Investigation report: Emergency neonatal blood transfusion at birth following acute blood loss during labour and/or delivery

This investigation looks at the issue of emergency blood transfusions given to newborn babies who need resuscitation when they are born. If a baby has lost blood before or during birth, efforts to resuscitate them may be less effective because they may not have enough blood to carry the oxygen their body needs. Delays in the administration of a blood transfusion in this scenario can therefore result in brain injury caused by lack of oxygen to the baby’s brain.

This investigation explores the issues that influence the timely administration of blood transfusion to newborn babies following acute blood loss during labour and/or delivery. Specifically, it explores:

• communication between the different medical teams involved in the care of women/pregnant people and their babies during labour and birth
• national guidance for medical staff on when to consider the option of a blood transfusion for a newborn baby.

As an example, which is referred to as ‘the reference event’, the investigation examined the experience of Alex and Robert, whose baby Aria was born by emergency caesarean section following an acute blood loss.

The investigation’s findings and safety recommendations aim to prompt clinicians to consider the option of a blood transfusion at an early stage in the resuscitation of a newborn baby.

Alex and Robert went to the hospital where they were booked for maternity care after Alex’s waters broke at home. While at the hospital Alex experienced sudden unexpected blood loss and was transferred to the labour ward. Baby Aria’s heartbeat could not be found on a cardiotocograph (a machine that monitors a baby’s heartbeat and a woman’s or pregnant person’s contractions during labour), but an ultrasound scan identified a slow heartbeat. Alex was transferred to an operating theatre for an emergency caesarean section and baby Aria was delivered soon after. Baby Aria required resuscitation and was given a blood transfusion before being transferred to the neonatal (newborn baby) unit. Baby Aria sadly died when she was 2 days old.

Delays in neonatal blood transfusion had emerged as a safety risk from investigations carried out under HSIB’s maternity investigation programme. HSIB contacted the hospital Trust where the reference event occurred. The Trust welcomed HSIB’s involvement and collaborated with information gathering. After the initial information had been gathered and evaluated, HSIB’s Chief Investigator authorised a national safety investigation.

• The administration of a blood transfusion as part of resuscitation requires a number of preparatory steps, including collecting the blood and undertaking various checks before using it. Inclusion in resuscitation training of a prompt for clinicians to consider the need for a transfusion, and to prepare for it if appropriate, may help reduce any delay.
• Involving members of neonatal teams (staff who specialise in the care of newborn babies) in multidisciplinary training in maternity units is not routine. Standardising their inclusion in such training would promote a shared understanding of relevant clinical information and ways of working.

1.1.1 Vaginal bleeding that occurs after 24 weeks of pregnancy is known as an antepartum haemorrhage (APH) and is a complication in 3% to 5% of pregnancies (Royal College of Obstetricians and Gynaecologists, 2011).

1.1.2 This type of vaginal bleeding may be caused by blood loss from a fetus, with the blood entering the woman’s/pregnant person’s genital tract, or blood loss from the woman/pregnant person. Initially it may be difficult to be sure of the origin of the blood loss. Additionally, the cause of such bleeding may not always be known (Royal College of Obstetricians and Gynaecologists, 2011). An APH of unknown origin is said to account for more than half of instances of APH (Chan and To, 1999).

1.1.3 The Royal College of Obstetricians and Gynaecologists (2011) says that the most ‘important’ causes of APH are:

• placental abruption, where the placenta (afterbirth) separates from the wall of the woman’s or pregnant person’s uterus (womb) before the baby is born, and
• placenta praevia, where the placenta attaches low in the woman’s or pregnant person’s uterus and covers the cervix (neck of the womb).

1.1.4 In the patient safety event examined by this investigation (referred to as ‘the reference event’), it was noted after the birth that Alex’s placenta had what is called velamentous cord insertion (VCI). Usually, the umbilical cord inserts into the middle of the placenta as it develops. In VCI the cord inserts into the membranes (the sac containing amniotic fluid in which the fetus develops) and the cord vessels travel within the membranes to the placenta (see figure 1). This leaves exposed cord blood vessels which are vulnerable to rupture, which can cause the baby to lose significant amounts of blood before birth. Given that a newborn baby born at full term will have 85ml/kg to 90ml/kg in their body (referred to as their ‘circulating blood volume’) (Hazinski, 2013), losing even a relatively small amount of blood can cause a baby’s health to deteriorate significantly.

Figure 1 Normal and velamentous cord insertion

1.1.5 A VCI is considered to be present in approximately 1% of singleton pregnancies (that is, pregnancies with one fetus rather than twins or triplets, for example) (Bohîlțea et al, 2016) and is strongly associated with vasa praevia (Buchanan-Hughes et al, 2020).

1.1.6 Vasa praevia is where blood vessels travelling from a baby to the placenta, unprotected by placental tissue or the umbilical cord, pass near to the entrance of the womb and are vulnerable to tearing during labour. In the reference event, following the birth of baby Aria, Alex was found to have a VCI and was suspected to also have had vasa praevia.

1.1.7 Vasa praevia is a very rare condition affecting between 1 in 1,200 and 1 in 5,000 pregnancies (Royal College of Obstetricians and Gynaecologists, 2018). Although VCI does not directly cause vasa praevia, the UK National Screening Committee, which advises ministers and the NHS on all aspects of population health screening, states it ’could be considered a marker’ due to the strong association between the two (UK National Screening Committee, 2017).

1.1.8 Many cases of VCI are diagnosed after birth, having had no impact on the woman/pregnant person or baby.

1.1.9 Occasionally vasa praevia can rupture. This is where vessels tear, causing a baby to lose blood and their condition to deteriorate rapidly. The death rate for babies in this scenario is at least 60% despite birth by emergency caesarean section (Royal College of Obstetricians and Gynaecologists, 2018).

1.1.10 Routine antenatal ultrasound screening for cord insertion into the placenta and vasa praevia is not currently part of the NHS fetal anomaly screening programme (which offers scans to all women/pregnant people to check on their baby’s development). The screening programme does require visualisation of the umbilical cord insertion into the baby’s abdomen (Public Health England, 2018).

1.1.11 In 2013 and 2017 the UK National Screening Committee considered screening for vasa praevia in the second trimester, as part of the fetal anomaly ultrasound scan. It did not find sufficient justification to include this as part of the fetal anomaly ultrasound scan. However, at the time of writing this report the committee was due to consider this issue again as part of its routine periodic review of evidence.

1.2.1 A blood transfusion is the process of giving a patient blood (that has been donated by another person) to replace blood they have lost, or to replace the liquid or cells found in blood.

1.2.2 For a planned (rather than emergency) blood transfusion, a sample of the patient’s blood is taken and analysed to ensure that that the replacement blood matches that of the patient.

1.2.3 This report focuses on emergency neonatal blood transfusion – that is, giving blood to a newborn baby – during resuscitation. In this report, ‘resuscitation’ refers to efforts to support the newborn baby’s airway, breathing and circulation, including the need for chest compressions. When a transfusion is given in these circumstances, blood is given which is compatible with all blood types.

1.2.4 The investigation was unable to find nationally published data about the number of incidences of blood transfusion being required for a newborn baby. One retrospective review (Wyckoff et al, 2005) found that, within the group studied, the incidence of cases where babies born at more than 34 weeks’ gestation received a transfusion in the delivery room to replace lost blood was 0.04%. That said, stakeholders who engaged with the investigation were unanimous in their view that it is a rare event, but one which can have significant consequences for the baby. Perhaps reflecting this view, the investigation is aware that a European Neonatal Transfusion Network is currently being set up to explore safety issues related to neonatal blood transfusion.

2.1.1 Alex, aged 32 years, booked for maternity care at 9 weeks’ gestation (9+0 weeks) in her second pregnancy. No pregnancy risks were identified.

2.1.2 Alex experienced a vaginal bleed at 13 weeks which went away on its own after about a week.

2.1.3 At Alex’s routine anomaly ultrasound scan at 20+3 weeks a low-lying placenta was identified. A further ultrasound scan and a follow-up appointment for the results were arranged for 32+3 weeks. At this appointment, it was noted that ‘the leading edge of the placenta cannot be visualised [on the ultrasound scan] … no evidence of placenta within 44mm of the internal os [neck of the womb]’. In essence, the placenta was no longer low lying so was not a cause for concern. A further scan and clinic appointment were arranged for 5 weeks later.

2.1.4 At 36+2 weeks Alex awoke to find that she was bleeding from her vagina. She travelled to the maternity assessment centre where she was seen by a midwife. The bleeding had stopped, and baby Aria had been moving less since the bleed occurred. Alex’s contractions and baby Aria’s heart rate were monitored using a cardiotocograph (CTG). This is a machine which provides a visual recording of the fetal heartbeat and the regularity of the woman’s or pregnant person’s uterine contractions.

2.1.5 An obstetrician reviewed Alex and an internal examination was performed which identified that there was no ongoing bleeding and that birth was not imminent. An ultrasound scan was performed which did not indicate any concerns about baby Aria and confirmed the placenta was in the expected position. Alex was moved to the labour ward for a further period of CTG monitoring and the documented plan was ‘+/- delivery if any further concerns’. Consent was obtained from Alex for a caesarean section.

2.1.6 While on the labour ward Alex experienced uterine contractions and medical staff discussed with her and Robert (baby Aria’s father) the possibility of a caesarean section. Reviews undertaken by obstetricians, the CTG monitoring and clinical markers (such as heart rate and blood pressure) did not suggest any further concerns. Therefore, the obstetric team decided that there were no immediate medical reasons to deliver baby Aria by caesarean section.

2.1.7 Alex’s contractions ceased. Following review by an obstetric consultant, a plan was made to discontinue the CTG, but for Alex to remain as an inpatient for observation and CTG monitoring for 24 hours after the bleeding stopped in case a further bleed occurred. This decision surprised Alex and Robert as they were not aware that the obstetric team had decided against a caesarean section at this time. They recalled staff on the labour ward telling them “you will meet your baby in 20 minutes”, rather than the caesarean section being a possible course of action that depended on the CTG, clinical markers or further bleeding.

2.1.8 At the next obstetric consultant review, approximately 4 hours after the previous one, the decision was made to discharge Alex home. Alex had had no further bleeding and all clinical markers (including observations such as blood pressure and temperature and blood test results) were within the expected ranges. Alex had a prearranged clinic appointment and ultrasound scan in 7 days’ time.

2.1.9 Alex’s waters broke at home that evening (36+3 weeks) and she was readmitted to the maternity assessment unit in the hospital. A CTG was started and an obstetrician examined Alex, confirming that her waters had broken.

2.1.10 While walking to the toilet Alex felt baby Aria making big movements and noticed heavy blood loss from her vagina, seeing drops of blood on the floor. She called for help and the staff member who arrived arranged for Alex to be taken immediately to the labour ward. She was taken there in a wheelchair in view of the ongoing blood loss.

2.1.11 On arrival on the labour ward a midwife attempted to locate baby Aria’s heartbeat using a CTG machine. The heartbeat could not be found by the Midwife or their colleague and, after momentary confusion about which bell should be used to summon assistance, the emergency bell was activated.

2.1.12 Members of the midwifery and obstetric teams attended immediately. Further attempts to find baby Aria’s heartbeat using the CTG were unsuccessful, but her heartbeat was identified on an ultrasound scan. Baby Aria’s heart rate was significantly below the expected range at around 80 beats per minute (bpm) (expected range 110bpm to 160bpm). A vaginal examination showed that Alex’s cervix was 2cm dilated, indicating that birth was not imminent. A decision was therefore made to move Alex to the obstetric operating theatre for an emergency caesarean section.

2.1.13 Baby Aria’s heart rate was again identified on an ultrasound scan in the operating theatre and was documented to be ‘very faint’.

2.1.14 The paediatric team, including a paediatric registrar (junior doctor undertaking specialist training in paediatrics) arrived just before the caesarean section took place. They were given a handover by the staff in attendance, in particular that Alex had had a significant antepartum haemorrhage (APH) of 300ml to 500ml and that baby Aria’s heartbeat could not be identified on the CTG. Importantly, they were not told that baby Aria’s heart rate had been identified on an ultrasound scan prior to the caesarean section.

2.1.15 Baby Aria was born 12 minutes after the decision to proceed with a caesarean section and 20 minutes after the blood loss was identified – times which are within national recommended timescales. Members of the paediatric team were present for the birth and had called for support from the neonatal nurses in anticipation of baby Aria requiring significant resuscitation.

2.1.16 At birth, baby Aria had no heartbeat, was making no respiratory effort (effort to breathe) and was described as ‘white’ in colour. She was transferred immediately to the resuscitaire (a piece of equipment which contains the components for managing emergencies, including resuscitation) and the care of the awaiting paediatric team.

2.1.17 Staff performed suction of baby Aria’s airway and attempted to stimulate baby Aria with a towel for 1 minute. They noted that there was no response.

2.1.18 Due to baby Aria’s condition, midwifery staff tried to contact the on-call consultant paediatrician (Consultant Paediatrician 1) using the hospital bleep system (see 4.1.5). The investigation found that this call was made when baby Aria was 1 minute old. When there was no response to several bleeps, a member of the midwifery team went to the neonatal unit to try and locate the Consultant, who they knew was in the hospital. In addition, the Paediatric Registrar telephoned Consultant Paediatrician 1 on their mobile phone and spoke with them, asking them to come to the operating theatre immediately. The off-site, on-call paediatric consultant (Consultant Paediatrician 2) was also contacted during this time and was asked to come to the hospital.

2.1.19 The placenta was delivered. No blood was present in the umbilical cord so cord blood samples could not be taken for testing.

2.1.20 Despite the efforts to resuscitate her, baby Aria initially showed no signs of life. Baby Aria was intubated (a tube was inserted into her airway to help her breathe) at around 3 minutes old.

2.1.21 When Consultant Paediatrician 1 arrived in the obstetric operating theatre they assessed baby Aria while resuscitation continued.

2.1.22 Following further resuscitation efforts and discussion, emergency neonatal blood was requested and baby Aria was given a blood transfusion via an intraosseous line (which involves inserting a needle into a bone to allow the delivery of medication or fluids directly into the bloodstream). This took place when baby Aria was 23 minutes old. Baby Aria’s colour was noted to change to pale pink on the administration of the blood. Consultant Paediatrician 2, who was on the way to the hospital, contacted Consultant Paediatrician 1 who gave them details of the situation.

2.1.23 At 30 minutes old baby Aria’s heartbeat had not been heard but support for her breathing was ongoing as she was making some efforts to breathe.

2.1.24 When Consultant Paediatrician 2 arrived (when baby Aria was 31 minutes old) they were informed that staff had not yet heard a heartbeat and there was a discussion about whether it was appropriate to continue resuscitation efforts.

2.1.25 A faint heartbeat was heard for the first time since baby Aria’s birth by Consultant Paediatrician 2. Because of this, resuscitation continued, with a plan to stabilise baby Aria’s condition and transfer her to the special care unit (SCU).

2.1.26 Once baby Aria was breathing on her own and her heart rate was stable she was transferred to the SCU. The documented plan was for baby Aria to be taken to the regional neonatal intensive care unit (NICU) where more specialised care was available.

2.1.27 Robert saw baby Aria for the first time in the SCU. He told the investigation that she was “as white as a sheet”.

2.1.28 When she arrived at the SCU, baby Aria again lost colour and her heartbeat was described to the investigation as “almost inaudible”. Efforts to resuscitate her began again, and further blood transfusions were arranged.

2.1.29 In advance of baby Aria’s transfer to the NICU, therapeutic cooling (a procedure where a baby is cooled with the aim of preventing further brain injury following a lack of oxygen) was initiated, and the target body temperature of 33.5C was reached when she was 2 hours old.

2.1.30 Baby Aria remained in an unstable condition when the regional neonatal specialist team arrived. When her condition was stable enough she was moved into the transport incubator and transferred to the regional NICU.

2.1.31 At the NICU staff noted that baby Aria was showing signs of multi organ failure due to hypoxic ischaemic encephalopathy (brain injury caused by an interrupted supply of oxygen to her brain). Baby Aria’s condition continued to deteriorate. Due to her poor prognosis and following discussions with Alex and Robert, care was redirected towards keeping her comfortable. Baby Aria died when she was 2 days old.

2.2.1 For resuscitation efforts to have an optimum effect, a baby’s airway/breathing and cardiac (heart) function need to be managed effectively, and they need an adequate circulating blood volume. Although not established at the time, the blood loss Alex experienced was due to an acute fetal blood loss that occurred before baby Aria’s birth. This resulted in baby Aria having a significantly depleted circulating blood volume. Resuscitation, therefore, would not have optimum effect until her circulating blood volume had been corrected through a blood transfusion. This was administered when baby Aria was 23 minutes old. A number of factors contributed to the blood transfusion not being administered earlier (see 4.1).

2.2.2 The investigation asked subject matter advisors (two consultant neonatologists) for their opinion on the impact of the blood transfusion not being given earlier. They told the investigation that, in their view, it was not possible to know whether earlier blood transfusion would have changed the outcome for baby Aria. They thought it was unlikely but reiterated that it is impossible to know for certain. In addition, representatives from the British Association of Perinatal Medicine said that in their view the delay in administering blood was no more than 5 minutes and this did not significantly contribute to baby Aria’s sad death.

2.2.3 Irrespective of whether the outcome was altered for baby Aria, the investigation acknowledges the psychological impact on Alex and Robert of not knowing whether their daughter was given blood at the earliest opportunity.

3.1.1 HSIB was notified of a patient safety incident relating to the death of a newborn baby. The notification was made by two NHS trusts, and Alex and Robert, the mother and father of baby Aria.

3.1.2 HSIB’s Chief Investigator authorised a scoping investigation.

3.1.3 The purpose of scoping investigations is to explore the identified patient safety risk(s), and to consider the practicality of, and opportunities for learning afforded by, a national investigation.

3.2.1 The Chief Investigator authorised a national investigation after assessing the findings of the scoping investigation against HSIB’s criteria for investigation (see below).

Outcome impact – what was, or is, the impact of the safety issue on people and services across the healthcare system?

3.2.2 Delay in initiating an emergency neonatal blood transfusion at birth can cause significant harm and death.

Systemic risk – how widespread and how common a safety issue is this across the healthcare system?

3.2.3 There is limited information on the prevalence, and administration of, emergency neonatal blood transfusions. The HSIB maternity investigation programme investigated 1,719 incidents between April 2018 and April 2021. Of these, 39 cases involved neonatal blood transfusions at birth. In 22 of the 39 cases there were concerns about the recognition of the need for, or the initiation of, emergency neonatal blood transfusion, making this a theme. Thus, emergency neonatal blood transfusions are rare during resuscitation, but the consequences of a blood transfusion not being undertaken at the earliest opportunity can be life threatening, making it a significant safety risk.

Learning potential – what is the potential for an HSIB investigation to lead to positive changes and improvements to patient safety across the healthcare system?

3.2.4 Given the very limited research literature on this safety issue there is the potential for an HSIB investigation to raise awareness and to make safety recommendations for improvement in this area of care.

3.2.5 An HSIB investigation provides an opportunity to share insights and learning from stakeholders on this safety issue.

After initial evidence gathering, it was agreed that the terms of reference were to:

1. Understand the context and contributory factors which influence the timely administration of emergency neonatal blood transfusion at birth.
2. Identify the environmental, human and other system-wide factors that help or hinder team working and communication when a baby’s symptoms suggest that an emergency neonatal blood transfusion is needed.
3. Identify the implications of the findings and whether opportunities exist to promote the early recognition and administration of emergency neonatal blood transfusion when required.

3.4.1 Multiple sources of evidence were gathered and analysed by the investigation, including:

• review of clinical records
• interviews and conversations (via video call) with Alex and Robert
• interviews with 17 staff members either directly or indirectly involved in the reference event, including paediatric, obstetric and midwifery staff
• review of national guidance.

3.4.2 Stakeholders and subject matter advisors informed the wider, national investigation. They provided authoritative comment on system-wide issues that impact on emergency neonatal blood transfusions at birth. Stakeholders and subject matter advisors included:

• Resuscitation Council (UK) – Newborn Life Support subcommittee
• Practical Obstetric Multi-Professional Training (PROMPT)
• NHS Resolution – Early Notification team
• British Association of Perinatal Medicine (BAPM)
• consultant neonatologist clinical advisors for HSIB.

3.5.1 The investigation used the Systems Engineering Initiative for Patient Safety (SEIPS) (see figure 2) to analyse the evidence gathered in the scoping investigation (Holden et al, 2013; Carayon et al, 2006). SEIPS provides a framework for understanding the work system (that is, the external environment, organisation, internal environment, tools and technology, tasks, and persons), work processes (including physical, cognitive and social/behavioural aspects) and the relationship between these and the resulting outcomes in healthcare.

4.1.1 Neonatal care in the hospital trust where the reference event took place (referred to in this report as ‘the Trust’) is provided by paediatricians and neonatal nurses. The paediatric team first became aware of baby Aria when the emergency call for a caesarean section was made.

4.1.2 The paediatric team was provided with verbal information by the midwives present.

4.1.3 The Paediatric Registrar who first attended the operating theatre said that they were informed there had been a significant antepartum haemorrhage (APH) of 300ml to 500ml and that baby Aria’s heartbeat “could not be found”. However, the Paediatric Registrar said they were not told that a heartbeat had been identified on the ultrasound scan moments before their arrival. They said this significantly contributed to their understanding of baby Aria’s condition, which indicated that they were attending the birth of a baby that was likely to be stillborn.

4.1.4 The context of this handover may have contributed to all relevant information not being communicated. The Paediatric Registrar described the atmosphere as “anxious”. In addition, there was a “rush to deliver the baby”. They described midwifery staff who were trying to locate baby Aria’s heartbeat shaking their head, indicating that they were unable to do this. This reinforced the paediatrician’s perception that baby Aria had likely died in utero (in the womb) and before birth.

4.1.5 In line with usual practice, a consultant paediatrician was on call for neonatal emergencies on the labour ward, the neonatal unit, and for paediatric emergencies on the children’s ward, assessment unit, and the emergency department. Due to the distance this consultant (Consultant Paediatrician 1) lived from the hospital, they stayed on site, in a room on the neonatal unit, when providing on-call cover. In addition, due to the COVID-19 pandemic, the Trust had arranged for a second consultant paediatrician to be on call (Consultant Paediatrician 2). This was to provide cover because the primary on-call consultant was clinically vulnerable to COVID-19 and therefore unable to see patients who had, or were suspected to have, COVID-19.

4.1.6 Staff told the investigation that as soon as baby Aria was born the Paediatric Registrar leading the resuscitation asked for Consultant Paediatrician 1 to be contacted as baby Aria was very unwell. They did not specify the method of contact so midwifery staff used the bleep system in line with their usual practice. Switchboard records reviewed by the investigation show that the bleep system was used to send an emergency bleep to the paediatric team, which midwifery staff believed included Consultant Paediatrician 1, for a neonatal emergency.

4.1.7 In addition to the emergency bleep via switchboard, staff recalled “bleeping the consultant” several times without a response. The hospital switchboard provides an audit trail of emergency bleeps only. The investigation found evidence of one emergency bleep at this time.

4.1.8 As there had been no response from Consultant Paediatrician 1, midwifery staff went to the neonatal unit in person to find them.

4.1.9 The paediatric consultant bleep is held between 09:00 hours and 17:00 hours. After this time contact with the consultant is via their mobile telephone. This practice contrasts with that in the obstetric team where contact with, and escalation to, the obstetric consultant is via a bleep during the day and overnight.

4.1.10 The investigation was told that the different mode of contacting the paediatric consultant overnight was well known within the paediatric and wider neonatal teams. However, the investigation found that this practice was not well known to the midwifery staff, who do not have direct access to the paediatric rota or contact telephone numbers for the paediatric consultants.

4.1.11 The investigation found that training for emergency situations and newborn life support, which would include how to contact a consultant out of hours, is not multidisciplinary. That is, the paediatric team undertakes this training separately from the maternity team (midwifery, obstetric, anaesthetic and associated roles) and vice versa. Thus, there is not an opportunity for these teams to rehearse together scenarios such as the reference event which would necessarily include simulated communications, such as out-of-hours contact with consultants from the different specialities. The investigation found that this may have contributed to the lack of shared awareness of the different mode of contact for the paediatric consultant overnight.

4.1.12 Evidence gathered by the investigation indicated that the different modes of contacting the teams out of hours created a safety risk. A standardised process for contacting consultants out of hours, across the obstetric and paediatric teams, may help to minimise this risk. With regard to using the bleep system for such contact, the investigation was told that the paediatric consultants are not required to be on site overnight (whereas the obstetric and anaesthetic teams are) so this is not possible as the bleep system will not work outside of the hospital. Changing to the use of mobile telephones to contact the obstetric team out of hours could potentially introduce new risks. For example, with the current system of an emergency bleep the consultant and other team members are summoned, providing a safety net that would not exist if an individual’s mobile number were to be used.

4.1.13 Stakeholders noted that issues regarding out-of-hours communication with the on-call consultant are not unique to this Trust. They raised different options to mitigate the risk, one of which was to include consultant emergency contact details on all resuscitaires and telephones on the labour ward.

4.1.14 In addition to midwifery staff going to the neonatal unit to summon the on-call paediatric consultant (Paediatric Consultant 1), the Paediatric Registrar involved in the resuscitation of baby Aria contacted the consultant directly on their mobile telephone.

4.1.15 Telephone records show that the Paediatric Registrar’s call to Consultant Paediatrician 1 was received 7 minutes after baby Aria was born. This was the first contact that the Consultant had received in relation to the emergency. They immediately went to the operating theatre and it is documented in maternity notes that they arrived 5 minutes after receiving the telephone call.

4.1.16 While some minutes were lost due to the incorrect process being used by the midwifery team to contact Consultant Paediatrician 1, the investigation was mindful that this Consultant Paediatrician was only on site due to the distance they lived from the hospital. Local policy and practice sets an expected attendance time of 30 minutes, and they arrived well within that time.

4.1.17 When baby Aria was transferred to the resuscitaire after the caesarean section, staff present told the investigation that she was “white”, was making no efforts to breathe and did not have a heartbeat. The Paediatric Registrar said that when they saw baby Aria, it confirmed their understanding and expectation that she had sadly died in utero.

4.1.18 The Paediatric Registrar stated that resuscitation was started pending the arrival of the Consultant; their expectation at this point was that these efforts were unlikely to be successful.

4.1.19 The resuscitation followed the newborn life support (NLS) algorithm, created by the Resuscitation Council (UK). This is the nationally recognised systematic approach used in the resuscitation of newborn babies (see appendix 1).

4.1.20 The Paediatric Registrar said they recognised that baby Aria had most likely suffered a brain injury caused by lack of oxygen due to her blood loss. However, they said they did not request a blood transfusion at that time because their perception was that this more invasive resuscitation effort was not appropriate in the context of baby Aria having no identified heartbeat prior to the caesarean section, which was the information they had been given at handover (see 4.1.3).

4.1.21 Subject matter advisors told the investigation that babies whose hearts have stopped beating will “look white” – as there will be no blood circulating around their bodies. In the scenario of a “white” baby (or one with significantly paler skin colour than expected) with a history of a significant APH, early consideration of a blood transfusion is appropriate. They pointed out that a baby must have an adequate circulating volume of blood for resuscitation attempts to have a chance of success. Conversely, having an adequate circulating volume of blood is only beneficial if the blood is being effectively oxygenated. The Chair of the Newborn Life Support subcommittee of the Resuscitation Council (UK), who is a consultant neonatologist, said that hypoxia (lack of oxygen) is the most common reason for a ‘white’ baby and addressing airway and breathing to ensure effective oxygenation is the first priority in a neonatal resuscitation.

4.1.22 Subject matter advisors highlighted other contextual factors. For example, it can be difficult to hear a baby’s heartbeat and in this case the Paediatric Registrar was not aware that baby Aria’s heartbeat had been identified just before the caesarean section. This influenced their decision not to request a blood transfusion immediately. In addition, the subject matter advisors pointed out that the need for an emergency blood transfusion is a very rare scenario and one a registrar is likely to have very limited experience of. The Paediatric Registrar told the investigation that they had not experienced another scenario like this.

4.1.23 Subject matter advisors raised the broader system-wide issue that national guidance does not emphasise the potential need for neonatal blood transfusion following an acute blood loss from the vagina which, as in this case, may be due to acute fetal blood loss. In their view, strengthening of this point in the guidance could prompt clinicians to consider this.

4.1.24 When Consultant Paediatrician 1 arrived and appraised the situation they requested emergency neonatal blood (baby Aria was 17 minutes old at this time). Consultant Paediatrician 1 told the investigation that when they arrived, the outlook for baby Aria appeared to be poor. They said that they went through in their mind the actions taken so far and requested a blood transfusion as the one outstanding clinical intervention.

4.1.25 The investigation found that there was a dedicated emergency blood refrigerator on the labour ward so blood was readily available when needed in the operating theatres. Staff interviewed were aware of its location, had access to the store, and were familiar with the checks required in its use.

4.1.26 The Paediatric Registrar who led the initial resuscitation efforts, before the arrival of Consultant Paediatrician 1, requested a peripheral cannula to be sited (a thin tube inserted into a vein in a patient’s arm or hand). They explained that at this point they were focused on supporting baby Aria’s breathing and trying to contact Consultant Paediatrician 1. They wanted rapid access for the administration of resuscitation medication and so asked the junior doctor present to site a peripheral cannula. The Paediatric Registrar added that he could see a vein in baby Aria’s left hand that was amenable to a cannula being sited. As they had anticipated, the cannula was successfully inserted on the first attempt and adrenaline (a resuscitation medicine) was administered when baby Aria was 11 minutes old. The Resuscitation Council (UK) (2015) states that during the resuscitation of a severely unwell infant at birth an umbilical venous catheter (UVC) (a thin tube inserted into the vein in a baby’s cord) is ‘the primary method of access during resuscitation’.

4.1.27 Subject matter advisors told the investigation that early UVC access would have been the ideal given that baby Aria was unresponsive due to hypovolaemia (reduced blood volume). A UVC allows administration of medication in close proximity to a baby’s heart, enabling the medication to take effect rapidly. In contrast, medicines administered via a peripheral cannula take longer to have an effect, compounded in this scenario by baby Aria’s low circulating blood volume.

4.1.28 The investigation sought subject matter advisor opinion on the decision to site a peripheral cannula as opposed to a UVC. They said that siting a peripheral cannula is a far more common procedure than siting a UVC and thus, for the team present, was the easiest method of obtaining quick intravenous access to allow, in the first instance, administration of emergency resuscitation medication. The subject matter advisors explained that siting a UVC is a more complex technical task, and this was a time-pressured, urgent situation where speed of access (and therefore the administration of resuscitation medicine) was critical.

4.1.29 When Consultant Paediatrician 1 arrived, they were informed of baby Aria’s condition at birth and the lack of response to resuscitation efforts so far. Consideration was given as to whether it was in baby Aria’s best interest to continue resuscitation. Consultant Paediatrician 1 said they felt it was inappropriate to stop resuscitation at this point as not all possible interventions had been tried – in particular, administration of blood and UVC access to optimise the administration of resuscitation medication. They therefore requested both these actions. Obtaining UVC access was unsuccessful, so intraosseous access (the insertion of a needle into a bone) was attempted and achieved. Blood was administered via the intraosseous access when baby Aria was 23 minutes old.

4.1.30 Commenting on baby Aria’s care, the British Association of Perinatal Medicine said that diagnosing neonatal blood loss is recognised to be very difficult. In their opinion, to have given a transfusion within such a timeframe (23 minutes) and in such a stressful situation should be acknowledged as an achievement by the team.

4.1.31 Baby Aria had lost a significant amount of blood before birth so had a reduced circulating volume of blood. Baby Aria was given a blood transfusion at 23 minutes old.

4.1.32 The handover received by the Paediatric Registrar did not include the fact that baby Aria’s heartbeat had been identified on an ultrasound scan before her birth by caesarean section. This, together with other factors, influenced the Registrar’s belief that they were attending the birth of a likely stillborn baby.

4.1.33 Trust processes for escalation to a consultant at night differ between the paediatric and obstetric teams. Midwifery staff interviewed were unaware of the different processes.

4.1.34 Peripheral rather than UVC access was obtained to administer resuscitation medication. The limited experience of the team and the urgency of the situation contributed to this decision.

4.2.1 Examination of Alex’s placenta revealed the cause of the blood loss and confirmed that this loss was from baby Aria rather than Alex.

4.2.2 The Consultant Obstetrician told the investigation that when they delivered the placenta there were no signs of placental abruption, such as bleeding or clots, to explain the cause of the acute blood loss.

4.2.3 As is standard practice, the placenta was passed to a member of the midwifery team so that it could be examined. The placenta was examined in the operating theatre. Staff explained that the placenta is examined for two main reasons: to ensure the placenta is complete, so that an obstetrician can complete the caesarean section; and to take samples of cord blood for testing. Staff told the investigation that an obstetrician will commonly ask “is the placenta complete?” and then ask for cord blood samples to be taken for testing. The Midwife who took the placenta said that from their recollection they were asked this question and they confirmed that the placenta was complete. The Midwife said that they told the Consultant Obstetrician that a blood sample could not be taken. There can be various reasons why it may be difficult to obtain a sample, but a discussion did not take place at this time about this.

4.2.4 The Midwife told the investigation that when they examined the placenta they recognised a velamentous cord insertion (see 1.1.4) and saw ruptured vessels potentially indicative of vasa praevia (see 1.1.6).

4.2.5 The Midwife explained that they were communicating with the Consultant Obstetrician at this time, but the immediate focus of attention was completing the caesarean section and their dialogue related to this. Thus, the details regarding the placenta seemed immaterial to the task in hand at that time. However, the Midwife stated that they informed the Senior Midwife present of their findings as they were unusual.

4.2.6 The Senior Midwife did not recall this conversation. Their recollection was that they examined the placenta just before the caesarean section was complete and saw the velamentous insertion and large ruptured vessel. They said that they alerted the Obstetric Consultant and Registrar to these findings, taking them to see the placenta once the caesarean section was complete.

4.2.7 The Consultant Obstetrician told the investigation that they initially understood the placenta to be “normal”; they did not recall being alerted to abnormal findings. They said that after the caesarean section was completed they examined the placenta to look for clues as to the cause of baby Aria’s condition and saw the velamentous cord insertion and ruptured vessels. They immediately told the paediatric team members, who were caring for baby Aria on the special care unit by that time. A paediatric consultant involved with baby Aria’s care said that when they heard about the placenta it explained the cause of her blood loss and made sense of what had happened.

4.2.8 The communication of abnormal placental findings was identified as a safety issue, which was explored further in the wider investigation (see section 5).

4.2.9 Examination of the placenta revealed that a ruptured velamentous cord insertion was the cause of the acute blood loss.

4.2.10 The placental findings were not communicated to the Consultant Obstetrician as the focus of attention was on the completion of the caesarean section. They were also not communicated to the paediatric team while baby Aria was in the operating theatre. This delayed an understanding of the cause of baby Aria being so unwell at birth.

5.1.1 In the reference event there were three occasions where the communication between teams was not effective:

• The handover given to the paediatric team did not result in an awareness that baby Aria’s heartbeat had been identified on an ultrasound scan immediately before the caesarean section.
• The midwifery team did not know how to contact the paediatric consultant out of hours.
• Findings from the initial examination of the placenta were not given to the paediatric team at the earliest opportunity.

The investigation found that the maternity team (obstetric, anaesthetic, midwifery and associated roles such as maternity support workers) and paediatric teams did not train together. Thus, there was no opportunity for the teams to practise how they would communicate with each other in simulated emergency scenarios. The investigation considered that this may have contributed to the lack of shared awareness of out-of-hours contact procedure and the relevance of all available clinical information being shared between teams.

5.1.2 The ‘Saving Babies’ Lives version two’ care bundle (NHS England, 2019), which sets out best care practice with the aim of reducing stillbirths and deaths of newborn babies, made a number of recommendations. One was for ‘all staff to undertake multidisciplinary training’ because of the benefits this brings to team working and shared knowledge and understanding of roles. The need for multidisciplinary team training involving the maternity, anaesthetic and neonatal/paediatric teams has been recognised by NHS Resolution, which stated that ‘Efficient, skilled team working can also be supported by inclusion of neonatal teams in relevant maternity emergency simulation training and skills drills’ (NHS Resolution, 2019).

5.1.3 Similarly, the interim report of the Ockenden review of maternity services at the Shrewsbury and Telford Hospital NHS Trust (Ockenden, 2020) details seven ‘Immediate and Essential Actions’ to help improve safety in maternity services across England. Essential Action 3 states that ‘Staff who work together must train together’.

5.1.4 The actions from the Ockenden report prompted the creation of a core competency framework for professionals working in maternity and neonatal services by the NHS Maternity Transformation Programme, in collaboration with other stakeholders.

5.1.5 This framework identifies ‘maternity emergencies and multi professional training’ as core requirements. It states that those involved in such training should be ‘midwifery, obstetrics, anaesthetics, neonates as well as representatives from medical and critical care specialists’.

5.1.6 HSIB has identified through its maternity investigation programme a common theme relating to the lack of multidisciplinary training across maternity and neonatal/paediatric teams. The issue persists despite the Ockenden report and the Maternity Transformation Programme’s competency framework.

5.1.7 Subject matter advisors explained that, in their experience, training involving both the maternity (obstetric, anaesthetic, midwifery and associated roles such as maternity support workers) and neonatal teams was not common. They told the investigation of a number of barriers or factors influencing this:

• the logistical difficulties of organising training at a time and location convenient to these teams
• competing clinical demands on both teams which take precedence over such training
• lack of dedicated time for training – training needs to be factored into rotas and reflected in staff resource
• traditionally separate training.

The subject matter advisors also echoed a point made by the Chair of the Newborn Life Support subcommittee of the Resuscitation Council (UK), who is a consultant neonatologist. They explained that scenarios are often “perceived as two separate events”. They said that there are training scenarios which are specific to obstetrics, such as a post-partum haemorrhage (when a woman/pregnant person bleeds after giving birth), in which the participation of neonatal team members would offer no benefit. However, there are scenarios which are both obstetric and neonatal emergencies and where advisors agreed that the teams training together would be of significant benefit. Reflective of this, a representative of the British Association of Perinatal Medicine and other stakeholders stated that it was vital to move away from the perspective of separate teams (obstetric, anaesthetic, neonatal, midwifery and associated roles such as maternity support workers and neonatal nurses) to the concept of one ‘perinatal team’ that is, all staff involved in the care of a woman/pregnant person and baby during labour and after the birth. The investigation noted that the concept of perinatal team working and perinatal team culture is a fundamental part of the British Association of Perinatal Medicine ‘s core strategy.

5.1.8 NHS Resolution administers a maternity incentive scheme to support the delivery of safer maternity care. The scheme, developed in partnership with national maternity safety champions, and the maternity incentive scheme Collaborative Advisory Group, financially rewards trusts that meet 10 safety actions designed to improve the delivery of best practice in maternity and neonatal services.

5.1.9 Guidance is provided annually on what trusts must do to comply with the scheme’s requirements. The year three guidance, when first published, included within it a requirement for ‘systems testing’ which stipulated:

‘Trusts should evidence that multi-professional system testing occurs in the clinical area at least once in the MIS [maternity incentive scheme] reporting period, and it is attended by the anaesthetic, maternity and neonatal teams. Risks and issues identified within the clinical environment are addressed accordingly.’
(NHS Resolution, 2021a)

This requirement was removed from the guidance in March 2021 when it was amended in light of the COVID-19 pandemic.

In August 2021 NHS Resolution published its maternity incentive scheme year four guidance.

5.1.10 This guidance acknowledges the competency framework and outlines what trusts must do to comply with it. In relation to multi-professional training, the guidance specifies the maternity staff that should attend. It states that ‘at least one of the four emergency scenarios should be conducted in the clinical area, ensuring full attendance from the relevant wider multi-professional team’ (NHS Resolution, 2021b). While the multi-professional team is defined in the core competency framework as including the neonatal team, the neonatal team is not named as part of the multi-professional team in the year four guidance. Disciplines mentioned are obstetric anaesthetists, operating theatre staff and maternity support workers. Planning for year five guidance will begin early in 2022.

5.1.11 HSIB acknowledges that it is common for neonatal and midwifery teams to train alongside each other in relation to newborn life support. However, there is a lack of multidisciplinary scenario-based training attended by both the maternity (obstetric, anaesthetic, midwifery and associated roles such as maternity support workers) and neonatal teams.

5.1.12 The investigation acknowledges that guidance and training are not strong interventions to effect safety improvements, as they are people-focused rather than system-focused. However, the maternity incentive scheme is a national programme with financial implications for trusts and is therefore influential. Thus, including by name the neonatal/paediatric teams within the multi-professional training element of the NHS Resolution guidance may encourage trusts to undertake combined maternity (obstetric, anaesthetic, midwifery and associated roles such as maternity support workers) and neonatal training and would follow the spirit of the core competency framework.

5.2.1 In the reference event, information received and initial assessment of baby Aria did not prompt consideration of a blood transfusion (see 4.1).

5.2.2 The newborn life support (NLS) algorithm in place at the time of the reference event did not mention the transfusion of blood during resuscitation. However, the Resuscitation Council (UK) guidelines (2015), which provide detailed information supporting the use of the algorithm, did include consideration of blood transfusion within the section concerning the use of resuscitation medication.

5.2.3 Staff involved in the reference event, as well as subject matter advisors, told the investigation that it would be helpful if the NLS algorithm included early consideration of the need for an emergency blood transfusion in resuscitation scenarios. There was a consensus that such a prompt would be beneficial both because of the time needed to prepare for the administration of blood, and because it was a rare event so may not readily come to mind.

5.2.4 The Resuscitation Council (UK) released updated guidance on newborn life support in May 2021, together with an updated algorithm (see appendix 2). At the end of the updated algorithm there is a prompt to consider other causes of cardiac arrest, which states: ‘Consider other factors e.g. pneumothorax [collapsed lung], hypovolaemia [lack of circulating blood volume], congenital abnormality.’ (Resuscitation Council (UK) 2021)

5.2.5 The investigation asked the Chair of the Newborn Life Support subcommittee of the Resuscitation Council (UK) about the potential for including in the algorithm a prompt to clinicians to consider a blood transfusion in cases of hypovolaemia (blood loss) prior to birth. The rationale being that this treatment requires preparations which are time consuming relative to those required for the other possible causes listed above.

5.2.6 The Chair explained that the algorithm begins with directions on early resuscitation, based on assessments of airway, breathing and circulation (ABC). They said these considerations “take precedence”, prior to considering blood transfusion. They added that working through this section of the algorithm takes “about 10 to 12 minutes”, after which the algorithm prompts the clinician to consider other factors such as hypovolaemia, although a blood transfusion is not explicitly stated.

5.2.7 The Chair said that the algorithm was intentionally designed for “simplicity and clarity”. It is intended to cover the key principles and “headline messages” for effective resuscitation, in a format that was as brief as possible to promote ease of use in an emergency situation. They said that to include every consideration would inevitably make the algorithm a large and unwieldy document which was unlikely to be helpful in such scenarios.

5.2.8 The Chair explained that staff attending Resuscitation Council (UK) NLS courses receive practical and simulation training on the use and application of the NLS algorithm, which is supported by additional information in the NLS course manual. They stated that this training provides the detail and nuance that is not appropriate to include within the algorithm itself. The Chair agreed that neonatal resuscitation teams should be aware of fetal blood loss as a potentially reversible cause of a life-threatening deterioration in a newborn baby. They said that consideration of, and preparation for, a transfusion of O negative blood to the baby (the type used if a patient’s blood type is unknown) may help it be rapidly available if required. They added that a history of significant acute blood loss may help indicate when such a transfusion is required.

5.2.9 The Chair thought it would be possible, and they believed appropriate, for these points to form part of the aforementioned training which accompanies the algorithm.

5.2.10 They agreed that a blood transfusion requires much more preparation than the other reversible causes – the blood needs to be collected and various checks carried out prior to its release. Furthermore, the location and procedure for release of blood varies between trusts and it can be a time-consuming process. Given these preparations, the Chair said a prompt may well be helpful and this will be emphasised in the NLS course’s ‘Resuscitation at Birth’ lecture, a change which could be effected quickly.

5.2.11 The Chair said that the formal guidance for the algorithm was continually reviewed and updated although the formal manual was only reprinted every 5 years.

• The administration of a blood transfusion as part of resuscitation requires a number of preparatory steps including collection of the blood and various checks prior to its use. Inclusion in resuscitation training of a prompt to consider the need for transfusion, and to prepare if appropriate, may help reduce any delay.
• Involving neonatal teams in multidisciplinary training in maternity units is not routine. Standardising their inclusion would promote a shared understanding of clinical information of relevance and ways of working.