Cardiff University.

• Professor Andrew Shennan, Professor of Obstetrics, King’s College London.

• Dr Geraldine O’Sullivan, Lead Clinician in Obstetric Anaesthesia, Guy’s and St Thomas’ NHS Foundation Trust (deceased 2012).

• Dr Matt Wilson, Consultant in Obstetric Anaesthesia/Senior Lecturer in Anaesthesia, Sheffield Teaching Hospital/University of Sheffield.

Trial Steering Committee

Independent members:

• Dr Paul Howell, Consultant Anaesthetist, St Bartholomew’s Hospital

• Professor Dame Tina Lavender, Professor in Midwifery, University of Manchester

• Professor Alan Montgomery (Vice-Chairperson), Professor of Medical Statistics and Clinical Trials, Nottingham Clinical Trials Unit

• Professor Stephen Palmer, Professor of Health Economics, University of York

• Ms Justine Pepperell (Consumer Representative)

• Professor Steve Robson (Chairperson), Professor of Fetal Medicine, Medical School, University of Newcastle.

Data Monitoring Committee

Independent members:

• Professor Christine Kettle, Professor of Women’s Health, Academic Unit of Obstetrics and Gynaecology, University Hospital of North Staffordshire and Staffordshire University

• Mr Stephen Walkinshaw (Vice-Chairperson), Consultant in Maternal and Fetal Medicine, Liverpool Women’s NHS Foundation Trust

• Dr Steve Yentis (Chairperson), Consultant Anaesthetist, Chelsea and Westminster Hospital

• Dr Pat Yudkin, Emeritus Reader in Medical Statistics, University of Oxford.