What we found
Averil Hart was a young woman with anorexia nervosa. She died on 15 December 2012, aged 19, following a series of failures that involved every NHS organisation that should have cared for her. Her death was avoidable. There were multiple opportunities between August and December 2012 to identify what was happening to Averil, to intervene to remedy the situation at that time, and therefore to prevent the subsequent course of events that led to the final emergency admission to hospital which culminated in her death. All of these opportunities were missed.
Averil’s illness had been serious enough that she was treated as an inpatient for over ten months from September 2011 in the Eating Disorders Unit in Cambridge, part of Cambridgeshire and Peterborough Foundation Trust (the Cambridgeshire and Peterborough Trust) but situated on the Addenbrooke’s Hospital site.
In April 2012, while still an inpatient, she received an offer from the University of East Anglia to attend university in Norwich, and was keen to take up the place that September. This would require her discharge from hospital, which is in any case a vulnerable time for those with anorexia nervosa, and it was clear that Averil’s illness was severe. Averil would need to move to a new location, a new environment at university and a new clinical service, the Norfolk Community Eating Disorder Service (NCEDS); all of these would potentially further increase the risk of relapse subsequent to discharge.
Averil was assessed as fit for discharge, although it was recognised that she was vulnerable to subsequent relapse, and left hospital on 2 August 2012. At her request, responsibility for her care was transferred to NCEDS from her move to Norwich on 23 September 2012, and she was not seen by the Suffolk service that covered her home area. NCEDS was managed by the Cambridgeshire and Peterborough Trust and had been explicitly set up to improve services in the aftermath of a previous death. Given that Cambridgeshire and Peterborough Trust knew the risks of leaving hospital for a new environment and the risk of relapse, it was particularly important that there was explicit documentation of warning signs of deterioration in her condition and the contingency plan to be invoked if they materialised. But the care plan at discharge failed to set these out robustly or explicitly enough.
Following discharge, joint working between the Eating Disorder Unit and NCEDS was poor, and NCEDS was operating with staffing shortages due to recruitment difficulties. Neither Averil’s weight nor her mental health was adequately monitored prior to her move to Norwich. Interim support was provided by the Cambridgeshire and Peterborough Trust, but there were gaps in weight monitoring from the outset. When she did move, there was a delay in allocating a care coordinator by NCEDS which meant that she was not supported or properly assessed for a further month, during which her condition deteriorated and her weight decreased significantly again. When she was first weighed in Norwich, she had lost 6kg, a very significant reduction for someone in Averil’s condition.
These failures – of assessment, of coordination and of care planning and implementation at a time of heightened vulnerability – all represent missed opportunities to recognise that Averil’s condition was deteriorating and that she was at significant risk, and missed opportunities to intervene to prevent the further deterioration that led to the final admission to hospital which ended with her death.
The care coordinator that was then appointed by NCEDS had no experience of looking after people with anorexia nervosa. While it is clear that much NHS care is properly carried out by newly qualified clinicians who are still gaining experience, it is a matter of significant concern that eating disorder services across the country are sometimes obliged to give lead responsibility for managing such a complex and challenging condition over an extended period to an inexperienced clinician. We are advised that this practice does sometimes happen in eating disorders services, because of the difficulty of recruiting and maintaining staff with the requisite skills and experience.
Although the appointment was not out of line with established practice in the specialty because of the difficulty of recruiting and retaining staff with the requisite skills and experience, we believe that this is not a satisfactory situation and requires national attention.
When there is no alternative to an inexperienced care coordinator, it is vital that they be properly supported both in this role and as the clinician in day to day contact with the patient. The support provided by the Cambridge and Peterborough Trust to this Eating Disorder Service clinician was inadequate. A multi-disciplinary team should have been arranged, to provide input from other clinical disciplines including psychiatry and to provide another clinician in contact with Averil. This is important in providing scrutiny and challenge that can detect the behaviours often used to hide weight loss by those with anorexia nervosa without jeopardising the therapeutic relationship between the patient and the clinician principally providing care. The failure to provide multi-disciplinary team support left the care coordinator as the sole point of contact with Averil, and impaired the ability to detect deterioration in her mental and physical condition.
The failure to provide for multi-disciplinary team care also meant that there was no opportunity to pick up an error by the care coordinator in calculating Averil’s four-weekly rolling average weight, contributing to an over-optimistic assessment of her weight loss. When the care coordinator went on leave, by which time Averil had become more unwell, no cover was arranged. The lack of handover to another identified clinician impaired the service’s ability to detect the deterioration in her condition.
Averil was under the direct care of NCEDS from her arrival in Norwich on 23 September 2012, and her condition deteriorated markedly through to late November. The Cambridgeshire and Peterborough Foundation Trust was responsible for NCEDS; its actions fell far short of what should have happened and constituted service failure. The failure to ensure adequate surveillance that was capable of detecting the change in Averil’s condition was another missed opportunity to prevent her further deterioration and the subsequent admission to hospital that culminated in her death.
Responsibility for monitoring Averil’s physical health once she arrived in Norwich rested with the GPs of the University of East Anglia Medical Centre (the GP practice). The requirements had been clearly set out in a referral letter at the time of her discharge from hospital in August, but the GPs failed to follow them. Averil was not seen regularly, or as often as she should have been. When she was seen, many of the observations required were not made. Signs and symptoms of Averil’s deteriorating health were missed.
On 8 November, she was asked to return in a month’s time, although monitoring was required weekly. There was no single point of oversight within the GP practice because Averil was not provided with a named GP.
Had the GPs done as they were asked, the deterioration in Averil’s physical condition would have been recognised sooner and action taken before she reached the point of collapse. The GP practice’s actions fell far short of what should have happened, and constituted service failure. Their failure to implement properly the straightforward monitoring that was required was another missed opportunity to intervene to prevent the subsequent deterioration that resulted in her final admission to hospital.
Averil’s care was divided over this period, with the GP practice monitoring her physical condition while NCEDS took responsibility for mental health and the behavioural aspects of her eating disorder. We are advised that joint arrangements such as this are common in this field of clinical work. There may be reasons to separate care in this way, but it places a premium on effective communication and effective care coordination, particularly when the patient may be an unreliable source of information because of their condition.
Communication is a two-way process that is the responsibility of both parties and not simply the transmission of a message, and in this case, it was not effectively established between the GP practice and NCEDS.
When Averil was seen by NCEDS in October, it was decided that responsibility for monitoring her weight would be taken on by them, with all of the rest of physical health monitoring continuing to be done by the GP practice. As two-way communication had not been established effectively, the GP practice concluded wrongly that, as care was now being undertaken elsewhere, they could reduce their already inadequate level of physical monitoring still further. The GP practice did not communicate their interpretation of the change of plan to NCEDS, who remained unaware of the error and unable to correct it.
This failure of communication between the two organisations meant that neither was effectively monitoring her condition, and was another missed opportunity to prevent the subsequent deterioration that resulted in her final admission to hospital.
As a direct result of the cumulative impact of this series of clear failures by the Cambridgeshire and Peterborough Trust, the GP practice and NCEDS, a very significant deterioration in Averil’s condition went unrecognised. By the end of November she was very unwell and her weight had dropped to an alarming extent.
When Averil’s father visited her on 28 November, he immediately recognised the marked deterioration in her condition over the preceding month. He sought to raise the alarm with her health carers. His initial approach was to the Eating Disorders Unit at Addenbrooke’s Hospital, where she had been treated as an inpatient prior to her discharge four months previously. Averil’s consultant during her admission declined to take the call although Mr Hart could hear her instructing that he be directed to NCEDS. A helpful and sympathetic response would have been to listen to his concerns, but she did not. Having then contacted NCEDS directly, Mr Hart was reassured to receive an email the following day telling him that action would be taken. He should not have been the first to recognise her condition, nor the first to prompt action.
Averil was due to attend NCEDS on Friday 7 December. Following her father’s intervention, and recognising relatedly that there was significant concern about her deteriorating condition, NCEDS arranged for this planned visit to include a medical review. Averil telephoned NCEDS during the evening of 6 December and left a message cancelling her appointment. Her care coordinator at NCEDS attempted to telephone Averil the next day, but received no reply.
On the morning of 7 December, Averil was found in her room in a state of collapse. She was transferred by ambulance to the emergency department of the Norwich Acute Trust. On admission, she was acutely unwell, with a low temperature, low blood pressure and low blood glucose. She was also very underweight. All of this indicated clearly that her anorexia had deteriorated severely and now constituted a medical emergency that was potentially life threatening. It should have been clear that she was in urgent need of refeeding.
The urgency of addressing Averil’s condition was not recognised by staff at the Norwich Acute Trust. Averil was allowed to walk around the ward (a common strategy to counteract feeding among people with anorexia that Averil was known to employ), and to feed herself from a trolley, so that her food intake was unknown. The medical team appeared focused more on pursuing other unlikely diagnoses than on the need to ensure an effective refeeding regime with support from mental health professionals. She saw no specialist eating disorders clinician for three days after admission, by which time her condition had deteriorated further. Nursing care was deficient and failed to monitor her condition effectively.
These clear failures of care wasted more time during which the continued further acute deterioration in Averil’s condition remained undetected. The Norwich Acute Trust’s actions fell far short of what should have happened and constituted service failure. This was another missed opportunity to intervene to prevent yet further deterioration in her condition, deterioration that culminated in her death.
When the seriousness of Averil’s condition was finally recognised, she was transferred to a gastroenterology ward at Addenbrooke’s Hospital, part of the Cambridge Acute Trust. She arrived at 2:40pm on 11 December, but was not seen by a doctor for almost five hours. Given her condition, this was a significant and inexplicable delay. Even when she was seen, the clinical assessment was cursory and no decision was recorded concerning her clinical risk or immediate care.
During the evening of 11 December, Averil’s blood glucose fell further to a level that was clearly life-threatening. She was offered treatment, but refused. Inexplicably, there was no proper assessment of her mental capacity to take this decision, and no mental health assessment with a view to establishing treatment under the Mental Health Act.
A healthcare assistant had been provided by the Cambridgeshire and Peterborough Trust to sit with Averil at this time, to guard against any attempts by Averil to sabotage treatment, for example by excessive mobility. By now, however, she was too weak to be mobile and the role was redundant. Nevertheless, the healthcare assistant remained, but took no part in her basic care. We consider that she should have done more to assist Averil and her family.
Overnight, there were further clear signs that Averil was increasingly critically ill, including extremely low blood glucose levels. Following an unsatisfactory telephone call between a junior doctor and a consultant that failed to result in effective communication, no definitive action was taken and she was found unresponsive the following morning. It became clear that she had severe brain damage due to extremely low blood glucose and that further restorative treatment was futile. Averil died at 11pm on 15 December 2012, with her family by her side.
These were multiple serious departures from the standards of care expected that meant that the critical nature of Averil’s condition was not recognised and treatment was not implemented promptly, as it could and should have been. The Cambridge Acute Trust’s actions fell far short of what should have happened, and constituted service failure.
This was the final failure that led immediately to Averil’s death, but it was the last of a long series of missed opportunities to recognise her deteriorating condition and intervene to prevent the need for her final hospital admission as an acutely ill medical emergency.
Following Averil’s death, as with any avoidable harm, the question naturally arises as to what happened and why, and how will it be prevented from recurring in future. Averil’s family, with Mr Hart taking a lead, rightly sought answers from the organisations involved. As happens to too many in these circumstances, he found the process to be difficult, unnecessarily painful and ultimately frustrating.
This was clearly a very serious clinical incident that required a commensurate investigation by the NHS. This would properly have been provided by an independent investigation, commissioned by all of the NHS organisations involved, which looked at the failures of care across all of the organisations and the failures of communication between them. In the event, the piecemeal investigations that were done comprised an unsatisfactory process that was unlikely to generate a complete account of what had gone wrong and how it could be remedied; nor was it likely to command the respect of the family that a thorough process had been undertaken.
The responses to Mr Hart’s requests for information were delayed and appeared evasive, and information he requested was often not provided. The responses to his complaints were equally unsatisfactory, and often appeared defensive or protective of the organisation concerned. Some information stored in electronic format turned out to have been deleted; the decision to delete material related to a significant safety incident was ill-considered and inappropriate. An anonymised account of Averil’s death was going to be used in the revised guidelines on management of severe anorexia nervosa (MARSIPAN) but was subsequently removed. Mr Hart and Averil’s mother were given an account that inappropriate pressure had been brought to bear on the author to withdraw the account, but the author denied this and we were unable to substantiate it.
There were clear failings in the response to Mr Hart. The Cambridgeshire and Peterborough Trust’s handling of Mr Hart’s complaint was so poor that it was maladministration. The GP practice’s investigation of Mr Hart’s complaint was deficient and did not uncover the serious failings in her care. Their complaint handling was so poor that it was maladministration.
The Norwich Acute Trust’s initial investigation of Mr Hart’s complaint was deficient and did not uncover the serious failings in her care. Their complaint handling was so poor that it was maladministration. The Cambridge Acute Trust’s investigation of Mr Hart’s complaint was deficient. It did not uncover the serious failings in her care or that her death should have been avoided. The Trust did not respond appropriately or sensitively to Mr Hart’s follow-up complaints and their complaint handling was so poor that it was maladministration.
Individually, these failures are seriously unsatisfactory. Taken collectively, they paint a consistent picture of unhelpfulness, lack of transparency, individual defensiveness and organisational self-protection that is of great concern. It is hardly surprising that this leads to a lack of trust from complainants, in this case Mr Hart. Equally unacceptable are the missed opportunities to learn and to improve services inherent in the incomplete and defensive investigations of safety incidents such as this.
When Mr Hart’s complaint reached NHS England, there was an opportunity for them to intervene to identify the inadequate and defensive nature of the previous investigations and complaint handling, and initiate an effective response. NHS England’s actions, however, were inadequate. They failed to demonstrate that they understood the very serious issues Mr Hart complained about and, due to a misunderstanding, they closed his complaint without investigating it. NHS England’s approach was not customer focused. It was so poor that it was maladministration.
The death of Averil Hart was an avoidable tragedy. Every NHS organisation involved in her care missed significant opportunities to prevent the tragedy unfolding at every stage of her illness from August 2012 to her death on 15 December 2012. The subsequent responses to Averil’s family were inadequate and served only to compound their distress. The NHS must learn from these events, for the sake of future patients.