Improving quality of care and outcome at very preterm birth: the Preterm Birth research programme, including the Cord pilot RCT

Article history

The research reported in this issue of the journal was funded by PGfAR as project number RP-PG-0609-10107. The contractual start date was in March 2011. The final report began editorial review in April 2018 and was accepted for publication in June 2019. As the funder, the PGfAR programme agreed the research questions and study designs in advance with the investigators. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PGfAR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Lelia Duley reports National Institute for Health Research (NIHR) Clinical Trials Unit support funding. Jon Dorling reports grants from NIHR during the conduct of the study, grants from NIHR and grants from Nutrinia Ltd (Ramat Gran, Israel) outside the submitted work. He also reports membership to the Health Technology Assessment (HTA) Efficient Study Designs 2015–16, HTA General Board 2016–19, HTA Maternal, Neonatal and Child Health (MNCH) Methods Group 2014–18, HTA MNCH Panel 2013–18, and HTA Post Board Funding Teleconference 2015–18. Andrew Weeks reports grants from Liverpool Women’s Hospital Charity during the conduct of the study. Gill Gyte reports personal fees from Cochrane Pregnancy and Childbirth Group outside the submitted work. Jim Thornton reports membership of HTA Clinical Trials Board 2010–15, HTA and EME Editorial Board 2016–18, Rapid Trials and Add on Studies 2012, CPR Decision-Making Committee. David Field reports HTA Commissioning Board 2013–2018, HTA and Efficacy and Mechanism Evaluation (EME) Editorial Board 2012–18. William McGuire reports grants from NIHR during the conduct of the study, and membership of the HTA Commissioning Board 2013–present and the HTA and EME Editorial Board 2012–present.

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Duley et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. 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.

2019 Queen’s Printer and Controller of HMSO

Background

Being born too early (preterm birth) has a major impact on survival and quality of life for the child, on psychosocial and emotional stress on the family, and on costs for health services and society. 1–3 Mortality is highest for infants born very preterm, before 32 weeks’ gestation. In the UK, infant mortality (deaths in the first year of life) for babies born very preterm is 144 deaths per 1000 live births, compared with 1.8 deaths per 1000 live births at term. 4 Although only 1.4% of live births in the UK are very preterm, these babies account for 51% of infant deaths. 4

The costs of neonatal care for infants born very preterm are high. For those born before 28 weeks’ gestation, duration of hospital stay is 85 times that for term births and hospital inpatient costs are £15,000 higher. For those born at 28 to 31 weeks, duration of hospital stay is 16 times that for term births and hospital inpatient costs are £12,000 higher. 5 Morbidity among children born very preterm who survive is also higher than those born at term. 3 Of very preterm infants who survive, around 5% develop cerebral palsy and those without severe disability have a twofold or greater increased risk for developmental, cognitive and behavioural difficulties. 1,2 These impairments may persist into adolescence and early adulthood. 6,7 Even modest improvements in outcome would be of substantial benefit to the children, their families and the health services.

Aims and objectives

The aims of this programme were to improve the quality of immediate care at very preterm birth, enhance family-centred care, and improve outcome for infants born very premature and their families.

Specific objectives were to:

• Develop a James Lind Alliance (JLA) Priority Setting Partnership (PSP) between service users and clinicians to identify and prioritise treatment uncertainties relevant to preterm birth (work package 1; Figure 1) and to:

  • identify and prioritise research gaps for preterm birth, using the methods developed by the JLA

  • describe how service users and clinicians interact when making collective decisions about research priorities, and how they communicate when deciding research priorities together.

• Develop strategies for providing initial neonatal care at birth beside the mother, rather than away from the mother, for preterm or sick babies by:

  • conducting a survey of current practice for initial neonatal care with the cord intact at NHS hospitals (work package 3.1)

  • describing parents’ experiences and views of care at very preterm birth, and developing a questionnaire tool to assess their views of care at very preterm birth (work package 3.2)

  • improving understanding of the physiology of transition from fetal to neonatal circulation by measuring umbilical flow at preterm birth and assessing how it varies with gestation (work package 3.3)

  • conducting systematic reviews and overviews (umbrella review) to assess the evidence for delivery room transitional assessment and support, and to identify research gaps (work package 2.1)

  • developing a mobile Bedside Assessment, Stabilisation and Initial Circulatory Support (BASICS) trolley to support providing newborn life support beside the mother, and with the cord intact (work package 3.4)

  • describing parents’ and clinicians’ experiences and views of neonatal care at birth beside the mother, and of the BASICS trolley (work package 3.5).

• Generate information that will inform the design of a high-quality large multicentre UK trial comparing a policy for very preterm births of deferred cord clamping and initial neonatal care with umbilical cord intact, against immediate clamping and initial neonatal care after cord clamping, by:

  • conducting a narrative systematic review (framework synthesis) to identify the ethical challenges and their potential solutions in consent to recruitment of preterm or sick infants to clinical trials (work package 2.2)

  • conducting a pilot randomised trial comparing deferred cord clamping and initial neonatal care with the cord intact, against immediate clamping and initial neonatal care after cord clamping for very preterm births, including follow-up of the women for 1 year and until 2 years of age (corrected for gestation at birth) for the children (work package 4)

  • establishing a collaborative group to conduct a prospective meta-analysis of trials evaluating alternative strategies to influence placental transfusion at preterm birth, and developing the protocol for this analysis (work package 5).

FIGURE 1.

Overview of work packages. WP, work package.

The prioritisation process ran in parallel to other work packages. It was added to the programme in response to feedback from the funding board at stage 1 of the grant application process. Hence, other work packages were not dependent on its outcomes. Although cord clamping did emerge as a top priority, we had already identified this as a priority via an informal process.

Priority setting for future preterm birth research

In the past, the health-care research agenda was determined primarily by researchers, and the processes for priority setting in research lacked transparency. 8 Often, research does not address the questions about treatments that are of greatest importance to patients, their carers and practising clinicians. 9,10 Many research questions have not been investigated and for many more the existing evidence is incomplete. The JLA (www.lindalliance.org/; accessed 24 January 2019) has developed methods for bringing patients and clinicians together to establish PSPs that then identify and prioritise ‘treatment uncertainties’ to inform publicly funded research. 11 This approach has been used successfully for a wide range of topics including asthma,12 urinary incontinence,13 vitiligo,14 prostate cancer15,16 and schizophrenia. 17 For urinary incontinence, of the top 10 priorities, five came from clinicians, four from patients and one from researchers. 13

When uncertainty about the effects of treatments relates to preterm birth, it seems particularly pertinent that research should address the most important priorities, and the most pressing needs, of these vulnerable children and their families and clinicians. Failure to identify and prioritise these uncertainties may result in suffering and death. 18 Examples in perinatal care where this failure to identify and prioritise important uncertainties has happened include the use of caffeine, which was shown to reduce the risk of cerebral palsy and developmental delay, and so was widely taken up in clinical practice, but only 30 years after being suggested for prevention of apnoea in premature babies. 19,20 In addition, the use of magnesium sulphate, which, after 60 years of controversy, was finally shown to be better for women with eclampsia than either diazepam or phenytoin. 21–23 Although the importance of women and their clinicians contributing to the perinatal research agenda is well established,8 we proposed the first PSP in the perinatal field.

To establish a PSP required bringing together representatives of women and their families and of clinicians and other health-care workers. Understanding of how people interact and make collective decisions in groups or committees with members from diverse organisations comes from health research, and from social psychology and business administration. 24 Larger groups may increase membership diversity,25 although this may be offset by a reduction in reliability of decision-making. The chairperson is crucial for establishing inclusiveness, openness and trust in the discussion. 25–28 If time is short, less knowledge is shared and evaluated and then decisions result more from judgement based on prior preferences than problem-solving,29 which may mean that they are more influenced by an individual’s status within the group. 25

When we planned the Preterm Birth PSP, evidence about how service users and clinicians make joint decisions on research priorities was lacking. 24 Improved understanding of how such partnerships work (e.g. if expertise based on qualifications, experience or problem-solving skills influences decisions30 or if the way arguments are framed changes as consensus develops) may offer insight into how the process could be improved.

The Preterm Birth PSP published a list of the top 15 research priorities,31 which is being used by the National Institute for Health Research (NIHR) and by researchers in planning new research. Study of the groups’ working processes has also led to recommendations about how decision-making might be improved.

Parents’ views and experiences of very preterm birth

Very preterm birth and subsequent hospitalisation of the baby can be an extremely distressing time for parents. 1,32 There has been little research into parents’ experiences at the time of very preterm birth, or their satisfaction with their care during the birth. In addition, previous work has often failed to include fathers. Women’s views and experiences during labour and childbirth are increasingly important to health-care providers and policy-makers,33 and may have an impact on subsequent health and well-being for the woman and her baby. 34,35

To ensure a strong parent focus throughout our programme, we planned qualitative interviews to explore parents’ experiences of very preterm birth and their first moments with their baby. 36–38 Regular multidisciplinary project meetings to plan these interviews, and to discuss the emerging themes and verbatim transcripts, were immensely valuable for subsequent work packages. These were particularly valuable for developing strategies for neonatal care beside the mother, and for planning the pilot trial. Our experience in conducting these interviews also contributed to further qualitative work exploring the views of both parents and clinicians of neonatal care at birth beside the mother,39,40 and of the new two-stage consent pathway in our pilot trial. 41,42

A number of instruments have been developed to assess women’s satisfaction with care during childbirth. When planning our programme we were not aware that any of these were designed to be used following very preterm birth, which is usually a different experience from giving birth to a healthy term baby. To confirm this we conducted a systematic review of available measures to assess parents’ satisfaction with care during labour and birth. 43 Having demonstrated there was not a suitable tool for very preterm birth, we developed a 20-item questionnaire called Preterm Birth experience and Satisfaction Scale (P-BESS) for use in this situation. 44 We used this questionnaire for the follow-up of women recruited to our pilot trial. It has been translated into Spanish and Portuguese.

Placental transfusion and neonatal transition

At birth, if the umbilical cord is not clamped then blood flow between the baby and the placenta may continue for several minutes. 45–48 This umbilical flow is part of the physiological transition from fetal to neonatal circulation, which, for very preterm infants, may improve resilience during this transition. 49–51 ‘Placental transfusion’ refers to the net transfer of blood to the baby between birth and cord clamping.

Cord clamping before umbilical flow ceases may restrict neonatal blood volume and red cell mass, and/or disrupt transition from fetal to neonatal circulation. For term births, umbilical flow usually continues for 2 minutes, but may continue for over 5 minutes. 46,48 The mean volume of placental transfusion at term is 100 ml, which is 29 ml/kg birthweight and 36% of neonatal blood volume at birth. 48 For preterm births, umbilical flow may continue for longer than for term births52 and is incomplete if the cord is clamped in 30–90 seconds. 53 This corresponds with development during gestation; at term, two-thirds of the fetoplacental circulation is in the infant, whereas for those born below 30 weeks’ gestation, a greater proportion of the fetoplacental circulation is in the placenta. 45 In addition, the preterm umbilical vein is smaller than at term, and uterine contraction less efficient; therefore, transition from the fetal to the neonatal circulation may be slower. Cord milking or ‘stripping’ seems to be an attractive option for preterm births, as potentially it increases neonatal blood volume without the need to defer cord clamping. 54 However, cord milking over-rides the infant’s physiological control of its own blood volume and blood pressure, and it interrupts transition to the neonatal circulation. 55

To improve understanding of the physiology of placental transfusion and assess when might be the best time to clamp the cord for preterm births, we measured umbilical flow at preterm birth. Although this proved to be more challenging than we had anticipated, the resulting data helped to inform the decision to wait at least 2 minutes before clamping the cord in the pilot trial.

Neonatal stabilisation and resuscitation at birth beside the mother

In the UK, about one-third of all newborn babies are attended at birth by neonatal resuscitation staff. For most, all that happens is an assessment, stimulation, thermal care and simple airway management. However, around 15% of these babies receive active stabilisation and/or resuscitation at birth, such as mask ventilation, intubation, cardiac massage or drug administration.

Transition to the neonatal circulation begins at birth when pulmonary vascular resistance falls as the lungs expand and fill with air, pulmonary blood flow increases and ductal blood flow declines as peripheral resistance falls. 56 Stabilisation at very preterm birth aims to assist this transition, and recommendations for newborn life support for very preterm infants are to prioritise establishment of respiration and a resting lung volume. 57 Traditionally, this was facilitated by immediate cord clamping, allowing the baby to be transferred quickly away from the mother for interventions such as airway opening manoeuvres, continuous positive airways pressure (CPAP) and tracheal intubation with or without prophylactic surfactant administration.

Providing initial neonatal stabilisation and resuscitation for very preterm or sick infants beside the mother would potentially allow newborn life support to be provided with the cord intact, facilitating a longer period before cord clamping than was possible at the time when we were planning this programme. It would also allow the woman and her partner to share the first moments of their child’s life, a more family-centred approach to care at birth. 58,59 Family-centred care in neonatal units, with improved communication and involvement of parents in their baby’s care, appears to benefit babies, is welcomed by parents60 and has been prioritised by the NHS. 61 Providing initial neonatal care beside the mother also has parallels with family presence during resuscitation of adults and children, an approach that is preferred by families and clinicians, and that appears to be beneficial. 62–65 Should the baby subsequently die, this time spent with the parents may prove to be an important factor in their experience of the life of their baby, and could potentially be of benefit in bereavement.

At the start of our programme, information about neonatal care beside the mother within the UK was anecdotal. 53,66 Therefore, we began by describing current practice at that time for care at birth67,68 and by assessing the evidence from randomised trials for delivery room neonatal interventions. We then developed and piloted strategies for providing initial neonatal care beside the mother, and assessed whether or not this was acceptable to parents and clinicians. This included both developing a new mobile trolley designed specifically for this purpose69 and adapting the existing equipment. 70 Our work demonstrating that providing newborn life support beside the mother is valued by parents and clinicians,39,40 and that care with the cord intact is feasible, contributed to the success of the Cord pilot trial. 71 Providing neonatal care beside the mother, with the cord intact, compared with after clamping and cutting the cord, has growing interest nationally and internationally. 72,73

Ethics issues in recruitment of preterm or sick infants to perinatal trials

Recruitment of preterm or sick infants to clinical trials requires approaching parents at a particularly difficult time, often with a tight time scale for making a decision. This raises challenges for obtaining informed consent to such research, especially issues regarding competence for consent, understanding of complex issues, insufficient time for parents to consider participation, and voluntariness if parents have a sense of obligation or feeling of debt to the clinician-researcher who is caring for their child. 74 On the other hand, if the problem of consent is not successfully addressed, this risks becoming an ‘orphan’ area of research. That is, if ethically permissible research cannot be designed, then this area of medical research will be abandoned.

Earlier work has explored these difficulties, specifically in the neonatal context. 74 Discussion of both the nature and the importance of consent, as well as empirical work on methods of obtaining consent and parental experience of those methods, has continued. 75–77 Hence, when we planned our programme it seemed timely to review understanding of these ethics issues in the light of this expanding literature, and of both the changing regulation for clinical trials and the increased public expectations of the conduct of research. Our overall aim was to identify a way of conducting ethical neonatal research in those circumstances where obtaining valid consent from parents has proved to be a significant challenge. The goal was to identify both the challenges to an ethically defensible consent process and their potential solutions. 78,79 This led to the development of a two-stage pathway for consent used and evaluated in the Cord pilot trial. This pathway accounted for almost one-third of recruitment to the trial, and was viewed positively by both parents and clinicians. 41,42 It was rapidly included in updated guidance for intrapartum research from the Royal College of Obstetricians and Gynaecologists. 80,81 The same two-stage approach has also been adapted for use in an acute stroke trial, for which randomisation is within 8 hours of the stroke. 82

Systematic review of timing of cord clamping at very preterm birth

Thirty years ago, it was first suggested that immediate cord clamping for preterm babies might increase the risk of intraventricular haemorrhage (IVH). 83 Postulated mechanisms for this increase were hypovolaemia or increased fluctuation in blood pressure during the abrupt transition from fetal to neonatal circulation.

At the time that we planned this programme, the Cochrane review of timing of cord clamping and other strategies for influencing placental transfusion at preterm birth included 10 trials, with 454 mother–infant pairs largely recruited before 33 weeks’ gestation. 84,85 In these trials, deferred cord clamping ranged from 31 to 120 seconds and immediate cord clamping ranged from 5 to 20 seconds. Many outcomes were reported by only a few studies, with potential for reporting bias. Immediate clamping was associated with an increased risk of transfusion for anaemia [relative risk (RR) 1.57, 95% confidence interval (CI) 1.14 to 2.16; four trials, 183 infants] or hypotension (RR 1.94, 95% CI 1.06 to 3.54; 3 trials, 90 infants) compared with deferred clamping. The risk of IVH on ultrasound scan was also higher for babies allocated immediate clamping (RR 1.90, 95% CI 1.27 to 2.84; seven trials, 329 infants), but for severe IVH (grade 3 or 4), a more reliable predictor of long-term outcome, there was no clear difference (RR 1.17, 95% CI 0.27 to 5.02; five trials, 269 infants). There was no clear difference between the groups in temperature on admission to a neonatal unit, but only three trials (143 infants) reported this outcome. One study had reported follow-up at a median age of 7 months for 58 out of 67 surviving infants, with no clear differences between the groups. 86

Alternative policies for timing of cord clamping at very preterm birth

In previous trials for deferred clamping, the decision about when to clamp the cord was usually a balance between allowing some umbilical flow and what was perceived as an acceptable delay in transferring the baby to the neonatal team. As standard practice was for the neonatal team to be located either at the side of the room or in a room nearby, this necessitated early clamping and cutting of the cord, particularly for infants requiring stabilisation or resuscitation at birth. Therefore, providing initial neonatal care at birth with the cord intact would make it feasible to defer cord clamping for longer than had previously been possible, including for high-risk infants who have the most potential for benefit.

Our programme to improve the quality of care and outcome at very preterm birth focused on care at the time of birth. We reviewed evidence from systematic reviews 36–38,67,68,70,78,79,87 relevant to delivery room neonatal care, surveyed current practice, described parent experiences, measured umbilical blood flow, developed strategies for providing newborn life support beside the mother, and reviewed ethics issues in recruitment of preterm and sick babies to clinical trials. These work packages all contributed to the design and conduct of the Cord pilot trial, a pilot randomised trial to assess the feasibility of conducting a large UK trial comparing alternative strategies for cord clamping. To provide adequate power for long-term follow-up of the children and inform the design of a large definitive trial, we planned a prospective meta-analysis of similar studies.

This focus on the time of birth and cord clamping was identified as a priority by informal discussion with parent representatives,88 and parent representatives worked in partnership with clinicians and researchers to develop the application and conduct the research. The topic had also been identified as a research priority by researchers,58,84,85,89 obstetricians,90 midwives,90 neonatologists (Duley L, Farrar D, McGuire W, Oddie S. Survey of the Extended Neonatal Network to Assess Views on Timing of Cord Clamping and Placental Transfusion: Report Prepared for the Extended Neonatal Network. 2009. Unpublished), the National Institute for Care Excellence (NICE),91,92 and the Royal College of Obstetricians and Gynaecologists. 59 Therefore, it was unsurprising that this uncertainty was included in the list of top priorities from our JLA Preterm Birth PSP. 31

Changes from the original programme plan

The phrase ‘preterm birth’ is broad. Our programme had a specific focus on care around the time of birth, and our intention was that this would be reflected in the scope of our JLA PSP (work package 1). However, from the initial scoping meeting, it was clear that both service users and clinicians had a different view and the discussion ranged from risk factors to prognosis and the grandparent perspective. This continued at the first meeting of the partnership steering group. As the partnership was between service users and clinicians, the researchers agreed that their views should determine scope. The challenge of too wide a scope led to delays in progressing the priority setting, and the process progressed only once it was agreed that this should be narrowed. This delay and the resource demands of the priority setting process meant that it was not possible to conduct the planned work on outcomes.

Although our sampling frame for the qualitative interviews with parents (work package 3.2) included one hospital where deferred cord clamping was in routine practice for very preterm birth, no parents interviewed had experienced neonatal care beside the mother. It was, therefore, not possible at that time to assess their experiences of this type of care. Instead, we did this by assessing parent and clinician experiences of neonatal care beside the mother using semistructured interviews, rather than the planned focus groups (work package 3.5). This was in order to gain more detailed accounts of the events and experiences than would be available in a focus group. In addition, because in focus groups individuals might be reluctant to disagree with the dominant view, you get a socially constructed view rather than people’s individual views and experiences.

Measuring placental transfusion at preterm birth proved to be far more challenging than our earlier work at term birth. 48 This was primarily because of the unpredictability of preterm birth, which made it difficult for the research team to be present and with the equipment set up in time for the birth. As we secured funding to assess feasibility of a similar trial in low and middle income countries, this study was replicated successfully in India. 47

The Cord pilot trial was conducted as planned. Added value of this multidisciplinary programme led to several additional elements. First, innovative methods for consent to participate in emergency perinatal trials was identified as a research gap by the ethics framework review. This led to the development of a two-stage oral assent pathway used in the trial when birth was imminent, which boosted recruitment. Second, to evaluate this pathway we used our experience of semistructured interviews with parents and clinicians (work packages 3.2 and 3.5) and of the framework analysis of ethics issues (work package 2.2) to design and conduct qualitative interviews with women and clinicians who had experience of consent in the Cord pilot trial. Finally, excellent recruitment contributed to the Trial Steering Committee (TSC) assessment that feasibility had been demonstrated and that we should seek funding to progress to the definitive trial. The TSC also recommended that recruitment continue while funding was sought, to avoid ‘stop/start’. Hence, recruitment was extended and only closed when the funding application was rejected. To maximise the value of this additional recruitment, follow-up of both women and children was also extended and data for all those randomised are presented here.

For the prospective meta-analysis, the first cycle of analysis had been planned within the programme. However, as data from the two largest studies (Cord pilot trial93 and the Australian Placental Transfusion Study94) were not available within the time scale of the programme, this analysis has been postponed. In addition, the collaborative group of triallists has agreed to expand the protocol to a retrospective individual participant meta-analysis with nested prospective meta-analysis.

Research aims

Our aims were to:

• identify and prioritise research gaps relevant to preterm birth that are most important to people affected by preterm birth and health-care practitioners in the UK

• describe how service users and clinicians interact when making collective decisions about research priorities, and how they communicate when deciding research priorities together.

Identifying and prioritising research gaps relevant to preterm birth

Prioritisation process

Using methods established by the JLA,95 we first identified unanswered questions about the prevention and treatment of preterm birth from people affected by preterm birth, clinicians and researchers. Then we prioritised those questions that people affected by preterm birth and clinicians agreed are the most important (Figure 2).

FIGURE 2.

Flow chart of the JLA Preterm Birth PSP. HP, health professional; PaPB, people affected by preterm birth.

Initiation of the partnership

The Preterm Birth PSP was initiated in November 2011, following an earlier introductory meeting of potential stakeholders. The 29 participating organisations were asked to complete a declaration of interests, and a steering group was convened that included members from nine of these organisations. At the introductory workshop it was clear that many participants felt that the scope of the partnership should be wider than was initially envisaged, for example including uncertainties about the causes of preterm birth, the prognosis following being born preterm, and interventions long before birth.

Consultation

As widening the scope too far risked making the prioritisation unachievable, the steering group restricted the scope to uncertainties about treatments and to interventions during pregnancy and around the time of birth or shortly afterwards (taken up to the time of hospital discharge for the baby after birth).

Research questions were gathered from people affected by preterm birth, clinicians and researchers using methods developed by the JLA. 95 This included a survey completed online by 349 people, and in paper format by 37 women attending specialist preterm birth antenatal clinics at two tertiary level hospitals and parents visiting two neonatal intensive care units. These 386 responses contained 593 potential research questions. In addition, 540 potentially relevant research questions were identified from systematic reviews of existing research and from national UK clinical guidelines. 96–107

Collation

All questions were screened to identify those sufficiently similar to either merge or group into broader questions and to remove any that were out of scope, unclear or being answered by a subsequent or in-progress randomised trial. This left 70 unanswered questions from the survey, 28 from systematic reviews and 24 from clinical guidelines remaining in the process. Of these 122 questions, 18 overlapped with other questions and were merged to give a final ‘long list’ of 104 unanswered research questions.

Prioritisation

This was a two-stage process. First, the long list of 104 questions was sent out for voting on the top 10 questions, online and in paper format, using a modified Delphi survey. The steering group reviewed ranking by the 507 people who voted, overall and by stakeholder group. They removed remaining overlap or repetition between questions, before agreeing the shortlist of 30 questions to go forward to the second stage: a prioritisation workshop. This workshop had 34 participants, including representatives from people affected by preterm birth and clinician organisations as well as parents of babies born preterm and adults who were born preterm. At the workshop, nominal group technique95 was used to achieve consensus on ranking the 30 questions.

Describing how service users and health-care professionals interact when making collective decisions about research priorities

Analysis

Transcribed data were coded using an analytical framework based on the Elaboration Likelihood Model111–113 using NVivo 10 (QSR International, Warrington, UK) software. We coded discussion as using either a central route (rational argument with evidence) for persuasion or a peripheral route (relying on emotional responses to cues such as authority, commitment, consistency, liking, reciprocation, social proof or scarcity). For example, prioritising a treatment because the speaker ‘wanted to know if there would be better outcomes’ was coded as ‘central’ because of its rational approach. Asserting ‘I think this topic has to be number one because that is sort of nice’ was coded as ‘peripheral’ with a ‘liking’ cue.

To investigate decision-making based on values, transcripts were coded into three types of discussion:114 informational (facilitator encourages participants to speak, defers controversy and lets participants know their ideas will not be evaluated), problematical (participants consider the information and/or values needed to address the issue intelligently) and reflexive (participants discuss their own discussion to learn from the process).

To understand the process of consensus development, the JLA approach was compared with the ‘Group Development Model’115 (Figure 3), which argues that every group experiences these five stages before becoming a self-reliant unit. At each stage, the dynamics of the group change from periods of inefficiency and uneasiness through to a period of high performance.

FIGURE 3.

Group Development Model versus JLA stages of partnership working.

Research aims

Our aims were to identify Cochrane reviews of delivery room transition support interventions, appraise review quality and identify important gaps in the evidence.

Methods for data collection

We undertook a systematic overview (umbrella review) using the standard methods of the Cochrane Collaboration and the Centre for Reviews and Dissemination. 119,120 We registered the overview on PROSPERO (registration number CRD42012003038). We searched the Cochrane Database of Systematic Reviews121 (Issue 6, 2015) for reviews evaluating any intervention for delivery room support of newborn infants. We excluded reviews of interventions that are more usually or feasibly delivered following admission to the neonatal unit, and those administered to all infants as part of routine practice.

Analysis

For each review, two reviewers independently extracted information on review quality characteristics, and on the participants, treatment and control interventions, and outcomes. Review quality was assessed using the 11-item AMSTAR (A MeaSurement Tool to Assess systematic reviews) tool. 122,123

Key findings

Eighteen Cochrane reviews were identified. Broadly, these reviews assessed delivery room interventions for airway management, respiratory or circulatory support (four reviews); surfactant replacement therapy for preterm infants with or at risk of respiratory distress syndrome (eight reviews); supplemental oxygen or other drugs for infants compromised at birth (five reviews); and strategies for influencing placental transfusion (one review). The overall quality of reviews was good, but methodological quality of the included trials varied greatly. Four reviews had no included trials. The most commonly prespecified primary outcomes were death, incidence of chronic lung disease and neurodisability. Two reviews prespecified surrogate outcomes, such as physiological measures, rather than clinically important primary outcomes. There are few data on long-term neurodevelopmental outcomes.

Of the four reviews that assessed interventions to support the infant airway and breathing, two did not find any eligible trials and one included only a single small trial. 124–126 The fourth review assessed delivery room airway support for infants at risk of meconium aspiration (four trials, 2884 participants) and concluded that there is no evidence that routine endotracheal intubation reduces mortality or morbidity in vigorous term babies with meconium staining compared with standard resuscitation. 127

The strongest evidence supported type and timing of surfactant administration for preterm infants. Two reviews, originally published in 1997, provided strong evidence that for very preterm infants, surfactant replacement reduced the risk of death by about 40%. 128,129 A related review concludes that natural surfactant is more effective than synthetic surfactant for reducing the risk of death. 130 These reviews are now regarded as ‘complete’, as further trials would be unlikely to change their conclusions. Subsequent reviews found evidence that early surfactant administration with brief ventilation reduces the need for mechanical ventilation and associated morbidity, but that prophylactic (delivery room) administration is not more effective than delayed selective administration when infants have prophylactic nasal continuous positive airway pressure support. 131,132 Uncertainty remains about the comparative effects of newer synthetic surfactants that contain surfactant protein mimics, and of novel non-invasive routes for administering these. 133,134

Five reviews assessed supplemental oxygen or other drugs for infants compromised at birth. One compared using air rather than 100% oxygen for newborn resuscitation at birth. 135 Five trials were identified (three were quasi-randomised), but the review concluded that there was insufficient evidence to support a recommendation for using either intervention. Three reviews assessed other drug interventions. 136–138 Of these, two assessed adrenaline or sodium bicarbonate during resuscitation and found insufficient evidence for reliable conclusions. 136,137 The third review assessed naloxone for infants exposed in utero to opiates and identified nine trials, but none assessed clinically important outcomes. 138 The fifth review examined interventions to prevent hypothermia in newborn very preterm infants,139 and it concluded that various measures, including plastic wraps or bags and warming mattresses, reduce the risk of delivery room hypothermia, but found insufficient data to assess effects on infant morbidity and mortality.

One review assessed timing of cord clamping and other strategies for influencing placental transfusion at preterm birth. 140 This review found evidence that deferring cord clamping for 30–120 seconds, rather than clamping before 30 seconds, probably reduced the need for blood transfusion and possibly reduced the risk of IVH. Data from 13 out of the 15 included trials did not identify a statistically significant effect on risk of death. Long-term neurodevelopmental outcomes were not reported.

Implications for research

Some reviews identified key evidence gaps. These were mainly related to pharmacological intervention for transition support or resuscitation of newborn infants, and the effectiveness of new, less invasive forms of airway management (and related issues regarding surfactant delivery).

Research aims

We aimed to synthesise:

• Observational and qualitative studies that explored the process of recruitment and consent, and reported parents’ and clinicians’ views and experiences.

• Analytical (or philosophical) studies that have examined the pertinent ethical questions. These concern the validity of consent, the proper understanding of the parental role in giving or withholding consent, varied possible methods of seeking consent, the best interests of those involved, issues of risk, and the parallels between consent processes in relevant research and clinical contexts.

The goal was to identify both the challenges to an ethically defensible consent process and their potential solutions. Ideally, these solutions would include strategies that we could implement to improve the consent process in the Cord pilot trial.

Methods for data collection

The methods used for this review conformed to those set out for a framework synthesis. 141 The first stage was the development of a tentative initial conceptual framework based on prior knowledge of the existing philosophical literature on informed consent, including in neonatal research. 142,143 This initial framework informed the criteria for including studies and it suggested terms for the literature searches. After refining the framework in the light of the literature from the first searches, the second searches were developed. All searches were updated (for examples of search strategies, see Appendices 6 and 7).

Analysis

Analytical studies

For the analytical studies, discussion of the initial themes that were identified confirmed the conceptual framework. Further searches were informed by this framework and by bioethics review methods studies. 144–146 We screened abstracts of papers from identified citations for inclusion, using the five criteria from the confirmed framework:

1. consent, participation or recruitment for neonatal research (relevant to clinical trials)

2. parental decision-making for treatment of, or research with, sick or preterm neonates

3. parental decision-making for birth and/or labour

4. methodology in emergency/urgent neonatal research

5. alternative ways of gaining consent for neonatal research.

Overall, we ‘included’ 49 empirical papers and 30 analytical papers (Figure 4). All studies met the quality assessment criteria. For the empirical papers, we identified five themes: (1) attitudes of parents, (2) attitudes of clinicians, (3) validity of consent, (4) different consent processes and (5) miscellaneous topics. For the analytical papers we identified four themes: (1) ethical basis of parental informed consent for neonatal research, (2) validity of parental consent, (3) other options for gaining consent, and (4) risk and the double standard between consent for treatment and consent for research. We coded articles and tabulated them against these themes.

FIGURE 4.

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram for the framework synthesis. Derived from Wilman et al. 79 [© Wilman et al. , 2015. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated] and Megone et al. 78 [© The Authors, 2016. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated].

Derived from Wilman et al. 79 [© Wilman et al. , 2015. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated]Derived from Megone et al. 78 [© The Authors, 2016. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated]

Key findings

Successes

This is the first review focusing on the ethics issues around consent to neonatal or perinatal research. Other reviews have focused on methods for increasing recruitment or improving how information is conveyed, but none has focused directly on the ethics issues. 61,147–151

Challenges

The key challenge lay in this project being a relatively novel departure for systematic review methods, as it involved reviewing systematically across the disciplinary boundaries between philosophy, social science and medicine. To address this, the multidisciplinary team included expertise in philosophy, clinical practice, social science, information science and advocacy for parents. For the philosophers, the systematic review methods were somewhat different from those that are conventionally adopted in work on ethics. The notion of a systematic review is uncommon, indeed almost unknown, as a method for conducting research in philosophy, both in philosophy as a whole and in philosophical medical ethics in particular. 144,152 On the other hand, for the social scientists and clinicians the standard philosophical means for resolving ethics problems were unconventional. As a result, the whole process of developing a systematic review in ethics through interdisciplinary collaboration has given rise to considerable reflection in its own right.

Implications for future research

We identified three important gaps relevant to consent for perinatal trials:

1. Studies on a new process for obtaining urgent or emergency consent in perinatal research where there is little time in which to seek consent to participation. This process involves seeking ‘assent’ to participation in the trial from parents, accepting that there is insufficient information for fully informed consent at that point, but then providing further information over time which allows the parents to give full consent (or an informed withdrawal).

2. Studies on fetal–maternal research, that is, trials in which both the mother and the fetus or neonate are participants and thus in which both fetal/neonatal interests and maternal interests are in question.

3. Empirical studies that report the views of bereaved parents on the consent process.

This review provided part of the rationale and justification for including a two-stage consent pathway in our Cord pilot trial (see Work Package 4), as part of the feasibility assessment for a large trial. Participation by parent representatives (members of NCT and Bliss) in the ethics review, in the development of the pilot trial protocol and in the programme steering group contributed to robust discussion about development of this pathway. They focused on how it should be presented to women and their partners and how it should be evaluated. Endorsement of the trial protocol from these organisations was important in securing ethics committee approval and in communication with parents. Hence, this programme has contributed to filling the first two research gaps identified in our review.

At the time of our research, empirical studies had almost completely excluded views of parents whose sick child participated in research and died. Since then, however, the Bereavement and RAndomised ControlLEd Trials (BRACELET) study addressing this gap has been published. 153

Research aims

This study aimed to describe current practice for providing neonatal care in the delivery room, and for providing neonatal care at birth beside the mother.

Methods for data collection

We surveyed all neonatal units in the UK using a short online questionnaire to ask about neonatal care in the delivery room at very preterm births. Units reporting they were using deferred cord clamping or cord milking were selected, using purposive sampling, to participate in semistructured interviews. Two researchers interviewed experienced practitioners either in person or by telephone. Interviews were recorded and transcribed.

Analysis

Overall, 197 survey responses were received from 199 hospitals, of which 186 (94%) were fully completed. Seven units participated in the semistructured interviews: five tertiary hospitals, and one large and one small district hospital. Overall, 33 staff members were interviewed: seven midwives, seven neonatologists, two paediatricians, six neonatal nurses, seven obstetricians and four managers. Constant comparative analysis identified five core themes: (1) variability in guidelines and practice, (2) assessing eligibility, (3) competing priorities, (4) anxiety about timing and (5) persisting uncertainty.

Key findings

There was variation in delivery room stabilisation practice for infants born very premature, and in tertiary units the care provided was more consistent with current international guidance than in non-tertiary units. 68 For example, tertiary units administered more surfactant in the delivery room (93% vs. 78%), were more likely to provide CPAP (77% vs. 50%) or positive end-expiratory pressure in the delivery room (91% vs. 69%), and were more likely to start resuscitation in air or blended oxygen (91% vs. 78%) than non-tertiary units. Routine out of hours consultant attendance at very preterm birth was also more common in tertiary units (82% vs. 55%).

Our interviews suggested that variation in how deferred cord clamping and other strategies to influence placental transfusion at very preterm birth were being implemented, both between units and within the same unit. 67 For example, there was variation in when the cord was clamped, position of the baby with the cord intact, timing of the prophylactic uterotonic drug, and which babies were seen as eligible for this intervention. Deferring cord clamping was felt to require multidisciplinary agreement because of the perceived conflict between waiting to allow placental transfusion and a wish to ‘get on’ with keeping the baby warm and providing newborn life support interventions. In some units, whether or not this happened depended on which staff members were present. No unit was providing neonatal care with the cord intact, and there was staff anxiety associated with this practice.

Implementation of deferred cord clamping, or cord milking, appeared most likely to be successful if there was strong local multidisciplinary support, with agreement on a single technique and the eligibility criteria. Clinical leadership and training in the practical techniques, combined with audit, also appeared to be helpful.

Successes

Use of telephone reminders for the online survey helped us achieve a high response rate. A strength of our interviews was the inclusion of all the relevant professional groups, and that the prior survey allowed us to identify units with relevant experience.

Challenges

Both the survey and the interviews relied on reported practice rather than direct observation. The interviews were based on a small sample and, as at that time deferred cord clamping was not widely practised, may not be representative of all UK units.

One of the reasons this study was included in the programme was to identify UK units where neonatal care was being provided with the cord intact, so that we could learn from their experience. No such unit was identified, although introducing neonatal care with the cord intact was being discussed at one unit.

Implications for future research

Since this work was completed,154 both the timing of cord clamping for very preterm births and providing neonatal care with the cord intact have become increasingly topical. This has been fuelled by new research, including the work conducted within this programme. A change in culture is leading to a shift away from the previous practice of immediate cord clamping at very preterm birth. In the UK, waiting at least 30 seconds, but no longer than 3 minutes, before clamping the cord of preterm babies is now recommended if the mother and baby are stable. 155 Hence it would be timely to repeat a survey of practice for timing of cord clamping and immediate neonatal care.

Research aims

The aims of this work package were to:

• Explore parents’ experiences of very preterm birth, and their first moments with their baby, through three separate analyses to explore:

  1. mothers’ and fathers’ initial experiences of the birth of their very preterm baby

  2. parents’ experiences and satisfaction with their care during very preterm birth, and to identify the domains associated with positive and negative experiences of care

  3. parents’ views and experiences of the care for their very premature baby on a neonatal intensive care unit (NICU).

• Systematically review available measures of parents’ satisfaction with care during labour and giving birth.

• Develop a questionnaire to measure parents’ satisfaction with care during very preterm birth (P-BESS).

Methods for data collection

Parents whose baby was born before 32 weeks’ gestation during a 6-month period at three NHS hospitals in the south of England were sent letters of invitation to participate in the qualitative interviews (see Appendices 10 and 11 for full details). Of 123 eligible parents, 39 (32%) agreed to be interviewed (32 mothers and 7 fathers). 36 Two babies died shortly after birth. Interviews contained 13 open-ended questions and lasted about 45 minutes. They were conducted by one psychologist and took place at the participants’ home, or in a private hospital room. Interviews were recorded and transcribed. Reporting complied with COREQ (Consolidated criteria for reporting qualitative research). 154

For the comparative review, studies were included if they reported use of a questionnaire that was a multi-item scale of satisfaction with care during labour and birth, and provided psychometric information (about questionnaire construction, reliability or validity) for the satisfaction measure. 43 To identify potentially eligible studies we used the search terms (Birth or Childbirth or Lab*r or Intrapart*) AND (Satisfaction or Perception or Evaluation) AND (Questionnaire or Measure* or Scale or Instrument). We searched Scopus, PsycArticles, PsycINFO, PubMed and Web of Science from inception to 30 July 2011, and checked reference lists in reports of included studies for additional studies. Duplicate citations were removed. Finally, Web of Knowledge and Scopus were searched for all reports that cited the final questionnaire measures; no new citations were identified. Data extraction was by two reviewers.

Initially we developed the questionnaire using data from the interviews and studies identified in the comparative review. We identified seven areas of satisfaction with care during preterm birth: (1) information and explanations, (2) emotional support, (3) encouragement and reassurance, (4) staff being confident and in control, (5) staff being calm in a crisis, (6) involvement of the partner and (7) birth environment. Thirty questions were included, both positively and negatively phrased, and responses scored on a Likert scale; we made minor changes following feedback from nine parent representatives. The questionnaire was then posted to parents of babies born very preterm during the previous 12 months at five tertiary care centres in England.

Analysis

We used inductive systematic thematic analysis to identify themes across interviews. Data were managed using NVivo software. For the systematic review, we assessed psychometric quality of each questionnaire using questionnaire construction (item generation, pilot study), reliability (internal consistency, test retest reliability) and validity (content, face, criterion and construct).

For the questionnaire, a factor analysis was conducted to explore whether questions could be combined into subscales that represent different aspects of satisfaction with care during very preterm birth. Three questions asked about partner’s involvement so, as they were not relevant for all women, they were excluded from the initial analysis. Hence, 27 questions entered into the factor analysis. The number of factors to be retained was determined using the scree plot and eigenvalues > 1. Questions that loaded on a factor at > 0.4 were considered significant and were retained. Questions that loaded on more than one factor ≥ 0.3 were removed and the analysis was rerun. To check whether questions and subscales for the women were applicable to partners, we conducted a confirmatory factor analysis.

Content validity is evident through the systematic series of steps taken when designing the questionnaire. Convergent validity was explored by examining the relationship between the total score (and associated subscales) with two questions assessing overall satisfaction with care during the birth, and reliability by looking at indicators of internal consistency.

Key findings

Comparative review of measures of parents’ satisfaction with care during labour and birth

Nine questionnaires measuring satisfaction with care during labour and birth were identified (Figure 5). For seven of these questionnaires, how the items were selected was described. Eight of these questionnaires had tests of internal consistency, but only one reported test–retest reliability. 43 At least one aspect of validity was reported for all questionnaires, but none reported criterion validity.

FIGURE 5.

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow chart for the comparative review. Reproduced from Sawyer et al. 43 © Sawyer et al. ; licensee BioMed Central Ltd 2013. This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reproduced from Sawyer et al. 43 © Sawyer et al. ; licensee BioMed Central Ltd 2013. This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Only two questionnaires assessed satisfaction with different aspects of care, as well as the perceived importance of these aspects of care. Three questionnaires were designed for particular types of births, two for operative births and one for uncomplicated vaginal birth. None of the nine questionnaires evaluated care for specific populations, such as parents of sick or preterm babies, or whose baby was stillborn. Parent experiences in these situations may be substantially different from giving birth to a healthy, term baby.

Successes

We achieved a good response to the invitation to be interviewed. The use of qualitative methods provides an in-depth insight into the experiences of parents who have had a very premature baby. The inclusion of fathers and bereaved parents also provides a valuable and unique perspective. Our data underline the importance of listening to women during preterm labour and of encouraging fathers to feel involved during the birth. Whenever possible, parents should be encouraged to visit the NICU before birth. If this is not possible, parents could be provided with a photograph of their baby in the neonatal unit before they visit.

Parents who are worried about touching their baby should be reassured and taught to recognise infant behaviour in response to touch.

Having identified only two existing questionnaire measures of satisfaction with care during labour and childbirth, both designed for term birth, we developed a new tool for use by both parents following very preterm birth. This tool is the first specific to preterm birth. The total score may be useful to compare satisfaction with care during very preterm birth across hospitals and differing practices, and individual aspects of care can be evaluated using the separate subscales. We used this tool to measure satisfaction with care in the Cord pilot trial (see Work Package 4), and it has been translated into Spanish and Portuguese.

Challenges

Although the response rate for our qualitative interviews was good for this type of study, we received responses from parents whose baby was born only from two of the three hospitals, reducing the generalisablity of our data. In addition, participants were mainly white, married women, which is not typical for very preterm birth. In common with other studies of satisfaction, parents may have been reluctant to criticise the professionals who took care of them and their preterm baby. This ‘halo effect’ may be even more pronounced for parents of premature babies, as the staff have been looking after their baby for many weeks. 44 Similarly, if women do not know what care during birth should be like they may just evaluate the status quo. 156,157 Our study used in-depth interviews by a researcher not associated with the hospital, which should have helped to pick up relevant negative experiences.

The response rate for development of the questionnaire was relatively low, although again this is a good response for studies of this kind. The sample size was relatively small for a factor analysis, which limits the validation process. In addition, the sample was not representative of all parents who have a very preterm birth, as it included largely white, highly educated, married/cohabiting women and their partners. Finally, as the same factor structure was not identified in partners as in women, only the total score is recommended for use with partners, which means that the individual factors of care cannot be explored for them.

Implications for future research

Further research is needed to replicate our findings about parent experiences at very preterm birth, and to explore whether or not there is variation for parents from different backgrounds. For the P-BESS, further studies are needed to test the refined instrument in a larger, more representative sample of parents. Fathers of preterm or sick babies are a difficult group to recruit into research. 158 Our work highlights the importance of including them in research studies to ensure that their perspective is represented.

Our experience in conducting semistructured interviews with parents following very preterm birth also contributed to the design and conduct of an evaluation of parent and clinician experiences of neonatal care beside the mother (see Work Package 3.5), and of women and clinicians’ experiences of the two-stage consent pathway in the Cord pilot trial (see Work Package 4).

Research aims

The aim of this work package was to measure the volume and the duration of placental transfusion for preterm births by weighing babies with the umbilical cord intact.

Methods for data collection

Recruitment was at three maternity units in England. Women likely to give birth to a live baby before 36 completed weeks’ gestation were offered participation. At each site, research staff undertook an initial training phase recruiting women giving birth at term. At birth, the baby was placed on the weighing platform and wrapped in warm towels with the umbilical cord intact. We used high-quality pharmacy scales (Excellence XS Precision Balance Model XS8001L, Mettler Toledo, Im Langacher, Switzerland), which calculated an average weight twice every second, with data stored in a linked computer. Before the birth, scales were zeroed to allow for towels, probes and any other equipment. To ensure that the baby was no higher than the level of the woman’s abdomen, the woman’s bed was raised or lowered if necessary. The baby was monitored throughout weighing using saturation monitors.

Analysis

Characteristics of the women, events during labour and mode of birth were described for all women. Placental transfusion was assessed by change in weight over time, as 1 ml of blood weighs 1.05 g. For each birth, two authors (JD and SO) independently assessed the weight change by inspecting the graphs of weight against time, using information on the time of cord clamping and when the baby was touched. Differences were resolved by discussion. Owing to the small numbers recruited, data are described only as summary statistics were inappropriate.

Key findings

Of 97 women approached, 33 gave consent, of whom six had their baby weighed. Reasons why the baby was not weighed for the remaining 27 with consent were: term birth (n = 10), research staff not available as out of working hours (n = 6), birth too rapid (n = 5), cord too short (n = = 2), consent withdrawn (n = 2), hardware problem (n = 1) and clinician decision (n = 1). For the six babies who were weighed, gestation was from 34⁺⁴ weeks to 36⁺⁵ weeks (Figure 6). Three were vaginal births and three were caesarean. The time of cord clamping ranged from 2 minutes to 3 minutes 57 seconds. For two babies, drying and applying probes led to a poor recording of weight for the first minute. An initial 10 seconds ‘hands-off’ period to give a baseline weight was therefore adopted. Weight change ranged from a 20-g decline to a 128-g increase. For one baby, weight change was continuing when the cord was clamped at 120 seconds; for another, weight change ceased at 2 minutes, and for the remaining four babies weight change appeared to continue for at least 3 minutes.

FIGURE 6.

Caesarean section at 35⁺⁶ weeks, weight gain estimated at 51 g.

This study is small; nevertheless, as with term births, there appears to be variation in the volume and duration of placental transfusion and, for some, net flow is to the placenta rather than the baby. For preterm births, umbilical flow may continue for more than 3 minutes. As placental transfusion may have a role in stabilising the cardiorespiratory circulation during transition from the fetal to neonatal circulation, the duration of time that the cord is left unclamped and umbilical flow continues may be as, or possibly more, relevant than the volume of any net flow. 49,159

Challenges

Major challenges were recruitment and, when appropriate women were recruited, having enough time for research staff to arrive and prepare the equipment, made even more challenging as births were often out of hours. Despite considerable effort at each site, these challenges were only partly overcome.

Implications for future research

Despite the small numbers, these data contributed to evidence supporting the decision to wait at least 2 minutes before clamping the cord for the intervention arm in the Cord pilot trial (see Work Package 4). These data also suggest that the effect of delayed cord clamping may not simply be to allow more blood to reach the baby from the placenta, but rather to allow the changes from an in utero placental circulation to an ex utero lung one, to occur more gradually. Improved understanding of the physiology of placental transfusion at very preterm birth would help identify the optimal strategy for cord clamping. Better methods for assessing placental transfusion are required. 46,72

Research aims

The aims of this work package were to develop a mobile trolley to provide newborn life support beside the mother, and with the umbilical cord intact, and to identify a commercial partner so that the trolley could be made more widely available.

Methods

A multidisciplinary team including clinical engineers, neonatologists, obstetricians, midwives and parent representatives met regularly to discuss and develop the design, using a prototype device based on an overbed hospital table with piped gasses, suction and a timer. Two simulated resuscitations provided assessment of functionality and practicality for both vaginal and caesarean births, following which the prototype was modified further.

A ‘second-generation’ prototype was developed in collaboration with Inditherm Medical (Rotherham, UK), a company specialising in the development and sale of neonatal equipment and devices. Following a third simulation to refine the design further, and certification with the Conformité Européenne (CE) logo, this was subsequently marketed as LifeStart™.

Key findings

The first prototype trolley was adapted from an overbed hospital table, which has a long base directly below the platform that slides under the bed, providing stability. However, modern operating tables and delivery beds have pedestal bases with little room beneath them. Therefore, the next prototype used a circular base, with larger wheels and enough weight to provide stability. This then led to the first commercially available trolley (Figure 7). The design struck a balance between trolley stability and reach of the platform. The platform is widest at the proximal end, allowing space for the baby’s shoulders and to access the head. The distal end is narrower, allowing the platform to get close to the mother. The baby is kept warm with an electric heated mattress produced by Inditherm; the company already produced the CosyTherm© (Inditherm Medical) mattress for neonatal cots, and this was reshaped to fit the trolley platform. To allow the platform height to be adjusted to be level with the mother, the central pillar adjusts up and down via an electronic mechanism. The trolley also has a battery driven digital timer with alerts at 1 minute and at 5 minutes, for use during resuscitation.

FIGURE 7.

First commercially available LifeStart trolley, October 2012. Reproduced from Thomas et al. 161 © 2014 The Authors, licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Reproduced from Thomas et al. 161 © 2014 The Authors, licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Medical gases for newborn resuscitation (air or oxygen) usually come from wall sockets; hoses connecting these sockets to the usual resuscitation equipment are short and tuck in against the wall. For a mobile trolley, long hoses connected to wall sockets create a trip hazard. Using small gas cylinders attached to the trolley was not possible at the time we developed the trolley, as the smallest size available for medical air was too large to attach to the trolley. Gasses are blended and the pressures are regulated using a Tom Thumb Infant Resuscitator^© (Viamed, Keighley, UK) and an oxygen blender (Inspiration Health Care Ltd, Leicestershire, UK) attached to the back of the trolley, along with a suction bottle (Oxylitre Ltd, Manchester, UK). Other equipment is attached, as required, using two medical equipment rails on the central pillar.

Successes

This is the first mobile neonatal resuscitation unit designed to facilitate newborn resuscitation beside the mother, and with the umbilical cord intact. Our initial prototype was entered into the Medical Futures Innovation Awards in 2011 (www.medicalfutures.co.uk/) and awarded ‘Best Redesign in Cardiovascular Medicine’. Further development has led to a successful commercial product, LifeStart, marketed in the UK and internationally, and cheaper than the usual equipment (around half to one-third of the cost). Regardless of timing of cord clamping, the trolley has a role in supporting neonatal care beside the mother and allowing her to share her baby’s first moments.

Challenges

Feedback to the manufacturer led to further modifications. For example, the platform was lengthened to extend its reach, an important development as premature babies tend to have shorter cords than those born at term.

The trip hazard caused by piped gas from wall sockets has reduced acceptability of the trolley in clinical care. A medical air gas cylinder small enough for the trolley is now available and can be used following a local risk assessment (as it is not covered by the CE mark). A disadvantage of this solution is that the cylinder requires frequent refilling. Another less practical option is to put the large cylinder on a second trolley and move this to the bedside with the resuscitation trolley.

Resuscitation can be provided without a pressurised gas supply, using a bag, valve and mask system, and room air. Disadvantages of this method for preterm resuscitation are the inability to control inspiratory time or provide positive end expiratory pressure, and also that oxygen would not be available for those babies who fail to respond to resuscitation with air.

Implications for future research

Once a trolley was available for use, we introduced it into clinical service and conducted a service evaluation161 followed by qualitative interviews to explore the experiences and views of parents and clinicians of neonatal care at birth beside the mother, and of their experiences and views of the trolley (see Work Package 3.5). 39,40 We also developed strategies for providing neonatal care beside the mother using the usual resuscitation equipment. 162 As discussed in Work package 3.5, further research is needed to evaluate the clinical and psychological impact of neonatal care beside the mother and of providing care with the cord intact.

Research aims

Our aims were to provide proof of concept that the trolley could be used in clinical practice, then to assess whether or not the trolley could be used as part of routine care within a busy tertiary care hospital, and whether or not it was acceptable to parents and to clinicians. Therefore, we aimed to conduct:

• a case study of first use of the trolley in clinical practice

• a questionnaire-based service evaluation

• qualitative semistructured interviews with parents and clinicians about their experiences.

Methods for data collection

First use of the trolley in clinical practice was for a baby born by caesarean section at 37 weeks’ gestation who required an EXIT (Ex utero Intrapartum Treatment) procedure163 at Liverpool Women’s hospital. We delivered the baby’s head and shoulders, while the chest and abdomen stayed inside the uterus, and administered tocolytics to maintain placental perfusion. Once a safe airway was achieved (by endotracheal intubation), the baby was born and the umbilical cord cut.

Following this the LifeStart trolley was introduced into clinical service at Liverpool. We conducted a service evaluation from March 2012 to October 2013. 161 The trolley was available at births for which hospital policy required attendance of an advanced neonatal nurse practitioner or a paediatrician. As this was the first time that the trolley was available for routine clinical practice because its use was initially restricted to ‘low-risk’ births. This was extended to high-risk births after review of data from the first 20 births, which was satisfactory. Babies were assessed and resuscitated at birth according to hospital guidelines. Overall, 78 births were included in the evaluation. Data were from the clinical notes, including demographics, temperature of the baby after resuscitation, care provided on the trolley, whether or not there was a need to move the baby to a standard resuscitation platform, and any problems experienced. For 61 of these births, clinicians completed a questionnaire asking their views of care beside the mother and of using the trolley, and any views expressed by the women or their partners. This questionnaire was not completed for 17 births, as the women were recruited into the Cord pilot trial.

For the qualitative interviews, we recruited parents and clinicians between November 2012 and January 2014. Interviews were conducted by a female psychologist or one of three midwives trained in qualitative methods, using a standard schedule of open-ended questions. They were recorded and transcribed, and reporting complied with COREQ. 154

Fifty-six women and their birth partners who had initial neonatal care beside the mother were invited to participate, of whom 30 were interviewed (19 mothers, 10 partners and one grandmother). 39 We conducted 19 interviews, as 11 women chose to be interviewed with their birth partner. Five of the babies required advanced neonatal resuscitation at birth. If their baby had not received intervention at birth, it was explained to parents what this might involve, and they were asked about how they might feel about being so close if this had been necessary. Interviews lasted approximately 30 minutes, and took place either in a private hospital room (17 women) or at the woman’s home (two women). We extracted demographic data from the clinical notes.

Using purposive sampling, 26 clinicians who were present at a birth where initial neonatal care was beside the mother were invited to participate. All initially agreed to be interviewed; as six later declined, 20 were interviewed. 40 Most were neonatal specialist doctors or nurses (Table 2). Five had provided or observed advanced resuscitation beside the mother. Interviews lasted approximately 20 minutes, and took place in a private room on the neonatal unit.

Analysis

For the service evaluation, we assessed usability of the trolley based on the range of resuscitation procedures performed while babies were on it; the feasibility of resuscitation with the cord intact, based on the proportion who received this; and the acceptability to clinicians based on questionnaire responses. The only anticipated safety concern was hypothermia. We monitored unexpected safety issues using the Hospital Incident Reporting System.

The qualitative interviews ended when we achieved data saturation. We used inductive systematic thematic analysis to identify themes across interviews. Data were managed using NVivo software. We identified five themes in the parent interviews: (1) reassurance, (2) involvement of the family, (3) staff communication, (4) reservations and (5) experiences of the trolley. For the clinician interviews, we also identified five themes: (1) parents’ involvement, (2) reservations about neonatal care at birth beside the mother, (3) practical challenges in providing neonatal care beside the mother, (4) comparison of the trolley with usual resuscitation equipment, and (5) training and integration into clinical routine.

Key findings

Successes

As far as we are aware, the work reported here describes experiences at the first hospital to provide initial neonatal care beside the mother within routine clinical service. We have shown that all common resuscitation procedures can be provided this way, with reassurance that this is also safe. Although we used the mobile trolley designed for this purpose, the same care can be provided using the standard resuscitation equipment, as happened in the Cord pilot trial. 70 Although larger than the trolley, standard resuscitation equipment is easily moved alongside the mother, the side panel dropped to allow the mattress on which the baby is placed to be moved closer to the mother, and the height adjusted. Using standard equipment has the advantages of being already available in the room or operating theatre, having the usual radiant heater, and requiring no equipment-specific training as staff are already familiar with its use. Data presented here on initial neonatal care beside the mother are therefore generalisable to settings in which care is provided using standard equipment, with the exception of specific practical issues with the trolley.

Although this requires further investigation, the positive view of a woman whose baby had care beside her and died shortly afterwards has important implications for bereavement.

Challenges

A practical concern about the trolley was the potential trip hazard, especially in the operating theatre, associated with the hose required for gas supply. Replacing one hose with a small oxygen cylinder fixed to the trolley has reduced this hazard. Removing the second hose requires a small air cylinder, which was not commercially available at the time. It is now available, but is not included in the CE mark. A concern about providing care with the cord intact at caesarean section is the potential to compromise the sterile field. Maintaining the sterile field and negotiation about space for the neonatal team at the operating table remain a challenge at caesarean births, especially in an emergency. Our experience is that this becomes easier as staff become familiar with providing this care.

Generalisability of these data is limited by recruitment of both parents and clinicians from a single hospital (where the trolley had been developed) with a particular interest in care beside the mother. Other limitations of the interviews with parents were that participants were not necessarily representative of the wider population, as they were primarily white, they were either married or living with their partner, the babies were all alive at the time of interview, and most babies had not required advanced resuscitation at birth. Parents may have been reluctant to criticise their care, but we mitigated this risk by having an interviewer who was not directly involved in their care or care of their baby.

Implications for future research

The work presented here should be repeated in other hospitals and should include using the standard resuscitation equipment as well as the mobile trolley to assess whether or not outcomes, experiences and costs are similar. Further research should include parents from more diverse backgrounds, those whose baby required advanced resuscitation and those whose baby died. Further research should also include follow-up of both parents and infants to assess any long-term impact.

The optimal study design for further evaluation would be a randomised trial comparing outcome for parents who are offered initial neonatal care beside the mother with the outcome for those offered traditional care only.

Research aims

The aim of the Cord pilot trial was to assess the feasibility of conducting a large multicentre randomised trial in the UK comparing deferred cord clamping and immediate neonatal care (if needed) with the cord intact, with immediate cord clamping and with immediate neonatal care after cord clamping for very preterm births. Recruitment was planned for 1 year and feasibility was assessed based on prespecified process measures, such as recruitment, compliance, completeness of data collection and participants’ views.

At an interim review, the independent TSC agreed that feasibility criteria had been met and advised that recruitment continue while funding was sought for the full trial. Recruitment closed in February 2015, when this funding application was unsuccessful. For the planned full trial, joint primary outcomes were death before discharge and IVH (all grades), hence these outcomes were considered the primary outcomes for the extended pilot trial. As the two-stage consent pathway was an important addition to the trial protocol, we also conducted a qualitative evaluation of the alternative consent pathways.

Methods for data collection

Recruitment to this pragmatic trial was at eight UK tertiary maternity units: five had contributed to earlier work in the programme and three had not been involved and so were more typical of UK sites. Three parent representatives were coinvestigators, involved in identifying the research question, securing funding, and designing and conducting the study. Target recruitment for assessment of feasibility was 100–110 women over 1 year, based on target accrual of 16–18% of eligible births. As we planned this as a pilot, there was no formal power calculation.

Women expected to have a live birth before 32 weeks’ gestation were potentially eligible. The intervention was umbilical cord clamping after at least 2 minutes and neonatal stabilisation and resuscitation, if needed, with the cord intact. Babies were placed onto a firm surface with access to resuscitation equipment; six sites used their usual resuscitation equipment moved alongside the woman’s bed162 (153 women recruited) and two sites used the mobile trolley (108 women). 69 The control was cord clamping within 20 seconds and neonatal stabilisation and resuscitation, if needed, after clamping. For both groups, neonatal care was based on local unit policy and consistent with newborn life support guidelines. 57,164

For the assessment of feasibility we prespecified measures of feasibility. For the planned main trial, joint primary outcomes were death before discharge and IVH (all grades);165 therefore, these were the main outcomes for the extended pilot trial. We adjudicated diagnosis of IVH. Secondary outcomes for the babies included severe IVH (grade 3 or 4),165 a range of measures of neonatal morbidity and neurodevelopment in early childhood. Secondary outcomes for the women included complications of the third stage of labour, well-being and satisfaction with care at birth. Follow-up for women was by self-completed questionnaire at 2–3 months and again at 1 year, and for the children at the age of 2 years (corrected for gestation at birth) by parent-completed Ages and Stages Questionnaire (ASQ)166 and an assessment at home using the Bayley Scales of Infant Development III (Bayley-III). 167 If no ASQ or Bayley-III data were available, we asked the site to complete an outcome form using routine clinical data. We agreed the statistical analysis plan for the extended study before unblinding the data.

Whenever possible, we offered women the usual written consent process (Figure 8). However, if birth was imminent and the attending clinician considered it appropriate, we offered women a brief explanation of the study and randomised those who said ‘yes’ (i.e. they gave oral assent). Before discharge from hospital, these women had an opportunity to discuss the study in detail and to give written consent for participation in follow-up. This two-stage process was developed in consultation with the NCT and Bliss.

FIGURE 8.

The two alternative consent pathways. a, Women approached to give oral assent in established labour or at emergency caesarean section only if the attending clinicians considered it to be appropriate. Women were not approached if there was insufficient time to give a brief verbal summary of the trial or if they did not speak fluent English and no translator was available. How long was required for oral assent depended on factors such as how much the woman already knew about the study, and her knowledge and wishes about care during the third stage. If recruitment was after oral assent, (1) women were approached before discharge to give written consent to participation in follow-up and (2) the chief investigator was notified of this within 15 days, and this recruitment was monitored by the TSC. Derived from Pushpa-Rajah et al. 168 © Pushpa-Rajah et al. 2014. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Derived from Pushpa-Rajah et al. 168 © Pushpa-Rajah et al. 2014. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Randomisation was by the attending clinicians, using sealed consecutively numbered opaque envelopes. Clinical staff were trained, with regular study-specific training provided at each site to as many neonatal staff, midwives, obstetricians and theatre staff as possible (see Appendix 17). We supported site training with film clips of simulations for providing immediate neonatal care with the cord intact, and of role-playing to offer consent via the two pathways. Multidisciplinary teamworking on site was essential for successful trial conduct.

We sent 179 women in the trial an invitation to participate in semistructured interviews. A total of 23 were interviewed (usual consent pathway, n = 18; two-stage pathway, n = 5). The interview schedule consisted of open-ended questions42 to explore the women’s views and experiences of being offered participation. Interviews lasted approximately 30 minutes and took place either at their home or by telephone. We sent 20 clinicians who recruited women invitations for semistructured interviews, and interviewed 17 from seven sites. The interview schedule consisted of open-ended questions, to explore clinicians’ views and experiences of offering participation. 41 Interviews took place either in a private hospital room or by telephone. They lasted 20–30 minutes and were audio-recorded and transcribed. Reporting complied with COREQ. 154

Analysis

For the assessment of feasibility of a large trial, the prespecified outcomes are presented based on recruitment at 1 year. For participants in the extended pilot, analyses were based on the groups as randomly allocated (intention to treat). Women who gave birth after 35⁺⁶ weeks were not included in the analyses, as outcomes for these babies are different from those born very preterm. Where appropriate, results are presented as RR or risk difference with 95% CIs.

For the qualitative interviews, we used inductive thematic analysis using NVivo version 10 software.

Key findings

Results by allocated group, for all women randomised

Of 945 women approached, 472 (50%) gave consent and 261 (28%) were randomised (Figure 9). 71 One-third of women were randomised before 28 weeks’ gestation and two-thirds of women were randomised before 30 weeks. Just over half were women in their first pregnancy and a similar proportion had a caesarean birth. Compliance was good, with median time to cord clamping 120 seconds for women allocated to cord clamping after at least 2 minutes, and 11 seconds for those allocated to cord clamping within 20 seconds. Neonatal care was comparable between the two allocated groups.

FIGURE 9.

The CONSORT (Consolidated Standards of Reporting Trials) flow diagram for outcome to discharge from hospital. a, Woman changed her mind (n = 2), intrauterine death (n = 2), equipment failure (n = 1), study closed (n = 1), randomisation website failed (n = 1). b, Compliance reported by baby. c, Birth at ≥ 36 weeks’ gestation (all singleton pregnancies). d, One woman and her baby withdrew, data reported for mortality only. Reproduced from Duley et al. 71 © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2018. All rights reserved. No commercial use is permitted unless otherwise expressly granted. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/.

Reproduced from Duley et al. 71 © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2018. All rights reserved. No commercial use is permitted unless otherwise expressly granted. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/

There were no clear differences between the allocated groups in either of the primary outcomes (Table 3). Three-quarters of infants who died were born before 28 weeks’ gestation. The three stillborn babies were all born before 28 weeks’ gestation and resuscitation was attempted at birth. There were no clear differences between the allocated groups for any other outcome for the baby, or in any outcome for the mother.

TABLE 3 - Primary outcomes at discharge from hospital for all babies

Includes the baby of one woman who withdrew.

Excluding death due to congenital malformation gives a RR of 0.50 (95% CI 0.21 to 1.20).

Major congenital malformation, not known before birth.

For live births: n = 134, clamp ≥ 2 minutes; n = 132, clamp ≤ 20 seconds.

IVH diagnosis based on scan adjudication for 81 infants, scan report only for 8, and site data only for 1.

Reproduced from Duley et al. 71 © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2018. All rights reserved. No commercial use is permitted unless otherwise expressly granted. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/.

	Clamp ≥ 2 minutes + neonatal care with cord intact, n (n = 135)	Clamp ≤ 20 seconds + neonatal care after clamping, n (n = 135)a	Risk ratio (95% CI)
Death	7 (5%)	15 (11%)	0.47 (0.20 to 1.11)b
Stillbirth	1	2
Early neonatal death	3	7
Late neonatal death	2	5
Post neonatal death	1	1
For babies who died, gestation at birth (weeks)
30 – 31+6	–	1c
28 – 29+6	1	3
26 – 27+6	–	4
< 26	6	7
Any IVH (grade 1–4)d,e	43 (32%)	47 (36%)	0.90 (0.64 to 1.26)
Alive at discharge	38	42
Severe IVH (grade 3–4)d	6 (4%)	7 (5%)

Successes

We conducted a substantial and successful pilot trial in a difficult clinical area, and demonstrated feasibility of conducting a large multicentre study. As the study was conducted within existing NHS clinical services, results are generalisable to similar settings.

A key limitation of previous trials is that they largely excluded babies who were likely to need immediate resuscitation at birth. 172 We developed two successful strategies for recruiting such high-risk infants. First, for women allocated deferred clamping, neonatal stabilisation and resuscitation was available with the cord intact. This had the additional advantage that stabilisation was equivalent in the two intervention groups. Second, if birth was imminent, we could offer women the opportunity to participate using the two-stage consent pathway. We recruited one-quarter of women using this pathway – women and babies it would not otherwise have been possible to recruit. Qualitative interviews and questionnaire responses from the women suggest the two-stage consent pathway was acceptable in the trial.

Design and conduct of the Cord pilot trial drew on outputs from earlier parts of the programme, and the experience we developed of working as a multidisciplinary team. In particular, the strong patient and public involvement in the programme and in the trial was essential for changing the way neonatal care was provided for very preterm births, and for developing the two-stage consent pathway.

Our study is also the first cord clamping trial to use independent adjudication of cranial ultrasound scans, thereby reducing potential for bias and improved reliability in ascertainment of IVH. 173 Other successes were being the one of the first deferred cord clamping versus early cord clamping trials at preterm birth to report outcome for the women, and outcome for the children at 2 years of age (corrected for gestation at birth).

Challenges

Although we demonstrated feasibility and had strong support for the main trial, both from the independent oversight committees and from clinical sites keen to participate, we were not able to secure the necessary funding. The reason given by the funding board was that they had concerns over the timing of the study given other relevant trials that had yet to conclude.

When an external pilot trial is successful, transforming it into an internal pilot by continuing into the full trial may maximise efficiency and value for money, but our experience demonstrates it is a challenge to achieve. 174 Nevertheless, data from our trial will contribute to the design of any future trials,72 in particular through the planned prospective meta-analysis and the experience gained in delivering the intervention.

As for many perinatal trials, long-term follow-up of the children was challenging and it required a range of strategies. The importance of complete follow-up to minimise potential bias is underlined by the apparent difference between the allocated groups in response to follow-up and that children without follow-up data in the intervention group were higher risk than those for whom data were available.

Implications for future research

With appropriate planning and training, large multicentre trials of interventions around the time of preterm birth are feasible within existing maternity and neonatal services in the NHS. To improve the opportunity for participation, such trials should consider including the two-stage consent pathway. Nevertheless, this two-stage pathway merits further evaluation in a wider range of studies. Providing neonatal care beside the mother so that she can see and, if she wishes, touch her baby, also merits further evaluation in a wider range of clinical settings, and with follow-up to assess any long-term effects for parents or baby.

Since our trial was planned, new research has emerged that should be considered in planning new trials of timing of cord clamping. This includes work with pregnant sheep which has improved understanding of the physiology of transition from the fetal to neonatal circulation51,159,175,176 and the publication of other relevant trials including a large international trial. 177 The design of any future trial, including the interventions to be compared, should therefore be informed by the systematic review and meta-analysis based on individual participant data (IPD) planned in our final work package.

Research aims

The aims were to establish a collaborative group of triallists to conduct an IPD meta-analysis to assess whether timing of cord clamping and other strategies to alter placental transfusion at preterm birth influence (1) the composite outcome of death or serious morbidity at discharge from hospital and (2) disability-free survival in early childhood (aged 2–3 years). This work is referred to as Cord Clamping and Placental Transfusion at Preterm birth (CCPTP). The unexpected rapid increase in small randomised trials being published led to agreement by the CCPTP collaboration that, if sufficient funding is available, the planned prospective meta-analysis should be expanded to a retrospective IPD meta-analysis, with a nested prospective meta-analysis.

Methods for data collection

The CCPTP collaborative group of triallists was established and the protocol (see Appendix 24) discussed at the annual face-to-face meeting. The PROSPERO registration identifier is CRD42013004405. Following quarterly searches of trial registries, the chief investigators of studies potentially eligible for the prospective meta-analysis are contacted and invited to join the collaboration. Studies are included if they are randomised trials meeting the eligibility criteria and the investigator(s) are blind to outcome data by intervention group at the time that they agree to the CCPTP protocol.

Eligibility criteria are that participants gave birth (if women were randomised) or were born (if baby was randomised) before 37 completed weeks’ gestation. Interventions are comparisons of alternative policies for timing of cord clamping, and other strategies to influence umbilical flow (e.g. lowering the baby below the level of the placenta, use of uterotonic drugs and umbilical cord milking or stripping). Studies are included if they compare strategies to maintain ‘physiological’ umbilical flow (i.e. none or minimal intervention) with strategies that aim to alter umbilical blood flow (e.g. using gravity by lowering the baby, cord milking or stripping). Studies comparing alternative strategies for influencing umbilical flow without a ‘timing of cord clamping’ arm will also be included.

Planned comparisons are:

1. Early cord clamping (short delay) versus deferred cord clamping (long delay).

2. Early cord clamping (short delay) versus umbilical cord milking or stripping.

3. Deferred cord clamping (long delay) versus umbilical cord milking or stripping.

4. Deferred cord clamping (long delay) plus umbilical cord milking or stripping versus early cord clamping (short delay).

As there is no consensus about the definition of ‘early’ and ‘deferred’ cord clamping, early clamping is defined whenever possible as within 30 seconds, and deferred clamping as after at least 1 minute. Trials that use different definitions will be included in the analysis.

As outcomes for babies born very preterm (< 32 weeks’ gestation) are different for those born moderately preterm (32 to 37 weeks’ gestation), separate analyses are planned for these two groups of infants. For those born < 32 weeks’ gestation, primary outcomes are (1) death or serious morbidity at hospital discharge (serious morbidity is defined as one or more of severe IVH, necrotising enterocolitis, late onset sepsis, chronic lung disease and retinopathy of prematurity); (2) chronic lung disease; and (3) neurodevelopmental delay in early childhood. Secondary outcomes include individual items in the composite outcome and measures of neonatal morbidity and adverse neurodevelopmental outcome in early childhood.

For infants born at or after 32 weeks’ gestation, the primary outcome is admission to NICU. Secondary outcomes are measures of neonatal morbidity, iron status in infancy, and adverse neurodevelopmental outcome in early childhood.

For the women, outcomes are complications of the third stage of labour, postpartum infection, breastfeeding and postnatal depression.

Analysis

The detailed statistical analysis plan will be developed and agreed before any analysis begins. Planned subgroup analyses are based on:

• participant-level characteristics – gestation at birth, whether singleton or multiple pregnancy, mode of birth and whether or not spontaneous onset of labour

• hospital-level characteristics – highest level of neonatal unit available at site, timing of uterotonic drug, timing of cord clamping, planned position of the baby relative to the placenta with the cord intact, and whether or not babies needing immediate resuscitation at birth were recruited.

To assess whether or not the results are robust to trial quality and different methods of analysis, the following sensitivity analyses are planned for the primary outcomes, if there are sufficient data: excluding studies with a high risk of bias; for comparison of early with deferred clamping, analysis weighted by observed between-arm difference in mean timing between the groups; analysis weighted by degree of separation in haemoglobin between the groups.

If a policy of deferred cord clamping appears to be beneficial, additional analyses will therefore explore the effects based on duration of allocated deferred cord clamping and, if the allocation was clamping after 20 to 30 seconds, whether or not neonatal care, if needed, was provided with the cord intact.

Successes

Currently we know of 53 trials, involving 11,811 infants (3020 for long-term outcomes). Our success in establishing the CCPTP collaboration emphasises the international interest in this topic. Annual collaborators meetings, held alongside neonatal conferences to minimise cost, have facilitated collaboration and biannual newsletters provide regular updates.

Success of the collaboration between the Nottingham Clinical Trials Unit and the National Health and Medical Research Council Clinical Trials Centre at the University of Sydney, which jointly support the secretariat for CCPTP, means that the first cycle of analysis is now planned for 2020 and funding has been secured to expand this to include a retrospective meta-analysis based on IPD, with a nested prospective meta-analysis (see Appendix 25). The collaboration also changed its name to systematic review and network meta-analysis with individual participant data on Cord Management at Preterm birth (iCoMP).

Challenges

The main challenge for this project is common to all planned prospective meta-analyses: the time scale for analysis is unpredictable. Results for the large study planned for inclusion in CCPTP, the Australian Placental Transfusion Trial, were published later than anticipated,177 as were results for the Cord pilot trial. Hence, the first cycle of analysis was not possible within the programme time scale. The number of new trials under way is accelerating, with 29 registered in the past 2 years. Many of these are relatively small and data analysis may have begun before the investigators agree the CCPTP protocol. Growing numbers being published risked undermining the relevance of a standalone prospective meta-analysis. Hence, we expanded the protocol to allow inclusion of a retrospective IPD meta-analysis, with network meta-analysis. Work commenced on this IPD in 2019.

Implications for future research

Placental transfusion at preterm birth is a current ‘hot topic’, with wide variation in clinical practice. This has led to a plethora of recent trials largely undertaken in isolation, with little standardisation of the intervention(s) being tested or the outcomes being collected. Value of these existing trials will be substantially enhanced by combining and fully utilising the IPD, as planned in CCPTP. Provision of individual participant data by each trial investigator will provide the required statistical power and enable reliable subgroup analyses to be undertaken.

Results of these analyses based on IPD from each trial will inform clinical practice and will identify the most promising intervention(s) for further evaluation. Co-ordinating international efforts in this way will also help achieve consensus on the most important clinical outcomes to assess in any future trials.

Limitations of our programme

Preterm birth is associated with factors such as lower socioeconomic status, ethnicity and maternal age. For our prioritisation process, despite implementing strategies to reach a representative population, respondents were primarily white and a relatively high proportion of respondents were homeowners and were therefore not representative of those most affected by preterm birth. This may limit relevance of the ranked priorities to other populations.

For the qualitative work on very preterm birth, limitations of our data were that participants in the interviews came from two hospitals in the south of England and they were mainly white married women; therefore, not typical for vey preterm birth. For development of the P-BESS questionnaire, the sample size was relatively small for a factor analysis and, again, was unrepresentative.

Generalisability of our findings on neonatal care beside the mother is limited by recruitment of parents and clinicians from a single hospital with a particular interest in care beside the mother, and where the trolley had been developed. In addition, we were again not successful in recruiting parents for interview who were representative of the wider population with very preterm birth. Furthermore, the babies were all alive at the time of interview and most had not required advanced resuscitation at birth.

Finally, although out pilot trial demonstrated feasibility and had strong support for progression to the main trial, we were not able to secure the necessary funding. Hence, we have not been able to conduct a definitive trial.

Implications for health care

Providing initial neonatal care at birth beside the mother is feasible within the NHS, acceptable to parents and to clinicians, and both parents and clinicians felt that it improved communication. Although further evaluation is required, our data suggest that, with adequate training, preparation and support, parents and clinicians prefer care beside the mother rather than moving the baby away from the mother for assessment, stabilisation and resuscitation at birth. This has major implications for maternity services. Although this is a relatively simple change in practice, implementation of neonatal care beside the mother at birth requires a change in culture to multidisciplinary teamworking. Care beside the mother can be done either by adapting the usual resuscitation equipment or by using a mobile trolley specifically designed for the purpose.

The Cord pilot trial demonstrates that neonatal stabilisation and resuscitation can be provided with the umbilical cord intact, providing a practical and generalisable strategy for supporting continued umbilical flow at birth. Results of this trial add to the growing body of data from randomised trials suggesting that there is no need to rush to clamp the cord at very preterm birth. Although the optimal policy for cord clamping at very preterm birth remains unclear, waiting 20 to 30 seconds seems prudent.

Only a small proportion of births are very preterm, and parents’ experiences of care during the birth can differ from those having a term birth. The P-BESS questionnaire was developed to assess experiences and satisfaction with care during very preterm birth and can be used by hospitals to assess their care, potentially identifying areas where care can be improved. A total score can be used to compare across hospitals and differing practices, whereas individual aspects of care can be evaluated using the subscales.

Recommendations for future research

• The 15 top research priorities should guide researchers and funding bodies when planning new research.

• When planning and funding preterm birth research, it may be relevant to consider not only the top 15 priorities but also the full list of 104 unanswered questions, the top 30 ranked by those affected by preterm birth and the top 30 ranked by health-care professionals.

• Future prioritisation processes, particularly those with a similar wide range of health-care professionals, should anticipate potential different perspectives and mitigate any imbalance where possible, and should report voting separately by ‘service users’ and health-care professionals. Health-care professionals who are also researchers should declare this potential conflict before participating in the prioritisation workshop, so that it can be taken into account.

• Important ‘evidence gaps’ for neonatal care at birth include pharmacological intervention for transition support or resuscitation of newborn infants, and the effectiveness of new, less invasive forms of airway management (and related issues regarding surfactant delivery).

• Since our programme started, both timing of cord clamping for very preterm births and providing neonatal care with the cord intact have become increasingly topical. New research has emerged, including the work conducted within this programme, which is changing opinion and clinical practice. Therefore, it would be timely to repeat a survey of practice.

• Further research is needed to replicate our findings about parent experiences at very preterm birth, and to explore whether or not there is variation for parents from different backgrounds.

• Further studies are needed to test the P-BESS questionnaire in a larger, more representative sample of parents.

• Fathers of preterm or sick babies are a difficult group to recruit into research. Our work highlights the importance of including them in research studies to ensure that their perspective is represented.

• Further research is needed to evaluate clinical outcomes, parent experiences and the costs of neonatal care beside the mother and of care with the cord intact. This should include a wider range of settings and using the standard resuscitation equipment as well as the mobile trolley. It should involve parents from more diverse backgrounds, those whose baby required advanced resuscitation, and those whose baby died. Future studies should also include follow-up of parents and infants.

• The optimal study design would be a randomised trial to compare outcomes (for children and parents) for parents offered initial neonatal care beside the mother, with outcomes for those offered traditional care.

• With appropriate planning and training, large multicentre trials of interventions around the time of preterm birth are feasible within existing maternity and neonatal services in the NHS.

• The two-stage consent pathway should be included in future perinatal trials to increase participation when there is little time in which to seek consent to participation. It also has potential to increase participation in other emergency or time-critical trials, but this merits further evaluation.

• Design of future trials comparing alternative policies for cord management should take account of results from our trial and the planned meta-analysis based on individual participant data from all relevant trials.

Programme steering group

Lelia Duley (chairperson), Susan Ayers, Zoe Chivers, Jon Dorling, David Field, Gill Gyte, William McGuire, Chris Megone, Sam Oddie, Sandy Oliver, Jim Thornton, Andrew Weeks and Charles William Yoxall.

Programme management team

Lelia Duley, Charlotte Lloyd (2011–13), Virginia Portillo (2014–15), Sarah Somerset (2016–17), Sarah Smith (2011–13) and Ruth Fletcher (2014–16).

Thanks to Lindsay Armstrong-Buisseret for assistance with compiling this report.

Work package 1

Work package 2.1

Work package 2.2

Work package 3.1

Work package 3.2

Work package 3.3

Work package 3.4

Work package 3.5

Work package 4

Work package 5

Contributions of authors

Lelia Duley (Professor of Clinical Trials, University of Nottingham) substantially contributed to the design, methodology, data collection, analysis and reporting for all of the work packages. She was the grant holder, the chief investigator for the Cord pilot trial and chaired the programme steering group. She led the writing, revision and approval of the final report.

Jon Dorling (Associate Professor of Neonatology, University of Nottingham) substantially contributed to the design, methodology, data collection, analysis and reporting for the measuring placental transfusion work package (3.3), and the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Susan Ayers (Professor of Maternal Health, City, University of London) led on the qualitative aspects of the programme. She substantially contributed to the design, methodology, data collection, analysis and reporting for the parents’ views of care at preterm birth work package (3.2), and the evaluating parents’ and clinicians’ views of neonatal care beside the mother work package (3.5). She also led on the qualitative aspects of the Cord pilot trial work package (4). She was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Sandy Oliver (Professor of Public Policy, University College London) led the preterm birth priority setting work package (1) including supervising the doctor of philosophy (PhD) student, and substantially contributed to the design, methodology, data collection, analysis and reporting for the framework synthesis of ethics issues in recruitment of preterm or sick infants to perinatal trials work package (2.2). She was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Charles William Yoxall (Consultant Neonatologist, Liverpool Women’s Hospital) substantially contributed to design and development of the BASICS trolley work package (3.4), and the design, methodology, data collection, analysis and reporting for the evaluating parents’ and clinicians’ views of neonatal care beside the mother work package (3.5) and the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Andrew Weeks (Professor of International Maternal Health Care, University of Liverpool) substantially contributed to design and development of the BASICS trolley work package (3.4), and the design, methodology, data collection, analysis and reporting for the evaluating parents’ and clinicians’ views of neonatal care beside the mother work package (3.5) and the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Chris Megone (Professor of Philosophy, University of Leeds) led the design, methodology, data collection, analysis and reporting for the framework synthesis of ethics issues in recruitment of preterm or sick infants to perinatal trials work package (2.2), and advised on development of the two-stage oral assent consent pathway in the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Sam Oddie (Consultant Neonatologist, University of York) substantially contributed to the design, methodology, data collection, analysis and reporting for the neonatal care at birth work package (3.1), the measuring placental transfusion work package (3.3), and the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Gill Gyte (NCT) provided patient and public involvement expertise across all work packages. In particular she substantially contributed to the design, methodology, data collection, analysis and reporting for the preterm birth priority setting work package (1), the framework synthesis of ethics issues in recruitment of preterm or sick infants to perinatal trials work package (2.2), the parents’ views of care at preterm birth work package (3.2), design and development of the BASICS trolley work package (3.4), and the Cord pilot trial work package (4). She was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Zoe Chivers (Bliss) provided patient and public involvement expertise across all work packages. In particular she substantially contributed to the design, methodology, data collection, analysis and reporting for the preterm birth priority setting work package (1), and the Cord pilot trial work package (4). She was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Jim Thornton (Professor of Obstetrics and Gynaecology, University of Nottingham) supported the measuring placental transfusion work package (3.3) and substantially contributed to the design, methodology, data collection, analysis and reporting for the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

David Field (Professor of Neonatal Medicine, University of Leicester) substantially contributed to the design, methodology, data collection, analysis and reporting for the Cord pilot trial work package (4). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Alexandra Sawyer (Research Fellow, University of Brighton) substantially contributed to the design, methodology, data collection, analysis and reporting for the parents’ views of care at preterm birth work package (3.2), the evaluating parents’ and clinicians’ views of neonatal care beside the mother work package (3.5), and the qualitative aspects of the Cord pilot trial work package (4).

William McGuire (Professor of Child Health, University of York) substantially contributed to the design, methodology, data collection, analysis and reporting for the evidence-based immediate care of the very preterm infant work package (2.1). He was a member of the programme steering group and contributed to the writing, revision and approval of the final report.

Publications

Uhm S, Liabo K, Stewart R, Rees R, Oliver S. Patient and public perspectives shaping scientific and medical research: panels for data, discussions, and decisions. Patient Intell 2012;4:1–10.

Arnold L, Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Qualitative Collaborative Group. Parents’ first moments with their very preterm babies: a qualitative study. BMJ Open 2013;3:e002487.

Sawyer A, Ayers S, Abbott J, Gyte G, Rabe H, Duley L. Measures of satisfaction with care during labour and birth: a comparative review. BMC Pregnancy Childbirth 2013;13:108.

Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Qualitative Collaborative Group. Parents’ experiences and satisfaction with care during the birth of their very preterm baby: a qualitative study. BJOG 2013;120:637–43.

Duley L, Uhm S, Oliver S, Preterm Birth Priority Setting Partnership Steering Group. Top 15 UK research priorities for preterm birth. Lancet 2014;383:2041–2.

Oddie S, Rhodes P, Very Preterm Birth Qualitative Collaborative Group. Barriers to deferred cord clamping in preterm infants. Arch Dis Child Fetal Neonatal Ed 2014;99:F391–4.

Pushpa-Rajah A, Bradshaw L, Dorling J, Gyte G, Mitchell EJ, Thornton J, Duley L, Cord Pilot Trial Collaborative Group. Cord pilot trial – immediate versus deferred cord clamping for very preterm birth (before 32 weeks gestation): study protocol for a randomized controlled trial. Trials 2014;15:258.

Russell G, Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Qualitative Collaborative Group. Parents’ views on care of their very premature babies in neonatal intensive care units: a qualitative study. BMC Pediatr 2014;14:230.

Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Collaborative Group. Measuring parents’ experiences and satisfaction with care during very preterm birth: a questionnaire development study. BJOG 2014;121:1294–301.

Singh Y, Oddie S. Marked variation in delivery room management in very preterm infants. Resuscitation 2013;84:1558–61.

Thomas MR, Yoxall CW, Weeks AD, Duley L. Providing newborn resuscitation at the mother’s bedside: assessing the safety, usability and acceptability of a mobile trolley. BMC Pediatr 2014;14:135.

Bradshaw LE, Pushpa-Rajah A, Dorling J, Mitchell E, Duley L. Cord pilot trial: update to randomised trial protocol. Trials 2015;16:407.

Sawyer A, Ayers S, Bertullies S, Thomas M, Weeks AD, Yoxall CW, Duley L. Providing immediate neonatal care and resuscitation at birth beside the mother: parents’ views, a qualitative study. BMJ Open 2015;5:e008495.

Wilman E, Megone C, Oliver S, Duley L, Gyte G, Wright JM. The ethical issues regarding consent to clinical trials with pre-term or sick neonates: a systematic review (framework synthesis) of the empirical research. Trials 2015;16:502.

Yoxall CW, Ayers S, Sawyer A, Bertullies S, Thomas M, D Weeks A, Duley L. Providing immediate neonatal care and resuscitation at birth beside the mother: clinicians’ views, a qualitative study. BMJ Open 2015;5:e008494.

Corbett MS, Moe-Byrne T, Oddie S, McGuire W. Randomisation methods in emergency setting trials: a descriptive review. Res Synth Methods 2016;1:46–54.

Megone C, Wilman E, Oliver S, Duley L, Gyte G, Wright J. The ethical issues regarding consent to clinical trials with pre-term or sick neonates: a systematic review (framework synthesis) of the analytical (theoretical/philosophical) research. Trials 2016;17:443.

Batey N, Yoxall CW, Fawke JA, Duley L, Dorling J. Fifteen-minute consultation: stabilisation of the high-risk newborn infant beside the mother. Arch Dis Child Educ Pract Ed 2017;102:235–8.

Chhoa CY, Sawyer A, Ayers S, Pushpa-Rajah A, Duley L. Clinicians’ views and experiences of offering two alternative consent pathways for participation in a preterm intrapartum trial: a qualitative study. Trials 2017;18:196.

Sawyer A, Chhoa C, Ayers S, Pushpa-Rajah A, Duley L. Women’s views and experiences of two alternative consent pathways for participation in a preterm intrapartum trial: a qualitative study. Trials 2017;18:422.

Duley L, Dorling J, Pushpa-Rajah A, Oddie SJ, Yoxall CW, Schoonakker B, et al. Randomised trial of cord clamping and initial stabilisation at very preterm birth. Arch Dis Child Fetal Neonatal Ed 2018;103:F6–F14.

Bradshaw L, Sawyer A, Armstrong-Buisseret L, Mitchell E, Ayers S, Duley L. Cord pilot trial, comparing alternative policies for timing of cord clamping before 32 weeks gestation: follow-up for women up to one year. BMC Pregnancy Childbirth 2019;19:78.

Bradshaw L, Sawyer A, Mitchell E, Armstrong-Buisseret L, Ayers S, Duley L. Women’s experiences of participating in a randomised trial comparing alternative policies for timing of cord clamping at very preterm birth: a questionnaire study. Trials 2019;20:225.

Armstrong-Buisseret L, Powers K, Dorling J, Bradshaw L, Johnson S, Mitchell E, Duley L. Randomised trial of cord clamping at very preterm birth: outcomes at 2 years. [published online ahead of print August 1 2019]. Arch Dis Child Fetal Neonatal Ed 2019.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

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, CCF, NETSCC, PGfAR or the Department of Health and Social Care. 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 PGfAR programme or the Department of Health and Social Care.

Work package 1

Duley L, Uhm S, Oliver S, Preterm Birth Priority Setting Partnership Steering Group. Top 15 UK research priorities for preterm birth. Lancet 2014;383:2041–2. See Appendix 2.

Oliver S, Uhm S, Duley L, Crowe S, David A, James CP, et al. Top research priorities for preterm birth: results of a prioritisation partnership between people affected by preterm birth and healthcare professionals. Submitted, in peer review. See Appendix 3.

Uhm S, Liabo K, Stewart R, Rees R, Oliver S. Patient and public perspectives shaping scientific and medical research: panels for data, discussions, and decisions. Patient Intell 2012;4:1–10.

Work package 2.1

Corbett MS, Moe-Byrne T, Oddie S, McGuire W. Randomisation methods in emergency setting trials: a descriptive review. Res Synth Methods 2016;1:46–54.

Moe-Byrne T, Brown JV, McGuire W. Naloxone for opiate-exposed newborn infants. Cochrane Database Syst Rev 2013;2:CD003483.

Work package 2.2

Wilman E, Megone C, Oliver S, Duley L, Gyte G, Wright JM. The ethical issues regarding consent to clinical trials with pre-term or sick neonates: a systematic review (framework synthesis) of the empirical research. Trials 2015;16:502. See Appendix 6.

Megone C, Wilman E, Oliver S, Duley L, Gyte G, Wright J. The ethical issues regarding consent to clinical trials with pre-term or sick neonates: a systematic review (framework synthesis) of the analytical (theoretical/philosophical) research. Trials 2016;17:443. See Appendix 7.

Work package 3.1

Singh Y, Oddie S. Marked variation in delivery room management in very preterm infants. Resuscitation 2013;84:1558–61. See Appendix 8.

Oddie S, Rhodes P, Very Preterm Birth Qualitative Collaborative Group. Barriers to deferred cord clamping in preterm infants. Arch Dis Child Fetal Neonatal Ed 2014;99:F391–4. See Appendix 9.

Work package 3.2

Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Qualitative Collaborative Group. Parents’ experiences and satisfaction with care during the birth of their very preterm baby: a qualitative study. BJOG 2013;120:637–43. See Appendix 10.

Arnold L, Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Qualitative Collaborative Group. Parents’ first moments with their very preterm babies: a qualitative study. BMJ Open 2013;3:e002487. See Appendix 11.

Russell G, Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Qualitative Collaborative Group. Parents’ views on care of their very premature babies in neonatal intensive care units: a qualitative study. BMC Pediatr 2014;14:230.

Sawyer A, Ayers S, Abbott J, Gyte G, Rabe H, Duley L. Measures of satisfaction with care during labour and birth: a comparative review. BMC Pregnancy Childbirth 2013;13:108.

Sawyer A, Rabe H, Abbott J, Gyte G, Duley L, Ayers S, Very Preterm Birth Collaborative Group. Measuring parents’ experiences and satisfaction with care during very preterm birth: a questionnaire development study. BJOG 2014;121:1294–301. See Appendix 12.

Work package 3.4

Weeks AD, Watt P, Yoxall CW, Gallagher A, Burleigh A, Bewley S, et al. Innovation in immediate neonatal care: development of the Bedside Assessment, Stabilisation and Initial Cardiorespiratory Support (BASICS) trolley. BMJ Innov 2015;1:53–8. See Appendix 14.

Work package 3.5

Sawyer A, Ayers S, Bertullies S, Thomas M, Weeks AD, Yoxall CW, Duley L. Providing immediate neonatal care and resuscitation at birth beside the mother: parents’ views, a qualitative study. BMJ Open 2015;5:e008495. See Appendix 15.

Yoxall CW, Ayers S, Sawyer A, Bertullies S, Thomas M, D Weeks A, Duley L. Providing immediate neonatal care and resuscitation at birth beside the mother: clinicians’ views, a qualitative study. BMJ Open 2015;5:e008494. See Appendix 16.

Thomas MR, Yoxall CW, Weeks AD, Duley L. Providing newborn resuscitation at the mother’s bedside: assessing the safety, usability and acceptability of a mobile trolley. BMC Pediatr 2014;14:135.

Work package 4

Duley L, Dorling J, Pushpa-Rajah A, Oddie SJ, Yoxall CW, Schoonakker B, et al. Randomised trial of cord clamping and initial stabilisation at very preterm birth. Arch Dis Child Fetal Neonatal Ed 2018;103:F6–F14. See Appendix 18.

Chhoa CY, Sawyer A, Ayers S, Pushpa-Rajah A, Duley L. Clinicians’ views and experiences of offering two alternative consent pathways for participation in a preterm intrapartum trial: a qualitative study. Trials 2017;18:196. See Appendix 19.

Sawyer A, Chhoa C, Ayers S, Pushpa-Rajah A, Duley L. Women’s views and experiences of two alternative consent pathways for participation in a preterm intrapartum trial: a qualitative study. Trials 2017;18:422. See Appendix 20.

Bradshaw L, Sawyer A, Armstrong-Buisseret L, Mitchell E, Ayers S, Duley L. Cord pilot trial, comparing alternative policies for timing of cord clamping before 32 weeks gestation: follow-up for women up to one year. BMC Pregnancy Childbirth 2019;19:78. See Appendix 21.

Bradshaw L, Sawyer A, Mitchell E, Armstrong-Buisseret L, Ayers S, Duley L. Women’s experiences of participating in a randomised trial comparing alternative policies for timing of cord clamping at very preterm birth: a questionnaire study. Trials 2019;20:225. See Appendix 22.

Armstrong-Buisseret L, Powers K, Dorling J, Bradshaw L, Johnson S, Mitchell E, Duley L. Randomised trial of cord clamping at very preterm birth: outcomes at 2 years [published online ahead of print August 1 2019]. Arch Dis Child Fetal Neonatal Ed 2019. See Appendix 23.

Bradshaw LE, Pushpa-Rajah A, Dorling J, Mitchell E, Duley L. On behalf of the ‘Cord Pilot Trial Collaborative Group’. Cord pilot trial: update to randomised trial protocol. Trials 2015;16:407.

Pushpa-Rajah A, Bradshaw L, Dorling J, Gyte G, Mitchell EJ, Thornton J, Duley L, Cord Pilot Trial Collaborative Group. Cord pilot trial – immediate versus deferred cord clamping for very preterm birth (before 32 weeks gestation): study protocol for a randomized controlled trial. Trials 2014;15:258.

Batey N, Yoxall CW, Fawke JA, Duley L, Dorling J. Fifteen-minute consultation: stabilisation of the high-risk newborn infant beside the mother. Arch Dis Child Educ Pract Ed 2017;102:235–8.

Work package 5

Duley L, Askie L, Montgomery A, Hunter K, Barba A, Seidler AL, et al. Systematic review and network meta-analysis with individual participant data on Cord Management at Preterm Birth (iCoMP): study protocol. Draft prepared for submission. See Appendix 25.