TEAMWORK: How a lack of teamwork at the Royal Cornwall Hospital contributed to the death of a child with autism

In memoriam  Coco Rose Bradford  2011-2017

This report can be downloaded in pdf format here.

Executive Summary

    • Coco Bradford, six years old, was admitted to the emergency department of the Royal Cornwall Hospital on July 25th, 2017, with a variety of symptoms. She was discharged home but taken back to Treliske the following day and admitted to the paediatric ward. On July 28th she was transferred to the adult Intensive Care Unit and then transferred to Bristol Royal Hospital for Children for paediatric intensive care. Tragically, she didn’t respond to treatment, and she died there on July 31st.
    • An independent investigation into Coco’s death was carried out by healthcare consultants Facere Melius. They documented the events that had taken place and identified a number of missed opportunities which might have averted this outcome. They made a number of recommendations that took the form ‘The [Royal Cornwall Hospitals] Trust should review …’, ‘The Trust should consider …’, ‘The Trust should ensure …’. But they said nothing about how these recommendations should be acted on – about organizational changes that were needed, for example.
    • This paper applies teamwork analysis to the information provided in the investigation report. It shows how failures in organization, communication and the culture of the paediatric department contributed to Coco’s death. It concludes by making a number of practical suggestions for eliminating such failures.
    • The paper may usefully be read alongside the investigation report. The bracketed numbers in the text are the numbers of paragraphs in the investigation report which are being referred to.

Coco’s story
Coco Bradford was clearly a remarkable little girl. From the age of two she uttered only disconnected words, but then astonished her family at breakfast two days before her fifth birthday by coming out with a complete, grammatically-correct sentence: ‘I want more toast, please.’ She had been diagnosed with autism, but was evidently not going to let it limit her life.[1] 

But at the age of six Coco’s life was cut short. On July 25th 2017 she was admitted to the emergency department of the Royal Cornwall Hospital at Treliske, near Truro, with a variety of symptoms. She was discharged home but returned to Treliske the following day, when she was admitted to the paediatric ward. On July 28th she was moved to the adult Intensive Care Unit and then transferred to Bristol Royal Hospital for Children for paediatric intensive care. Tragically, she didn’t respond to treatment, and she died there on July 31st.

Coco’s parents raised questions about the clinical decision-making processes and treatment provided to their daughter while she was in the care of the Royal Cornwall Hospital Trust (RCHT). In February 2018, the Trust’s medical director commissioned Facere Melius, a healthcare consultancy, to undertake an independent investigation with reference to NHS England’s Serious Incident Framework. Investigations under this framework are designed to support learning and prevent recurrence, not to assign blame. Facere Melius submitted their 91-page report on their investigation to the Trust in September 2018, and it was published on the RCHT website the following month.[2]

The investigation identified numerous failures in the treatment that Coco received at Treliske. It wasn’t recognised that she was clinically dehydrated and in clinical shock when she arrived at the emergency department; these conditions weren’t treated appropriately; and official guidelines weren’t followed while she was on the paediatric ward. Crucially, she didn’t receive the rehydration treatment that she should have been given. The abnormal results of tests weren’t identified and acted on, her blood pressure wasn’t obtained until nearly 36 hours after she was admitted to hospital, and there was a delay in starting antibiotics when her clinical picture suggested she had developed sepsis.

Faulty too was the assessment of the pain that Coco was in. On many occasions she was described as ‘distressed’, ‘inconsolable’ or ‘agitated’, but her pain score was recorded as zero, not having pain, or not recorded at all, and the standard pain assessment tool was not made use of. The hospital’s Learning Disability team was not called in to help in establishing a relationship with this autistic child who, according to Coco’s mother, was thought by some members of staff to be ‘uncooperative’ and ‘non-compliant’.[3]

The serious incident investigation and its limitations
Investigations under the serious incident framework are intended to support learning and prevent recurrence. What was learned in this case? The report identifies seven major failures in Coco’s treatment, presented as ‘main findings’ and ‘missed opportunities’. (Investigation report, pp 3-4) It identifies no fewer than 16 ‘opportunities for improvement’ (Investigation report, p. 62) and makes seven ‘key recommendations’. (Investigation report, pp 5-6)

Facere Melius say they used a technique known as ‘root cause analysis methodology’, but their report, while describing many instances of how members of staff interacted with one another, does not penetrate behind these observations to analyse the organization’s structure, processes and culture. Thus the investigation report contains no organizational chart which would show the positions of those involved, areas of responsibility, lines of authority and channels of communication. It uses the term ‘paediatric team’ very loosely, not even saying who its members were. The report makes just two references to the ‘paediatric department’, and gives no information as to who were members of this department, how it was organized, whether there was someone in post as its head.

A consequence of these omissions is that, as the report itself demonstrates, the investigators aren’t able to suggest to the Trust what organizational changes would be required to put their recommendations into effect. They had no alternative but to express their recommendations in very general terms: ‘The Trust should undertake a review …’, ‘The Trust should agree a policy …’ and – it is alarming that this should be thought necessary – ‘The Trust should ensure that all staff involved in the care of a patient should follow professional standards in relation to the documentation of clinical records.’ (Investigation report, pp 5-6)

These limitations of the investigation report leave a gap which this paper is intended to help fill.

Analysing teamwork: Terminology
The analysis presented in this paper makes use of a small battery of concepts which provide us with questions to ask about what took place in Coco’s case. It distinguishes between the core team of paediatric doctors and nurses who attended Coco, and the wider team, which comprises the core team and everyone outside that core who provided contextual information, specialist knowledge and judgment, and technical services, such as analysing blood samples. It notes the existence of a rulebook which is intended to govern some aspects of their roles and behaviour.

As in all hospitals, there exists what we can recognise as a structure comprised of both formal relationships and informal relationships. Within that structure we can identify those people who occupied named positions which denoted the formal roles that they were expected to play. And we can distinguish processes that took place: broadly speaking, a process is made up of a succession of events, occurrences, including the successive stages in the movement of an individual through a hospital from admission to discharge and stages that form part of procedures set out in the rulebook.

Within processes decisions are taken: these are acts of conscious choice made on the basis of information, which may be generated in various ways, and is received through a variety of communication channels, along with a mixture of prior knowledge, assumptions and inferences. (Sometimes, of course, events take place that are not attributable to acts of conscious choice.) And we may be able to distinguish different informal roles that the participants in processes have played.

Finally, all these activities take place within a culture, which we can take as made up of shared beliefs and conventions, customary ways of doing things. Some of these are enshrined in the rulebook.

Analysing teamwork: Questions to ask
The investigation report examined the actions of those involved but did not comment on their roles in any analytical way, so we have to interrogate the chronology, identifying the clues that we have and asking what we can infer from them.

The core team
Who were the members of the core team? At the heart of the core team we can place Coco herself. Albeit unwittingly, she provided information in the form of her behaviour and the symptoms that she exhibited. Attending her were paediatric staff: doctors – paediatric consultants (senior doctors) and registrars (junior doctors in the fifth and final year of their training) – paediatric nurses, responsible for caring for her nursing needs round the clock, and healthcare assistants.

The wider team
Who made up the wider team? Potentially significant members of the wider team were Coco’s parents, by virtue of being long familiar with her behaviour and responses to stressful situations. Paramedics in the ambulance service attended Coco on two occasions and brought her to Treliske. Within the hospital doctors and nurses in the emergency department played a role when Coco was first brought into the hospital, as did doctors in the adult Intensive Care Unit towards the end of her stay. The hospital had a learning disability team: they were a potential source of assistance with caring for Coco but were not called upon. Technicians at Treliske analysed blood samples, took X-rays, etc. Doctors at Bristol Royal Hospital for Children were consulted by those at Treliske before Coco’s transfer, and then cared for her until her death.

The rulebook
Rulebooks set out guidelines and ‘standard operating procedures’, ways of doing things in specified circumstances. In Coco’s case guidance set out by NICE (the National Institute for Health and Care Excellence) and RCHT itself were applicable. Rulebooks enable people to know what they are expected to do. So long as everyone conforms to the rulebook, they know where they stand, and also where other people stand, so again this should imply predictability. Thus abiding by the rulebook can play an important part in establishing good quality working relationships. So in examining team behaviour one of the penetrating questions we can ask is: Did people abide by the rulebook?

The organizational structure, formal and informal roles and relationships
Within the core team and the wider team, there exist, as in all hospitals, what we can recognise as an organizational structure, made up of formal relationships between people in certain designated roles. We need to have a picture of the relationships that existed between them. For example, senior doctors (consultants) will have ‘their’ juniors (such as registrars) ‘under’ them: the relationship between them is one in which seniors report on the capability and progress of their juniors, so the latter depend on the former for their career progress. What informal relationships existed alongside the formal ones? What positive informal relationships, such as mutual understandings and shared attitudes, that assist teamworking, can we observe? And do we see negative ones, such as hostility and suspicion, that are unhelpful to teamworking?

Processes and events, decisions, information and communication channels
The chronology of this case provides us with an overall process, comprising the events that took place during Coco’s movement through Treliske from admission to her transfer to Bristol. Significant ones with respect to teamwork are the decisions that were taken, and we need to ask what these were, and what information – in addition to prior knowledge, assumptions and inferences – was ‘fed into’ them. And did the decision taker have the requisite experience and expertise to assess the quality of that information and make a judgment as to how best to use it?

We can question the absence of action in a similar way. Was responsibility for taking a decision placed appropriately? Did the decision taker not have the experience and expertise required? Was there an absence or blockage of communication channels?

In a sense, a hospital runs on information. When a patient is admitted, it is the information gained at his or her initial assessment that determines which department they are sent to, or whether they are kept in for further observation (i.e. gaining further information), or are sent home. Patient records are crucial when it comes to forming a diagnosis and prescribing treatment, keeping diagnosis and treatment under review and monitoring the patient’s progress or lack of it. As we shall see, issues to do with information are particularly salient in Coco’s case.

We can take take ‘culture’ as made up of shared beliefs and conventions, customary ways of doing things. Some of these may be hospital-wide: for example, there appears to be a convention that between departments (and indeed between hospitals too) consultants speak to consultants, registrars speak to registrars, etc. And it will be the consultants who set the tone of a department, so we need to ask: What example were consultants setting to their juniors? Did they act as mentors or masters? Did they check everything that was done under their aegis?

What makes an effective team?
In a nutshell, an effective team will have the following characteristics:

Members of the core team will have rapport with one another, a shared understanding of and feelings about the job in hand and their respective responsibilities: these help them to communicate easily. While many teams have an official leader, with a job title denoting their status, if the circumstances are unusual or unprecedented, for a team to be effective the leadership role may have to be taken by others with more relevant experience and expertise.

Members of the wider team will have access to relevant members of the core team, who will be aware of their existence as potential contributors to progressing the job in hand, and will have open communication channels with them, so members of the wider team feel able to volunteer, as well as provide when asked, potentially useful information.

There will be a rulebook which sets out basic procedures to be followed in normal, non-exceptional cases, and core team members will be individually and collectively accountable for following them. It should be clear to them what the purposes of the rules are, and how far they can exercise individual discretion.

The formal structure will usually be in the background, so no-one will be standing on formality when there is a job, especially an urgent one, to be done. At such a time relationships within the core team may be informal but they will be purposeful.

Relationships are important because they significantly affect the communication of information and the quality (accuracy, reliability etc.) of that information. For example, a registrar who is intimidated by his or her consultant – a negative relationship – may be reluctant to question an action that the consultant proposes to take or point out a salient fact that has been overlooked. In an effective team there will be a positive relationship, one where the junior person is positively encouraged to ask questions of the senior, where the relationship between them is one of mentoring rather than command and control. This implies a level of predictability and the confidence that goes with that: these come only with experience so take time to develop.

Communication channels will be open and relevant information sought from all sources, but it will be very clear and understood by all whose responsibility it is to take decisions.

The culture of the team will be affected by the example set by its senior members, the consultants. In an effective team, they will have a practice of checking all decisions and actions that are taken; they will ensure that proper records are kept; they will plan the work of their team and ensure that stresses are shared equitably among members; they will mentor junior members and check that they are learning from their experiences; and they will care for newcomers and make sure they take on board what is expected of them.

The missed opportunities at Treliske
The Facere Melius report identifies a number of missed opportunities to restore Coco’s health. (Investigation report, p.3) Some are specifically to do with medical practice, but from the chronology we can identify five where teamwork could have made a difference to Coco.

A. The first time that Coco came to Treliske, the history and the absence of evidence of her tolerating the fluid challenge would suggest that she was at risk of becoming dehydrated if her symptoms continued. This should have led to a period of observation in hospital, when dehydration would have become apparent. (Investigation report, p.20)

The events When the ambulance delivered Coco at the Treliske emergency department on July 25th, she was seen by a paediatric nurse and reviewed by a locum registrar, Dr A. Coco was given a ‘fluid challenge’ to assess her ability to tolerate oral fluids. This is carried out by giving small amounts of fluid at regular intervals. The results are normally recorded on an input-output chart, but such a chart was not found within her clinical records. Dr A examined Coco and recorded that she tolerated a fluid challenge, although this was not her mother’s recollection: ‘We were given a fluid balance sheet to fill in, but every time Coco had even 10 mls it would come back up.’ (8.1.6/7/8/9) There was poor recording of clinical information, and the doctor’s entry was not dated or timed. Coco’s blood pressure was not taken, her pain score was not assessed, nor was she weighed. (8.1.10)

Discussion What we see here, at the ‘gateway’ to hospital care, is:
(a) Failure to collect information (blood pressure, pain score, weight);
(b) Failure to record information that was collected;
(c) Failure on Dr A’s part to date and time his entries;
(d) A disputed record (Coco’s tolerance of the fluid challenge);
(e) Records going missing (the fluid challenge chart).

Not only is this a remarkable catalogue of lapses from good practice: it demonstrates an absence of any sense on the part of the staff who saw Coco that they were potentially at the start of a care process, and that information that they collected would be needed to ‘build the clinical picture’ (Investigation report, p.28) and determine where Coco moved on to from the emergency department. Evidently there was no sense on their part of being part of a team of people to whom Coco’s care would be entrusted.

The poor recording of clinical information by Dr A raises questions about his training and experience, about the induction (if any) that he received at Treliske, and about the supervision (if any) that he was currently receiving. Effective team working depends on being able to assume that all members of the team are doing their job properly.

Dr A’s failures also raise further questions: Would not a consultant, a senior doctor, someone with more experience, be better equipped to examine and ‘signpost’ a patient who is brought in to the emergency department, or determine that further observation was needed? And who was it, presumably in some managerial role, who decided that this was an appropriate position for someone with Dr A’s limitations?

As for the discrepancy over the fluid challenge, this reminds us that parents can play a valuable role as monitors of their child’s progress and as checkers of any error or discrepancy in the data recorded. Coco’s mother could usefully have served as a member of the wider team here.

B. On Coco’s second visit to Treliske, on July 26th, it was not recognised that she was clinically dehydrated and in clinical shock when she arrived at the emergency department. (Investigation report, p.27)

The events On Coco’s second visit to Treliske, Dr B (presumably a registrar) was the first doctor to see her in the emergency department. Dr B ‘escalated’ her to a paediatric registrar, Dr C, with whom she had previously worked. If Dr B had followed normal practice, an emergency department senior doctor (consultant) would have reviewed Coco first. (8.2.5)

Three members of the nursing team in the emergency department saw Coco. She was triaged by a nurse and reviewed by a healthcare assistant, who ‘remembered Coco being pale and agitated but was unsure whether the agitation was due to her acute illness or whether it was due to her autism’. (8.2.2) A paediatric nurse also saw Coco and remembered her as being pale. However, she did not record this assessment. When questioned by the investigation team, she thought with hindsight Coco could have met the criteria of being clinically shocked. (8.2.3) The investigation team also noted that the main point of contact and the recorder in the clinical records was the healthcare assistant, not a nurse. (8.2.4)

Discussion Not referring Coco to an emergency department consultant is an instance of the rulebook not being followed. That rule has a purpose: to take advantage of the experience of senior emergency consultants, who see more patients requiring fluid resuscitation than do paediatric departments. Dr B was evidently unaware of the rulebook or for some reason chose not to follow it, despite Coco’s condition and her parents’ account of how she had been over the last 24 hours triggering a ‘red flag’ for her. (8.2.5)

The events in the emergency department that day are remarkable for the non-involvement of doctors. Three nurses, along with a paediatric nurse, saw Coco and noted her condition, but they did not bring their concerns to the attention of a doctor, and they left record-keeping to the healthcare assistant. They evidently had a very limited view of their responsibilities, which manifestly did not extend to working as a team with doctors.

Nurses clearly can make a contribution to diagnosis as well as to the care of patients. In Coco’s case nurses made valid observations of her paleness and agitated state but it appears that no doctor asked them for their opinion: evidently the culture of the emergency department was such that they were not regarded by doctors as members of the team, people with potentially useful observations to contribute.

C. ‘Coco’s clinical dehydration and clinical shock were not treated appropriately throughout the period she was on Polkerris ward. An inadequate fluid management plan was implemented which did not follow NICE guidance or RCHT paediatric department guidance. Abnormal blood and gas results were not identified and acted upon to build a clinical picture and inform her diagnosis and treatment plan. On several occasions there was no thorough review of her most up-to-date clinical observations, examination findings and blood results …’ (Investigation report, p.3)

The events Coco was admitted to Polkerris ward at 7.30 pm on July 26th. (8.3.2) Her first medical review on the ward was at 10 pm, when she was seen by the registrar, Dr P, who was also providing cover to other paediatric wards. (8.3.5)

Dr P prescribed an intravenous fluid infusion to replace fluids lost by dehydration, but the amount of fluid she prescribed was only half that specified by NICE and by the RCHT paediatric department’s own guidance. ‘Dr P said she was aware of NICE guidance but was following normal practice for the department.’ (8.4.5)

At 10.05 am on July 27th, Coco was given a fluid infusion which was recorded in the fluid prescription chart but not recorded in the clinical notes. (8.4.8) When Dr E, Coco’s named consultant, saw her for the first time on his morning ward round at 10.43 am, he appears not to have been aware that Coco had just received a fluid infusion and that it may have had an impact on how she presented clinically. Dr P did not enlighten him.

In that ward round, too, the results of Coco’s blood and blood gas samples were reviewed. The blood gases reviewed were the very latest ones, taken at 8.46 am that morning, but the blood results reviewed were from the previous day at 3.38 pm. This discrepancy was not picked up by Dr E, and Dr P did not enlighten him about the existence of an up-to-date blood sample. (8.4.11)

Dr E told the investigation team he could not recall knowing the blood gas result or the clinical observations at the time of his review, and he did not remember viewing the fluid prescription nor being aware of the way in which the intravenous infusion rate had been calculated. He agreed that the usual infusion used in dehydrated children would be 20 ml/kg but could not explain why 10 ml/kg had been used overnight, nor why the replacement fluid had been calculated at half the usual rate. (8.4.12/14)

Coco’s clinical observations were not reviewed in a graphical format where trends would have been visible. ‘Dr E said his normal practice would be to use the computer on the nursing station to look at the observations in a graphical format, but he couldn’t recall why he didn’t do so on this occasion.’ (8.4.15)

After the ward round Dr E discussed Coco’s condition with her parents and told them the most likely diagnosis was bacterial gastroenteritis, so he recommended holding off prescribing antibiotics at this stage. (8.4.13) He did not check the fluid prescription and the way in which the intravenous infusion rate had been calculated. (8.4.14) Nor did he look at the observations in a graphical format where trends in Coco’s condition would have been visible. (8.4.15)

Discussion Here we have a striking example of the rulebook not being followed, with the fluid infusion given being at only half the usual rate. Failure to record a fluid infusion in the clinical notes is a further illustration of poor practice. Neither of these infringements was noticed by Dr E or any other paediatric consultant.

There is a question to be asked here about clinical records and how they are kept. In the past, clinical observation charts in graph format were kept on each patient’s bed, so everyone – including children’s parents – could see them, and any deterioration or lack of observations could easily be seen. In effect, this enabled a ‘cross checking’ on information such as one would expect to find in an effective team. Has computerized record-keeping eliminated this safety mechanism?

These instances also raise questions about relationships between consultants and registrars in the paediatric department. Clearly Dr P’s departure from standard practice was not being monitored by her consultants. Clearly too there was not the ease of communication between Dr P and Dr E that is a prerequisite for good teamwork. If Dr P had realised that the recent fluid infusion could have influenced Coco’s appearance at the morning ward round, and if she had picked up on the discrepancy between the blood and blood gas results, as she clearly should have done, something was preventing her from pointing these out.

There are also questions to be asked about the culture of the department and the tone set by consultants. Did Dr E see himself as a mentor to Dr P, or was he oblivious to her training needs? And without making a judgment about Dr E’s memory or whether his mind was on the job, it is fair to say that it behoves a team member whose memory is not what it once had been to be punctilious about keeping a visible-to-all record of their actions and the reasoning behind them.

Dr E’s behaviour was manifestly not that of an experienced mentor and team leader.

Moreover, to judge by his interaction with Coco’s parents after the ward round, he did not take the trouble to check the fluid infusion figures or the trends in her condition before delivering his opinion that the most likely diagnosis was bacterial gastroenteritis and that antibiotics should not be prescribed at this stage. We can see how Coco’s mother could have felt that she was being treated with ‘arrogance’,[4] and certainly not as a member of a wider team.

D. The chronology of Coco’s stay at Treliske records no fewer than ten occasions when her pain score was either assessed as zero or not assessed at all. Had more attention been paid to the pain that she was suffering, taken in conjunction with clinical indications of Coco’s declining physical state, this would have emphasized the strength of the case for escalating her treatment to the ‘critically ill’ category.

The events
1. When Coco was first brought in to Treliske, on July 25th, her pain score was not assessed.

2. Nurse 4 described how, on July 26th, on her admission to the ward, Coco was ‘physically upset and rolling around the bed’. But nursing observation did not include a pain assessment. (8.3.3)

3. At 10.36 pm Coco was given intravenous paracetamol. This was the first time she was provided with pain relief. However, there is no corresponding pain score recorded at this time. (8.3.11)

4. At 2.54 am the following morning (July 27th), Coco’s observations were taken. Her pain score continued to be recorded as zero: this was despite her being described as ‘restless and rolling around the bed’. (8.4.2)

5. At a nursing review at 11.50 am on July 27th, Nurse 6 described Coco as ‘so distressed’, ‘flinging her arms’. Throughout the daytime shift Coco’s pain score was recorded as zero. ‘Whilst staff felt that one reason it was difficult to record a blood pressure was that Coco had a diagnosis of autism, no referral was made to the learning disability team.’ (8.4.18 & Commentary)

6. It was recorded in the nursing notes that Coco remained restless. At 8.30 pm she was given paracetamol. Her pain assessment continued to be recorded as zero – the absence of pain. During this shift Coco was not weighed and no blood pressure recording was made.’ (8.4.28)

7. Coco’s mother requested that she be given some additional pain relief to settle her. At 10.48 pm she was given a dose of chloral hydrate: this is a sedation agent, not pain relief. The clinical record did not record the reason for prescribing chloral hydrate. Coco was unable to tolerate it and vomited after it was given. Her pain score had been recorded as zero. (8.4.29)

8. Coco’s observations were taken at 1.09 am and 2.16 am. Her pain was recorded as zero. (8.5.3)

9. At 4.00 am on July 28th Dr P reviewed Coco. She was recorded as being ‘restless’. No assessment of her pain was made. (8.5.7)

10. At 4.30 am Nurse 5 reviewed Coco. Her parents were concerned that she was in pain and very restless. A dose of intravenous morphine was given. Her pain score continued to be assessed as zero. The medical record showed that morphine had been given, but not the reason why. (8.5.11)

Discussion The investigation report notes that the assessment of pain is an example of where the nursing staff did not apply the FLACC assessment tool (based on observations of Face, Legs, Activity, Cry and Consolability, this is a method of assessing pain for children under 7 and individuals who are unable to describe their pain) to support their clinical judgement. On many occasions Coco was described as ‘distressed’, inconsolable’ or ‘agitated’, but her pain score was recorded as zero, not having pain. (

The investigation report also notes that during interviews with the nursing staff they were unable to say why these non-verbal signs were not considered to be indications of pain. The nursing staff were reliant on Coco’s parents’ feedback, rather than formally assessing pain. ‘Although parents’ views might be useful in understanding how children express pain, this should not be considered in isolation. Appropriate clinical assessment tools and clinical expertise should be used.’ (

A lay observer might with justification say that what we see here is insensitivity amounting to callousness.

This sad sequence of events raises questions about teamwork, or the lack of it, where Coco’s care was concerned. The rulebook on pain assessment was not being followed by either doctors or nurses. So senior nurses were not instructing junior nurses about using the FLACC assessment tool, and no doctor was asking nurses about the pain that Coco was in. Communication on this important subject did not exist. Moreover, although staff were aware that Coco had a diagnosis of autism, no referral was made to the Learning Disability team. This is another communication failure. There was no ‘team awareness’, let alone teamwork.

The investigation report notes: ‘The Learning Disability team told the investigation team they would have been able to assist the nursing team to obtain a blood pressure reading using distraction techniques. However, it would be expected that the paediatric ward team would be skilled in distraction and play therapy.’ (9.7.2) This comment by the investigation team demonstrates a worryingly narrow focus on technique: members of a learning disability team are likely to bring to a patient not just skill but empathy, the ability to relate sympathetically to a patient, something all too likely to be lacking in staff who do not record a pain score.

E. Because of how Coco was accommodated on Polkerris ward, she was not treated as a patient receiving or requiring high dependency care. When a communication channel failed there was no alternative channel to serve as a backup.

The events Coco was admitted to Polkerris ward at 7.30 pm on July 26th. The ward was described to the investigation team as ‘busy’. With her diarrhoea and vomiting, for reasons of infection control she needed to be isolated, and she was admitted to a side room at the bottom of the ward. This room was away from the main thoroughfare of the ward. Although it was referred to as a high dependency unit (HDU) side room it was not commissioned for this purpose. (8.3.2)

This meant that, despite Coco having been reported by Dr C as ‘one to watch’ and noted by Dr E as a child who needed ‘further assessment’, she was not treated as a patient receiving or requiring high dependency care. Access to the electronic observational charts that provide an overall view of trends and pattern of results was not available in the side room. (8.3.1/2)

On the afternoon of July 27th, there were three patients in the high dependency unit. (8.4.20) Coco was handed over to Dr H as an HDU patient and in need ‘of a clinical review but not an emergency review’. Dr H saw her in the side room opposite HDU and it was clear to Dr H that Coco was not well. She told the investigation team that she remembered the nurse was concerned about Coco and she noted that her resting heart rate had been abnormally high all day. (8.4.21)

Dr H reviewed Coco at 6.40 pm and made a management plan for her: to repeat the blood tests and dependent on the results to decide whether to start antibiotics. Dr H told the investigation team that it was very important that blood tests were ‘done as soon as possible’. She went home after that review. Before leaving the ward she requested that ‘the evening registrar and consultant ensure these bloods were taken urgently’. (8.4.25) The investigation report does not say in what form that request was made. Dr F, who was the resident evening consultant on duty, was busy with the flow of new referrals. (8.4.20)

Coco’s nursing care was handed over to Staff Nurse 7 at 8.10. pm. She was also looking after another very sick child who was in a side room at the opposite end of the ward. This meant that she was continually on the move between the two rooms and unable to keep a close eye on either patient. (8.4.26)

The investigation report notes: ‘Information sharing between those involved in Coco’s care was not always effective, and in some instances did not take place. For example, the evening consultant on July 27th (Dr F) did not recall being asked to ensure that blood tests were done urgently that evening.’ (9.6.1)

Discussion We aren’t told in what form Dr H put her request that blood tests were carried out urgently, but what is revealed here is the danger of relying on a single communication channel. Although it is conceivable that had there been access to electronic observational charts, and had the staff nurse not been under considerable pressure, the outcome might have been different, there was no systematic backup to make sure that the message got through.

The investigation report devotes only a single paragraph to communication. It says: ‘Good communication between clinical professionals is essential to protect patient safety. The timely and up to date exchange of clinical information between teams is a fundamental aspect of this.’ But the report has nothing concrete to say about how good communication is to be ensured. In this case it was clearly lacking.

But we may also question why Dr H did not feel, as a matter of professional responsibility, that she should wait for the results of the blood tests – required ‘as soon as possible’ – so she could act on them before she went home. Clearly, if those results were required urgently, important decisions about treatment would be waiting on them. In an effective team people take responsibility when the situation demands it. Dr H chose to go home. She was on call, but a call never came and she did not call the ward to find out why.

A hospital under pressure
For the first two days when Coco was in Treliske (July 25th and 26th), the hospital was under considerable pressure and ‘unable to deliver comprehensive care’ (NHS England’s Operational Pressures Escalation Framework, Level 4). Over the next two days the pressure was slightly less, with the hospital still experiencing ‘major pressures compromising patient flow’ (Level 3). It is clear from the account given in the investigation report that departments in the hospital were understaffed and doctors, nurses and others were under considerable stress.

This helps us to understand some of what went amiss in Coco’s case, but it would be wrong to use stress as an excuse for the failings noted in the investigation report. Rather, since an effect of stress is to ‘fragment’ the system, driving people to consider only their immediate personal task in hand, we can use those failings to identify weak points in the system. We can then ask how teamwork can assist in addressing these.

Developing teamwork: some questions and suggestions
1. The investigation report uses the term ‘team’ remarkably loosely, which suggests that it may be something of a foreign concept to the healthcare consultants who wrote it and to those who work in hospitals. Simply asking the question ‘What team do you belong to?’ would be a start. The answers might be revealing: do people define themselves by the profession to which they belong, or their level in the organization, or the ward or laboratory in which they work? Where do their loyalties lie? Are they aware of a wider team beyond the ‘core’?

2. ‘How do you see your role in the process that patients go through?’ would be another good question, especially if backed up by ‘Who do you depend on to achieve a good outcome?’ That should help to develop an appreciation of the contribution that members of the wider team in the hospital make. It would encourage staff to consider the whole process that patients go through, and it should develop some grasp of the hospital as a unified enterprise.

Taking a ‘process view’ can yield further insights. As we have seen, on the first occasion that Coco came to the emergency department she was seen by a paediatric nurse and reviewed by a locum registrar. Her blood pressure was not taken, her pain score was not assessed, nor was she weighed. Arguably seeing and assessing a patient on arrival at the emergency department is a task that a senior, experienced doctor could perform more accurately, more speedily and at less cost to the hospital and the patient.

3. Familiarity with members of the wider team should generate a deeper and more subtle appreciation of the part that they can play in caring for a patient. As we saw from the comment in the investigation report about how the Learning Disability team might have been able to assist the nursing team to obtain a blood pressure reading, the idea that members of that team could bring to Coco not just skills but empathy seems not to have occurred to the investigators.

4. Attitudes to members of other professions can foster teamwork or inhibit it. Does a doctor regard a nurse as a fellow professional, with valuable eyes and ears and skills, or merely as someone providing a service? If a nurse ventures an opinion or asks a question about a diagnosis, is her contribution taken seriously or does it get a frosty ‘who do you think you are?’ reception? It is of course the fellow professional who is listened to who makes the greater contribution to teamwork. Someone who is treated with arrogance will refrain from offering a contribution.

Arrogant behaviour is consistent with a claim to the status of ‘professional’, which invariably implies a claim to autonomy: ‘I’m a specialist, I have received specialist training, this is my field, I’m experienced, I know what’s best for the patient, don’t tell me what to do, I don’t need to consult you, decisions should be taken by me.’ Such a claim to autonomy will be bolstered by the use of a specialized language that can’t be understood by lay people.

I would suggest that any member of the wider team – including parents and other relatives – should feel entitled and able to ask a question and receive a respectful, thought-about, understandable answer. There is no monopoly of wisdom. Every now and again, one of these will give rise to a suggestion worth acting on. No member of the wider team should feel, as Mrs Bradford did, that they were treated with arrogance.

5. With regard to record keeping, as the investigation report points out, in Coco’s case there were instances where record keeping did not happen. This is inexcusable. Record keeping is an essential part of diagnosis: it enables trends to be detected, and it is essential as a means of building the clinical picture and communicating with other staff who are looking after a patient.

But there is a broader point to be made about record keeping. Records are a means of communicating between the doctors and nurses who are caring for a patient. They should be available to all of them. It cannot be acceptable for a doctor going off-shift to leave a request that blood samples are taken urgently and the doctor replacing her not registering the request. Requests should be put on record, and the record should be available for other staff to see, so that there is a backup communication channel if the intended recipient does not register the message. Record writing and reading should not be a chore. It is an integral part of the process.

6. Hospital culture needs attention. Consultants train the junior doctors assigned to them by setting an example. They demonstrate techniques, of course, but they also demonstrate attitudes. Consultants set the ‘tone’ and thus the culture of the department. A slapdash attitude to record keeping and to following the rulebook teaches junior doctors and other staff that these are acceptable behaviours. Likewise an arrogant us-and-them attitude on their part to non-medical staff also teaches behaviours that are inimical to good co-operation and thus to the welfare of patients. The importance of a teamworking culture is something that needs to be instilled in medical professionals from the start of their education and reinforced during their working life.

There is a further point to be made about teams. They make effective pressure groups. An alliance of paediatric doctors and nurses, especially if supported by their wider team, could have lobbied the hospital management for a reorganization of the physical layout of Polkerris ward and for extending the High Dependency Unit within the ward. The likelihood of a successful outcome is certainly much greater when people act together than when they don’t.

7. Finally, where was management before and during Coco’s short stay in Treliske? Why were problems of short staffing, excessive pressures on paediatric care, defective record keeping and ignoring the rulebook not being picked up and addressed?

The investigation report is remarkably silent on management aspects of Coco’s case. It does not define the ‘paediatric team’, nor does it tell us whether there was a paediatric department or who, if there was, headed it.

An exciting time at Royal Cornwall?
In July 2018, while the investigation into Coco’s death was under way, the RCHT Chief Executive, Kathy Byrne, resigned. The hospital trust now has an interim Chief Executive, Kate Shields. Under her aegis, on November 19th, 2018, the Trust posted an advertisement for clinical directors.[4] Here are extracts:

‘It’s an exciting time at Royal Cornwall as we implement our proposed new Care Group structure, moving from four divisions to seven distinct care groups – designed to deliver better patient care, improve accountability and make our colleagues’ working lives more fulfilling. If you’ve got the passion, commitment, patient focus and skills we need, and you can really demonstrate our values in your leadership style then we’re eager to hear from you.

‘We’re recruiting [for] demanding, strategic and visible senior roles that we’ve put in place to really bring about positive change in our hospitals and services; establishing more manageable portfolios and bringing leadership closer to our all-important front-line colleagues and our patients.

‘Our proposed structure includes the following seven care groups – General Surgery & Cancer, Specialist Surgery, Clinical Support, Women, Children & Sexual Health, Anaesthetics, Critical Care & Theatres, Specialist Medicine and Urgent, Emergency & Trauma.

‘Starting in January 2019, we are launching an extensive leadership development programme across the Trust. You’ll be one of the first leaders to experience this innovative and exciting programme which will deliver real benefits to how we work together to achieve the best outcomes possible for our patients and colleagues.’

An ‘exciting’ time? ‘Challenging’ would be a better word. And note that it is impossible to tell from the third paragraph whether ‘Children & Sexual Health’ constitutes a ‘care group’ or the relevant grouping is ‘Women, Children & Sexual Health’. We may question whether it is going to be advantageous to child patients to be bundled together with women and people with sexual health issues. The rationale for doing either is not apparent.

But it is telling that, despite the aspiration to ‘work together’, this advertisement emphasises ‘leadership’ rather than teamwork. This fits in with today’s NHS culture. There is an NHS Leadership Academy, the task of which seems to be to groom junior managers to take on senior positions. Worryingly, its Healthcare Leadership Model: The nine dimensions of leadership behaviour[5] is more suited to an academic treatise than a practical guide.

Some practical suggestions for effective teamwork
Drawing on Coco Bradford’s experience, we can identify some genuinely practical and down to earth suggestions for hospital staff.

  1. Make sure that everybody reads and follows the relevant guidelines, the rulebook that governs the task in hand. It has a purpose. It is designed to incorporate best practice and it will enable you and your colleagues to follow each other’s work.

  2. Keep meticulous records. Do not regard record-keeping as a chore. It plays an important part in tracing and showing a patient’s progress (or regress), and in building up a clinical picture.

  3. Senior doctors: remember that you are mentors of junior doctors. Keep an eye on registrars and other juniors, put them at their ease so they feel comfortable in asking for your advice. Importantly, remember that every aspect of your behaviour is setting them an example, so demonstrate good practice: make it a good example.

  4. Doctors: remember that nurses see more of the patients than you do. Ask what they have noticed about patients and their behaviour, and explain your reaction, so nurses can learn from you and develop a sense of what is significant in a patient’s behaviour, what to look out for.

  5. Communicate, communicate: Make sure that other people know what you’re doing, and why. Be aware that multiplying communication channels isn’t wasteful, it’s a security measure that helps to ensure that crucial information will get through.

  6. Adopt the principle that everybody is entitled to ask a question, to speak and be listened to, and to be given a thought-about answer. ‘Everybody’ should include not only colleagues but also parents, other relatives, and patients themselves. Among staff, everybody dealing with patients brings knowledge and expertise in their area of care. Appreciation rather than self-justification should be the watchword.

  7. If a child or an adult is special – has an autistic condition, for example – make sure there is somebody with empathy as well as training to work with them. Empathy is crucial.

  8. Pay attention to a patient’s pain. Nobody, even a patient who can’t be calmed and resists being touched, deserves to be in pain. Staff should be familiar with methods of gauging and ameliorating pain, and practised in employing them.

  9. Understand that you and others have to be flexible. This is a matter of professional responsibility. Most visibly it applies to time-keeping. You don’t clear off home before an operation is completed if it overruns the allotted time. Likewise you wait for urgent blood results to come through, rather than going home and leaving it to others to – you hope – pick them up and act on them. In a well-functioning team it will be the norm for colleagues to cover for one another, so over time no-one will lose out.

  10. If something isn’t working, get together with all your core team and your wider team to decide what needs to be done and press for it. A team is more powerful than a single individual in dealing with management. And you will be surprised by how much energy you generate.

Notes and references

* The title of of this paper was amended on December 19th, 2018 to replace the words ‘little girl’ with ‘child’ to avoid the paper being blocked by Safe Search.

  1. This paragraph was amended on December 2nd, 2018, to correct an inaccuracy in contemporary press reports.

  1. Facere Melius, Independent investigation into the death of Coco Bradford, September 2018

    Appendices at

  2. Graeme Wilkinson, ‘Royal Cornwall Hospital slammed over mistakes following death of Coco Bradford aged 6’, Cornwall Live, October 26, 2018 (This incorporates a statement by Coco’s mother, Rachel Bradford)

  1. Advertisement in Health Service Journal, November 19, 2018.

  1. NHS Leadership Academy, Healthcare Leadership Model: The nine dimensions of leadership behaviour, 2013