Investigation report: Provision of mental health care to patients presenting at the emergency department

Diane, a 57-year-old woman with a history of mental health problems, was in the care of the community mental health service. As her mental state fluctuated, she experienced increasing levels of anxiety, self-harmed and expressed thoughts of suicide. Over a two-year period she had received treatment from her GP, the local crisis resolution and home treatment team, the ambulance service and the emergency department of the local district general hospital.

Diane presented four times to the same emergency department following self-harm, receiving different levels of care on each occasion. Her physical health was generally well attended to by the emergency department staff. National guidelines recommend those who have self-harmed should receive a psychosocial assessment from a specialist mental health professional. The liaison mental health service team was located close to the hospital and was commissioned to operate between 08:00hrs and 23:00hrs; Diane was referred for assessment on the first two occasions but not thereafter. Consequently, the community mental health team was unaware of Diane’s crises when she attended the emergency department following self-harm on the last two occasions.

Six weeks after she had presented to the emergency department for the third time she received a visit from her care co-ordinator.

After this visit Diane reported she had taken an overdose and the next day presented to her GP who advised her to go to the emergency department. However, she did not go there and later that day her carer called 999. Diane arrived at the emergency department by ambulance at 20:19hrs. Following prolonged pressure on services, the emergency department was on ‘black status’ and experiencing its busiest day of the month.

After waiting for almost one hour, Diane was assessed, and her self-harm was recorded. The clinical notes remark that she wanted to go home. Her physical health was attended to; however, no referral was made to the liaison mental health team and Diane left the department sometime in the early hours of the morning. Later that morning she attended her GP practice for a repeat prescription but the GP was reluctant to prescribe, and deferred the decision until later that day.

In the early afternoon, Diane left a note on the railway station platform before lying in the path of an oncoming train. Following treatment at the scene, she was airlifted to a major trauma unit, where she died from her injuries.

There is a strong link between self-harm and suicide. Diane’s case highlights the challenge to the healthcare system when treating people experiencing a mental health crisis. Emergency departments treat approximately 220,000 cases of self-harm a year. The Five Year Forward View for Mental Health is attempting to address this situation by increasing the presence of liaison mental health services in acute hospitals.

Following an initial investigation which reviewed Diane’s four presentations to the emergency department, HSIB progressed to a full investigation. As part of its investigation, the HSIB conducted a series of observational studies, interviews and discussions with subject matter experts to establish how risk to mental health is assessed and then managed nationally in the emergency department.

Diane did not come to direct harm during treatment in the emergency department.

• Diane’s final two presentations at the emergency department represented missed opportunities to intervene and to take measures that may have helped to improve her mental state.
• The provision of liaison mental health services was variable across England and there was no consensus on commissioning models.
• Liaison mental health services had a positive influence on managing the care of patients in the emergency department and were most effective when services had a permanent integrated presence in the emergency department.
• The benefits of liaison mental health services were difficult to quantify in financial terms for commissioners. However, they were broad and stemmed from the integration of mental health professionals in the general hospital and the consequent shift in attitudes towards understanding the complexities of mental health.
• The process for triage and initial assessment completed by emergency department nurses was effective at identifying physical health problems but lacked structure when assessing mental state.
• There was the potential for misunderstanding in the self-harm guidance around interpretation and use of the Australian mental health triage tool.
• The national guidance issued to emergency department staff for the initial assessment of people who have self-harmed lacked coherence between documents and did not consistently describe a detailed process.
• In the absence of clear national guidance on the conduct of initial assessments, emergency departments continued to use locally developed, unvalidated tools of varying standards.

1.1.1 In England, it is estimated that 5% of all hospital emergency department attendances are primarily due to mental ill-health. Self-harm accounts for at least 220,000 emergency department attendances per year. A recent study suggests that this figure may underestimate the true rate by 60%.

1.1.2 Self-harm rates appear to follow the trend of suicide rates and 68% of people who had died by suicide within three months of contact with mental health services had a history of self-harm.

1.1.3 A recent Royal College of Emergency Medicine safety alert highlighted that absconding from care was a theme in 14% of all reported clinical incidents[5]. The extent to which the emergency department can provide a safe and caring environment to treat those experiencing a mental health emergency, influences their willingness to receive a psychosocial assessment and be referred to the wider service.

1.1.4 To understand what influenced the present-day provision of mental health care in the emergency department, several key documents were identified:

No health without mental health[6]

1.1.5 A cross-government paper published by the Department of Health in 2011 which set out a mental health outcomes strategy. It recognised that good mental health and resilience are fundamental to physical health, relationships, education, training, work and to realising potential.

Whole-person care: from rhetoric to reality: Achieving parity between mental and physical health[7]

1.1.6 Published by the Royal College of Psychiatrists in 2013, this report established the definition and vision of achieving ‘parity of esteem’. It described parity as “Valuing mental health equally with physical health”. One of its recommendations was that people with mental health problems who are in crisis should have an emergency service response of equivalent speed and quality to that provided for individuals in crisis because of physical health problems.

Mental Health Crisis Care Concordat[8]

1.1.7 The Mental Health Crisis Care Concordat - improving outcomes for people experiencing mental health crisis (2014), was a national initiative led by the Department for Health, which called on those responsible for commissioning, providing and delivering the mental health services to commit to a set of core principles around caring for those experiencing a mental health crisis.

1.1.8 It set out a requirement that whatever the circumstances of their arrival, people in mental health crisis should expect emergency departments to provide a place for their immediate care and adequate liaison psychiatry services to ensure the necessary and on-going support required was delivered in a timely way.

1.1.9 It stated that clear responsibilities and protocols should be in place between emergency departments, mental health and substance misuse services, and other agencies or departments, to ensure people receive treatment on a par with standards for physical health.

1.1.10 The report highlighted some key lessons, including services that should be commissioned to meet local needs. It also highlighted that agencies involved in crisis care should take a more joined-up approach.

1.1.11 It described ‘Local Crisis Care Concordat Groups’ taking a major role in making sure pathways for crisis care provide the right care to people in crisis, when they need it.

Right Here Right Now[9]

1.1.12 In 2015 the Care Quality Commission explored the lived experience of people during a mental health crisis. It noted that services across England varied greatly in providing a timely and high-quality response to people experiencing a mental health crisis.

1.1.13 The report found that access to, and the quality of, services after 17:00hrs was “not good enough”.

1.1.14 The report noted that peak hours for admissions to hospitals via the emergency department owing to self-harm were between 23:00hrs and 05:00hrs[10].

The Five Year Forward View for Mental Health[10]

1.1.15 This report outlined a strategic plan to deliver mental health care for all age groups. It recognised that mental health had not had the priority awarded to physical health, the service had been short of qualified staff and it been deprived of funds. The report recognised the importance of a joined-up approach across teams providing urgent and emergency mental health response for people attending the emergency department or admitted as inpatients to acute hospitals. It stated that by 2020/21, no acute hospital should be without all-age mental health liaison services in emergency departments and inpatient wards, and at least 50% of acute hospitals should be meeting the ‘Core 24(A)’ service standard as a minimum; ‘Core 24’ is explained in more detail in Section 5.2 of this report.

1.1.16 It acknowledged mental health services had been plagued by years of under investment and proposed that payments for services should incentivise the provision of integrated mental and physical healthcare.

1.1.17 The report set-out improving the 7-day crisis response service across the NHS would help save lives and was a key component to reducing suicide by 10% by 2020/21.

Achieving Better Access to 24/7 Urgent and Emergency Mental Health Care[11]

1.1.18 In response to the Five Year Forward View for Mental Health, NHS England published the evidence-based treatment pathway (EBTP) for urgent and emergency liaison mental health services for adults and older adults presenting at the emergency department in 2016. The guide covered the first 24 hours of care following presentation. The pathway included recommended standards (called ‘EBTP standards’), which require delivery of an evidence-based package of care, informed by NICE guidance and underpin the Seven Day Services Clinical Standards for the NHS.

Treat As One[12]

1.1.19 Published by the National Confidential Enquiry into Patient Outcome and Death in 2017, this report studied a cohort of patients aged 18 years and older, who were admitted to a general hospital for a physical health condition but also had a significant, known mental health condition. In just one month in 2014, there were 11,980 patients who met the criteria in the UK.

1.1.20 The report made 21 recommendations including:

• National guidelines should be developed outlining the expectations of general hospital staff in the management of mental health conditions.
• In order to overcome the divide between mental and physical healthcare, liaison psychiatry services should be fully integrated into general hospitals.
• Record sharing (paper or electronic) between mental health hospitals and general hospitals needs to be improved.

National Confidential Inquiry into Suicide and Homicide (NCISH) Annual Report 2017[13] (B)

1.1.21 The NCISH was commissioned to undertake the Mental Health Clinical Outcome Review Programme on behalf of NHS England.

1.1.22 The 2017 annual report provides findings relating to people who died by suicide or were convicted of homicide between 2005 and 2015 across all UK countries.

1.1.23 It reported that between 2005 and 2015 , NCISH was notified of 49,545 deaths in the general population of England that were registered as suicide or “undetermined”, an average of 4,504 per year.

National Suicide Prevention Strategy in England[14]

1.1.24 Having identified people who self-harm as a high-risk group in 2012, this 2017 report recognised the need to address self-harm as an issue in its own right. The scope of the government’s national strategy was expanded to include self-harm as a key area.

2.1.1 Diane was a 57-year-old woman with a history of mental illness who died by suicide. Almost two years prior to her death, Diane moved to a new part of the country to be near a close relative, who became her carer.

2.1.2 Diane had a diagnosis of mixed anxiety, depression and agoraphobia with an appropriate prescription of medication.

2.1.3 Over the next 22 months, Diane accessed various local mental health services, including:

• Community mental health team (CMHT)
  A multi-disciplinary team of specialists providing services to adults of working age with moderate to severe mental health problems or needs.
• Care co-ordinator
  A nominated member of the CMHT, often a social worker or mental health nurse, who should build a helpful and supportive partnership with the service user. The care co-ordinator should discuss plans, give counselling, information and advice and ensure there is a clear ‘care plan’ that sets out how the service user is going to be helped.
• Crisis resolution and home treatment team (CHRT team)
  Commonly referred to as the crisis team, it consists of mental health professionals i.e. psychiatrists, mental health nurses, social workers and support workers, who can support someone in their home. The CHRT team will decide if someone in crisis should be admitted to hospital or offered alternative treatment in the community. The crisis team will also offer home support to help patients leave hospital.
• Psychiatric liaison services (PLS)
  A psychiatric liaison team works across all departments in general hospitals (i.e. not mental health units). They provide psychiatric assessment and treatment to those patients who may be experiencing mental health problems while in general hospitals and provide the interface between mental and physical health. PLS models are covered in more detail later in the report in section 5.2.

A detailed chronology of events is given in Appendix 1. The key events are described below.

2.2.1 A month after arriving to live with her carer, Diane was urgently referred to the community mental health team (CMHT) by a GP following a self-harm(C) incident and expressing suicidal thoughts. Diane was seen the next day by a consultant psychiatrist who diagnosed:

• mixed anxiety and depressive disorder
• agoraphobia
• traits of emotionally unstable personality disorder with a history of deliberate self-harm

2.2.2 The planned treatment included a change to Diane’s medication, a referral to the crisis team, allocation of a care co-ordinator within the CMHT, a self-referral for cognitive behavioural therapy (CBT) and a review by a consultant after 14 days.

2.2.3 Diane did attend CBT and was assessed as having severe depression, severe anxiety and severe levels of psychological distress. This was communicated to the GP and CMHT.

2.2.4 A month later, Diane called the crisis team having self-injured and the ambulance service was called. After dressing her wounds, the attending paramedic determined that Diane was now less anxious and resting, so no further emergency treatment was provided.

2.2.5 The paramedic discussed this with the crisis team and the CMHT was informed.

(C) The term self‑harm is defined in NICE Quality Standard 34 as any act of self‑poisoning or self‑injury carried out by a person, irrespective of their motivation. This commonly involves self‑poisoning with medication or self‑injury by cutting. Clinical Guideline 16 adds self-harm is an expression of personal distress, not an illness, and there are many varied reasons for a person to harm him or herself.

2.2.6 Following heavy alcohol intoxication, Diane self-harmed again and said she was having suicidal thoughts. Her carer called 999 and Diane was taken to hospital by ambulance, arriving shortly after 09:00hrs. She was triaged by emergency department staff and a local safeguarding and managing risk tool (SMaRT) form(D) was completed.

2.2.7 Contact was made with the psychiatric liaison service (PLS) and at 10:15hrs Diane was seen by a mental health nurse in the emergency department. She was then referred to the crisis team and the hospital alcohol dependency team.

2.2.8 An assessment conducted by the crisis team at 20:35hrs did not recommend admission to hospital or a home treatment plan.

2.2.9 Diane was seen by the hospital alcohol dependency nurse and advised to seek help for alcohol detoxification. Following her discharge from the emergency department, the PLS shared the assessment with the CMHT and her care co-ordinator followed this up the next day.

2.2.10 Twelve days later Diane was admitted for a period of residential alcohol detoxification. She experienced a period of improved mood and stability following this treatment.

2.2.11 Six months later Diane was showing signs of increased anxiety. Her carer believed that this was in response to episodes of ‘paralysis’ induced by stress, which Diane believed were not being considered in her treatment.

(D) SMaRT was a locally produced form which used observed behaviours and symptoms to allocate a level of risk and suggest management strategies. See section 5.4

2.2.12 Diane presented several times at her GP for self-harm and described her anxiety levels as “creeping up”. The GP was aware she was receiving ongoing mental health care from the CMHT.

2.2.13 Following an unusually long and intense episode of ‘paralysis’, Diane dialled 999 in the middle of the night. The crisis team attended and determined that no immediate response was required. At this time Diane was calm, and the crisis team planned to update the CMHT in the morning.

2.2.14 The next day the care co-ordinator tried, but failed, to contact Diane. By mid-morning Diane returned the call, reporting she had taken a significant overdose of paracetamol. The CMHT called 999. Within two hours Diane had been transported to hospital and triaged in the emergency department. She was placed on a trolley in the area of the emergency department reserved for those requiring minor treatment and a SMaRT form was completed. A note was made on the clinical record of Diane’s medical history.

2.2.15 After five hours in hospital Diane removed the cannula and tried to leave the emergency department. She was persuaded to stay by an emergency department nurse. Diane was transferred to the observation bay overnight and an intravenous infusion for her overdose was re-introduced.

2.2.16 Medical notes made almost 20 hours after admission state “the PLS liaison have accepted referral and agreed to make contact to see if she is willing to be assessed prior to her treatment finishing”. A psychosocial assessment was completed by the PLS with a referral to the CMHT.

2.2.17 Diane was discharged from hospital after 30 hours. The next day she received a follow-up call from the CMHT and saw her GP. Diane was invited to attend anxiety management classes which were due to start the following month.

2.2.18 Leading up to event four, Diane had been attending anxiety management classes, but it was noted these classes increased her anxiety. During this period her care co-ordinator had also changed. There were multiple missed appointments and re-appointments for her anxiety management therapy. This resulted in Diane being taken off the list for the therapy. Her carer recalls a conversation with Diane about her being told she was “non-compliant” and that she had been “discharged” from care.

2.2.19 Diane reported to her care co-ordinator that her anxiety levels were “creeping up”; however, she was not considered high risk by the mental health trust.

2.2.20 The fourth event involved Diane going to the local railway station with a plan to take her life. A rail employee intervened and coaxed her to safety. Diane was taken to the local emergency department by ambulance with the police in attendance. Diane was triaged, it was noted she was very drowsy and wanted to go home. A locally produced SMaRT form was completed and Diane was assessed as ‘amber’(E).

2.2.21 Limited details of her previous attendance and medical history were noted. One set of observations of Diane’s physical health was completed at triage, together with an electrocardiogram (ECG) and blood samples. She was not referred to PLS.

2.2.22 She left the department after approximately three hours. A missing persons procedure was commenced, and the police were informed.

2.2.23 Diane’s carer contacted the CMHT an hour after Diane was reported as missing from the emergency department to advise the CMHT that Diane was at home; she did not return to the hospital and there were no medical entries in her notes other than her discharge from care.

2.2.24 There was no follow-up by the hospital to the CMHT, although the GP was notified of the attendance.

2.2.25 Diane saw her GP three days later and discussed the incident and hospital attendance. They talked at length about her bouts of paralysis and her anxiety.

2.2.26 There was no contact between the CMHT and Diane for a further six weeks.

2.2.27 Diane received a home visit from her care co-ordinator. During the visit Diane commented she felt suicidal most of the time but had no plan or intent at that moment.

2.2.28 The next morning Diane presented to the GP, reporting she had taken an overdose. She was advised by the GP to go to the local emergency department. Diane’s carer was called to take Diane to the hospital, but Diane refused, stating “There’s no point, they don’t do anything - I still won’t get any help. All that will happen is I’ll be there for hours and hours and then they’ll send me home.” Concerned for Diane’s wellbeing, her carer called 999 and Diane arrived at the emergency department at 20:19hrs.

2.2.29 The emergency department was experiencing high levels of activity and was on ‘black status’, indicating prolonged pressure on services which had been escalated to senior managers.

2.2.30 After a wait of almost an hour, Diane was triaged and admitted to the majors(F) department. A set of observations and a venous blood gas sample were taken. With a raised lactate level, the administration of PlasmaLyte intravenous fluids was commenced.

2.2.31 The hospital had a psychiatric liaison team, available between 08:00hrs and 23:00hrs. No referral was made to the psychiatric liaison team and no SMaRT form was completed. A nurse remembered talking to Diane and giving her a hot drink. The investigation could not identify any record of Diane being seen by an emergency department doctor or any medication chart for the intravenous fluids.

2.2.32 The next entry in the patient record at 02:37hrs states, “the patient did not wait”. The department attempted to contact Diane without success. No missing persons procedure was initiated. At 09:40hrs the next day, Diane presented at her GP practice for a repeat prescription of medication. The GP who saw her was reluctant to prescribe under the circumstances and deferred the decision for review later that day.

2.2.33 In the early afternoon, Diane left a voicemail message for her carer and a note on the railway station platform before climbing down onto the tracks and into the path of an oncoming train. Following treatment at the scene, she was airlifted to a major trauma unit, where she died of her injuries.

3.1.1 HSIB conducted a review of the Strategic Executive Information System(G) (StEIS) for incidents involving the provision of mental health care to service users in, or who have recently been discharged from, an acute hospital. A reference safety event was identified and, with the full co-operation of the Trust involved, a scoping investigation commenced.

(G) StEIS is a national reporting system that captures all Serious Incidents.

3.2.1 Following the scoping investigation, the HSIB’s Chief Investigator reviewed the findings and authorised a full investigation as the incident met the following criteria:

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

3.2.2 For someone suffering a mental health crisis, the emergency department is frequently the last resort because it is the only 24/7 service. Emergency department staff are trained and equipped to deal with physical trauma, but the increasing prevalence of mental health problems requires staff to have a broader skillset.

3.2.3 Research has shown:

• Individuals who attended hospital following self-harm in Manchester and Salford were up to 20 times more likely to die by suicide than the general population, 35% of those within six months of the episode of self-harm[15].
• 68% of people who had died by suicide within three months of contact with mental health services had a history of self-harm[16].
• The risk of suicide within the first year following self-harm is 49 times greater than that in the general population risk[17].

3.2.4 The report Right Here Right Now[18] found that only 35% of people with urgent mental health needs received the help they needed in a timely way in the emergency department.

3.2.5 Between 1 April and 21 June 2017 there were nine cases reported to national incident reporting databases of service users attempting to take their own lives on or near hospital premises while undergoing treatment.

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

3.2.6 In 2015, 4,820 people died by suicide in England[19]. Hospitals in England deal with 220,000 episodes of self-harm each year[20]. There has been an increase in the number of people seen in an emergency department for self-harm in the three months before suicide, from 11% in 1999 to 18% in 2013[21].

3.2.7 Despite National Institute for Health and Care Excellence (NICE) guidance that all episodes of self-harm should receive a full psychosocial assessment by a mental health professional[22], referral rates to liaison health services for psychosocial assessment varied from 22% to 88%[23]. Another study showed that 83%[24] of service users were discharged without being referred to liaison mental health services.

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

3.2.8 HSIB identified the potential to develop an understanding of the variability in the provision of care to service users with mental health problems in the emergency department. By studying where things have gone wrong as well as where there is good practice, the investigation aimed to highlight where the system supports practitioners in providing the safest possible care to patients.

3.2.9 Understanding how healthcare professionals can identify and control risk in the emergency department and provide a more therapeutic environment for those experiencing a mental health crisis, could have a significant impact on outcomes for patients presenting at the emergency department when experiencing a mental health problem.

3.3.1 Methods used in this investigation included:

• With the co-operation of the general practitioner, the ambulance Trust, the mental health Trust and acute Trust involved, a review of patient records and relevant policies and procedures.
• An interview with Diane’s next of kin to seek their perspective on her care.
• A review of the Trust serious incident investigation.
• Interviews and group discussions with over 70 individuals and subject matter experts on the provision of mental health care in the emergency department.
• Observations and visits to emergency departments at seven sites with varying models of liaison mental health services (two large teaching hospitals, two district general hospitals, an inner-city hospital with similar demographic challenges to the reference case, and two hospitals in Wales).
• A review of national incident report databases to identify similar reported events.
• A literature review.
• Interviews and communication with relevant national organisations:
• NHS England – senior programme manager for mental health
• NHS Improvement – national clinical director for mental health
• National Confidential Enquiry into Patient Outcomes and Death (NCEPOD)
• National Confidential Inquiry into Suicide and Homicide (NCISH)
• National Institute for Health and Care Excellence (NICE)
• The charity Mind
• Wales Applied Risk Research Network (WARRN)
• Royal College of Emergency Medicine (RCEM)
• Royal College of Psychiatrists (RCPsyc)
• National Centre for Collaboration for Mental Health (NCCMH)
• Psychiatric Liaison Accreditation Network (PLAN)
• Suicide Prevention Information Network
• East and Mid Surrey Mental Health stakeholder group
• Healthwatch England

3.3.2 Timing

The case history was compiled largely using clinical records. Recognising the limits of human memory, interviews were of benefit to establish ‘what normally happens in the emergency department’ rather than ‘what actually happened on the night’.

3.3.3 Research

The observations and visits were undertaken to compare and contrast the performance of liaison mental health services in emergency departments with differing contextual factors in England and Wales. Findings were supplemented by insights from subject matter experts with extensive clinical experience of emergency medicine and liaison mental health services. The findings were also underpinned by a targeted literature review of significant material in this field.

3.3.4 Investigation scope

The initial investigation touched on the involvement of primary care, ambulance service, 999 pathways, mental health and crisis care in the community, Network Rail, the police service and the emergency department in a general hospital. Aware of the challenges in other areas of mental healthcare, HSIB applied the investigation criteria and subsequently bound the lines of inquiry to the assessment and management of risk in the emergency department.

4.1.1 An analysis of Diane’s case was completed to determine where safety controls in the self-harm pathway in the emergency department had worked and where they were absent. This analysis was not to establish causation, but to understand the systemic challenges when treating someone experiencing a mental health crisis when they present at the emergency department.

4.1.2 Within the local area there was a higher than average proportion of working age people with limiting long-term illness and a high number of drug and alcohol attributable hospital admissions (over 850 admissions/100,000 population).

4.1.3 Services at the Trust were commissioned by four different clinical commissioning groups. In the preceding months, services at the emergency care centre in a neighbouring hospital had been reduced to an urgent care centre. The result had been an increased in demand at the reference emergency department.

4.1.4 The mental health liaison service was commissioned for 15 hours a day (08:00hrs to 23:00hrs) and operated from its own, co-located premises on the hospital site. The team was consultant led and staffed by seven nursing and social work practitioners. It received between 170 and 190 referrals per month and this number had increased steadily over the previous year.

4.1.5 Staff recruitment in the emergency department was reported to the investigation as “challenging”. On the night of Diane’s final presentation, although there was a full complement of nursing staff within the emergency department, 50% were agency nurses. The high level of activity had been escalated to senior management. At the time Diane arrived at the emergency department there were 63 patients requiring treatment.

4.1.6 The investigation classified and analysed evidence in four categories (Figure 1).

• Individual actions
  Discreet events in the reference event timeline associated with decisions or actions by individuals.
• Preconditions
  Environment or contextual factors that influenced the outcome.
• Local organisation
  Factors in the local system which influenced the provision of care.
• National organisation
  Factors on a strategic level that influenced the provision of care.

FIG 1 Classification of evidence from the scoping investigation

4.1.7 The investigation analysed the local pathway in the emergency department to identify the steps that were designed to ensure a person in crisis received appropriate care. In safety terms, these steps were considered to be safety defences or barriers (Figure 2) that provided the structure needed to identify and control a safety event. In Diane’s case the safety event was considered to be a patient leaving the emergency department without an assessment by an appropriate specialist. Other steps were intended to recover a situation which had gone wrong or attempted to minimise the impact of the event.

4.1.8 The investigation assessed how effective the barriers were and considered whether they were supported by documentation, training and adequate resource.

4.1.9 This broad analysis gave sufficient information to allow the investigation to identify themes for learning that could then be tested in a wider national investigation.

4.1.10 The investigation conducted a review of the relevant literature to assess whether Diane’s case was representative of the wider national picture. The review found:

• Physical health and life expectancy are severely compromised in individuals who self-harm compared with the general population[25].
• The repetition of self-harm is common and occurs quickly – 10% of subjects repeat self-harm within five days of the index incident, 13.6% within 12 month[26].
• Non-fatal self-harm is one of the most frequent reasons for emergency hospital admission.
• Over half of patients who die by suicide have self-harmed at some time[27].
• Non-fatal self-harm is also the strongest risk factor for subsequent suicide[28].
• There is evidence that completing a psychosocial assessment reduces the risk of future self-harm[29].
• A systemic review of risk factors and risk scales[30] concluded that a comprehensive psychosocial individual assessment was the only clear intervention that was beneficial to the management of people who have self-harmed.

4.2.1 The investigation noted the joint serious incident (SI) investigation that had been conducted locally between the acute Trust and the mental health Trust.

The findings included:

• use of SMaRT/SMaRT plus tool
• omissions in care
• delay in the patient care pathway
• emergency department access to psychiatric liaison service (PLS)

4.2.2 Following the SI, the acute Trust took the following action:

- review and scoping of the redesign of SMaRT plus to include a plan section

- review of the monthly audit of records to enable capture of SMaRT completion

- review of the requirement for implementation of the Trust’s missing persons policy

Additionally, the Trust had already been granted funding for a 24/7 psychiatric liaison service at the hospital and had an emergency department improvement plan in place.

5.2.1 In 2013, the Royal College of Psychiatrists described liaison psychiatry for acute hospitals31 and the Developing Models for Liaison Psychiatry Services[32] report published in 2014, explored the model further.

5.2.2 Liaison mental health services are designed to operate outside of traditional mental health settings, routinely in emergency departments, wards and medical or surgical outpatient departments. In acute hospitals they address the mental health needs of people being treated primarily for physical health problems and symptoms. The teams that provide this service go by various names including liaison psychiatry, psychiatric liaison services, psychological medicine and emergency department psychiatric services. The evidence based treatment pathway (EBTP) refers to liaison mental health services.

5.2.3 Effective liaison services are multidisciplinary, led by a consultant psychiatrist, and supported by specialist nurses and psychological therapists. The number of practitioners is dependent on the hours of service provision and the functions it performs[33]. The capability of the services is described as either ‘Core’, ‘Core 24’ (described below), ‘Enhanced 24’ or ‘Comprehensive’[34].

5.2.4 The Five Year Forward View set out the government’s commitment for 24-hour, all age, liaison mental health services in all emergency departments in England by 2021, with half of them meeting the ‘Core 24’ specification for the staffing of services for adults. NHS England shared with the investigation a longer-term ambition that all type 1 emergency departments(H) should be Core 24.

5.2.5 To meet the minimum Core 24 criteria[35], a service must be:

• commissioned to operate 24/7, as an on-site distinct service
• sufficiently staffed with appropriately skilled people
• able to respond to emergency referrals within an hour, and urgent referrals from inpatient wards in the wider hospital within 24 hours

5.2.6 In the 3rd Annual Survey of Liaison Psychiatry in England, 201636, 53% of the hospitals that responded had mental health staff on site 24/7. Not all of these services were staffed to meet the Core 24 standard.

(H) Emergency departments which are consultant led and provide a 24 hour service with full resuscitation facilities and designated accommodation for the reception of accident and emergency patients

5.2.7 Diane was referred to liaison mental health services on two of her four visits ’to the emergency department.’

5.2.8 There was no clarity among the emergency department staff about the operating hours of the mental health liaison service. The service operated between 08:00hrs and 23:00hrs. However, some staff commented to the investigation that a referral was futile after 21:00hrs.

5.2.9 The liaison team had offices in a building owned by the mental health Trust adjacent to the hospital. There was no access to electronic mental health records in the emergency department, which meant the liaison team staff had to return to their office to write up notes.

5.2.10 The peak time for referrals to the liaison mental health service was 08:00hrs at the start of the working day when service users who had presented overnight were first able to see a mental health professional.

5.2.11 When Diane was referred to the liaison service and received a full psychosocial assessment and a referral from a mental health professional as part of her treatment in the emergency department, the outcome was a successful de-escalation of her mental health crisis. Communication with the community mental health team worked effectively followed by a period of relative stability.

5.2.12 When the liaison service was not involved, the only follow-up to the presentation was a standard notification from the emergency department and ambulance service to the GP. Diane did see her GP following her penultimate presentation at the emergency department, but the care co-ordinator remained unaware of her crisis.

5.2.13 The Joint Commissioning Panel for Mental Health[37] said of liaison mental health:

“There is currently no single, uniform model for liaison services across the country. Where such services exist, they are often provided by the local mental health trust within the estate of the acute hospital trust, which may present logistical and operational challenges.”

5.2.14 The outcome of more recent government policy supporting the effectiveness of liaison mental health services, has resulted in significant increases in funding specifically for urgent and emergency mental health care.

5.2.15 The government spending review[38] in 2015, provided additional dedicated funding streams for core priorities. For mental health liaison services in emergency and urgent care, one of the most notable commitments was the release of transformation funding by NHS England in two ‘waves’ to accelerate provision of services in those emergency departments closest to the minimum Core 24 level.

5.2.16 The funding was to meet the specifications set down in the NHS Operational and Contractual Planning Guidance 2017-2019[39]. This stipulated that by 2020/21, 50% of acute hospitals across England will meet the Core 24 service standard as a minimum. At the time of writing, 11% had achieved this benchmark.

5.2.17 Refreshing NHS Plans for 2018/19[40] specified that each clinical commissioning group (CCG) must meet the Mental Health Investment Standard (MHIS). This required investment in mental health to rise at a faster rate than the overall programme funding.

5.2.18 The CCG Improvement and Assessment Framework[41] provides assurance to NHS England that NHS commissioning is in line with national directives, requiring CCGs to be audited to prove they are meeting the MHIS.

5.2.19 NICE guideline NG94 acknowledged the lack of evidence on cost-effectiveness or optimal models of care[42]. Measuring the outcomes and performance in liaison mental health services is challenging; interventions are complex and supported by many services, making it difficult to isolate specific impacts. The focus of liaison services differs depending on whether they are working with emergency departments, acute hospital wards or outpatient services[43].

5.2.20 The investigation heard from commissioners who had adopted a short-term approach to liaison services with a one-year financing arrangement. Some acute trusts were employing temporary mental health specialists to react to surges in demand, while others were supplementing liaison services with mental health nurses on the emergency department staff. Two of the trusts visited had funding approved but could not recruit into the advertised positions. In a 2016, survey of Liaison Psychiatry in England, 53% of the hospitals that responded had mental health staff on site 24/7. Not all of these services were staffed to meet the Core 24 standard[44].

Without data to substantiate the demand on services it is difficult for Trusts to rationalise the case for investment.

5.2.21 A comparison of rates of self-harm based on routinely collected Hospital Episode Statistics (HES) against data based on detailed self-harm studies found HES may underestimate overall rates of hospital presentations for self-harm by around 60%[45]. Improvements to the Emergency Care Data Set proposed by NHS England to collect data on referrals, response times, interventions and repeat referrals will help illustrate the true extent of demand on services.

5.2.22 The National Suicide Prevention Strategy[46] also introduced an indicator within the Public Health Outcomes Framework in 2017 to ensure all Trusts collect data on emergency admissions to hospital for self-harm.

5.2.23 Commissioned by the NHS Institute for Research in 2015, the Liaison Psychiatry: Measurement and Evaluation of Service Types Referral Patterns and Outcomes (LP-MAESTRO) research programme aims to evaluate the cost-effectiveness and efficiency of particular configurations of liaison mental health services for specified target populations. It will report in late 2019 and may add weight to commissioning models.

5.2.24 When funding is based primarily on return-on-investment, cost avoidance and organisational efficiency, the complexities of capturing data to justify liaison services was reported as challenging[47]. The investigation heard funding models rely on proof of financial effectiveness to support commissioning rather than quality, safety and improved patient experience[48].

5.2.25 Improving provision of liaison mental health services has now been in NHS planning guidance for the past three years[49]. In 2016, over half of services reported they were better resourced than the previous year[50] (Figure 3). The investigation saw this as a positive picture overall, but auditing will be required to ensure that trajectory remains achievable.

5.2.26 The investigation identified examples of integrated teams working together and highlighted benefits, including:

• Access to specialist mental health professionals as an integral part of the emergency department team.
• Improved flow through the emergency department.
• Raised levels of general awareness about mental health among emergency department staff.
• Improved sharing of patient data.
• ‘On-hand’ advice about mental health legislation.
• Improved wellbeing of emergency department staff.

The investigation was also made aware of some challenges:

• Liaison teams that could not get office space in or near the emergency department.
• Liaison teams that chose to be located away from the emergency department.
• Liaison teams that were 24/7 but were seen to be ‘less present’ during the night by emergency department staff.
• Strain between emergency department staff and liaison staff around response times to referral and breaches to the four-hour standard.
• IT systems that were incompatible.
• An ‘us and them’ culture between emergency department and mental health staff with hierarchies causing communication difficulties between emergency department doctors and mental health nurses.

5.2.27 The annual survey of liaison psychiatry[51] made the following observation:

“...there is an unfortunate distinction between ‘doing Liaison’, which would include seeing a patient in an acute hospital, and ‘being Liaison’ which is being a team which undertakes such clinical activity.
It is thought that the benefits of Liaison are enhanced when the mental health workers are well known to other professionals in the acute hospital. This is ‘being Liaison’ and is unlikely to be emulated by visiting professionals who merely ‘do Liaison’ and are based elsewhere”.

5.2.28 Many liaison services are funded and employed by the local mental healthcare provider and therefore integration in the general hospital governance structure is not a given. Recent advice states “liaison mental health services should have joint ownership and governance arrangements between acute trusts, mental health trusts and other local providers including senior clinical and operational leadership from those providers” [52].

5.2.29 The Psychiatric Liaison Accreditation Network (PLAN)(I) and the Care Quality Commission[53] both use integration as a measure of performance. When liaison services are involved on committees, in teaching, writing policies or incident investigation, this is seen as a positive indicator.

5.2.30 The Academy of Medical Royal Colleges in 2008[54] recommended:

“Clearer standards would help in making explicit the responsibilities of commissioners in ensuring service delivery. At present there is no clarity regarding how these services should be commissioned. Should they be commissioned via acute services, or mental health services? We recommend that national guidance, alongside national standards, is issued to commissioners.”

5.2.31 The Joint Commissioning Panel for Mental Health’s guidance to commissioners (2013)[55] stated:

“A good liaison service functions best as a discrete, specialised, fully integrated team comprising multi-professional health care staff, under single leadership and management. … There should be universal agreement to commission liaison services as part of the acute hospital care commissioning process"

(I) PLAN - A series of standards and quality assurance measures for liaison mental health services which enable a trust to become accredited.

5.2.32 The investigation observed various liaison service models, including the model employed in Wales where services were integral to the health board. There was a general acceptance around the evidence base for liaison psychiatry put forward by Aitken in 201456. The investigation found no consensus on commissioning models. The most common model for service provision was for the mental health Trust to provide services in the acute Trust.

5.2.33 Healthcare professionals commented to the investigation that the challenge for such a model revolved around the need for clear lines of clinical supervision and governance for professional competency and the need to have close ties to the local referral network. The perception among mental health professionals was that independence afforded them a buffer against the pressures of meeting emergency department targets. The perception of those responsible for managing care within the emergency department focused on the ability to influence a service from an external organisation and the impact on emergency department four-hour performance standards.

5.2.34 Liaison mental health services are a positive influence on managing the safe care of patients in the emergency department. They have the greatest impact when services have a permanent presence in the emergency department and are part of the multidisciplinary team.

5.2.35 The effectiveness of the liaison mental health team in treating Diane was influenced by the hours of operation, their presence and integration within the emergency department. Early intervention of trained mental health professionals had a positive effect on outcomes for Diane.

5.2.36 Provision of liaison mental health services is variable across England and there is no consensus on commissioning models.

5.3.1 HSIB investigators observed various mental health presentations in emergency departments. They also reviewed anecdotal evidence and reports of people experiencing mental health problems who had come to harm while receiving urgent and emergency care in, or having absconded from, acute hospitals.

5.3.2 The following is a service user’s perspective on the pathway:

“I came to A&E two days ago, I was feeling suicidal…I did not know where to get help for the way I was feeling and just came to A&E, I did not know that a specialist team was based there, it was random…when I first approached the receptionist at A&E, I just told her how I was feeling, she did not ask too many questions…. I went to see a nurse/ doctor initially, they checked me out and took us to another waiting room, they did not inform me who I was going to see and how long I would wait for, I had no idea that I would be seeing someone from the [mental health liaison] team…also no one from the [mental health liaison] team spoke to me whilst I was waiting. The waiting room was not ideal…we heard nurses joking and laughing …we did not want to be there…the environment of A&E and A&E staff made me feel fidgety and anxious…”

5.3.3 Every presentation to the emergency department following self-harm represents an opportunity to intervene and improve outcomes. Of note was the variability with which the person arriving in the emergency department was assessed and then signposted to the appropriate healthcare professional.

5.3.4 HSIB investigators consulted with healthcare professionals on where the variability occurred and originated. The interpretation and understanding of guidance around the initial assessment was highlighted.

5.3.5 NICE produces clinical guidelines on the management of clinical conditions.

5.3.6 Guidelines are recommendations that are not mandatory. Guidelines should be taken into account, but they do not override the responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual.

5.3.7 NICE Quality Standards are based on the recommendations in the Guidelines and set out the priority areas for quality improvement.

5.3.8 There is no single guideline for the delivery of mental health care in the emergency department. The investigation reviewed many guidelines for their applicability to the investigation and acknowledged the varied causes for service users presenting to the emergency department when in crisis. Notwithstanding that, given its relevance to the reference event, the investigation focused on the guidance for self-harm:

• NICE Quality standard (QS34) Self-Harm (2013)[57]
  NICE quality standards describe the aspirational indicators of high quality that a service user should expect during their treatment.
• NICE Clinical Guideline (CG16) Self-harm in the over 8s: short-term management and prevention of recurrence (2004)[58]
  This covers the treatment of people in the first 48 hours following self-harm. It was last reviewed in 2016.
• NICE Clinical Guideline (CG133) Self-harm in over 8s: long-term management (2011)[59]
  This guideline follows on from CG16 and deals with the longer-term psychological treatment and management of self-harm. It specifically does not address practice in the emergency department.
• Achieving Better Access to 24/7 Urgent and Emergency Mental Health Care (2016)
  The evidence-based treatment pathway (EBTP) was published by NHS England in 201660. It covers the first 24 hours of care following presentation. The pathway includes recommended standards which require delivery of an evidence-based package of care, informed by NICE guidance.

5.3.9 As the first point of contact with a service user, the initial assessment is the critical opportunity to recognise a mental health need and provide

QS34 defines an initial assessment as:

“... the first assessment by a healthcare professional after an episode of self harm. It should be undertaken each time a person presents following an episode of self harm.”

5.3.10 The NICE standards, guidance and the EBTP describe a two-stage process:

• The first stage is conducted by emergency department staff when the service user initially presents. It is variously called triage, initial assessment or a basic review.
• The second stage is a full psychosocial assessment conducted by a mental health professional and/or specialist management of any medical condition.

5.3.11 Diane’s four presentations emphasised the potential for variability in initial assessments. She was dealt with differently on each occasion.

5.3.12 Diane’s physical health needs were generally well recorded and scored using clinical screening tools. Appropriate tests were completed and, where applicable, the overdose was recorded as the reason for presentation.

5.3.13 Although her self-harm was noted, there was no process for the triage nurse when assessing mental state. The information recorded about Diane’s mental state during initial assessment was unstructured and lacked detail. Following triage, the next assessment in three of the four presentations was by a doctor when she had been in the emergency department between one to three hours. This clinical assessment addressed Diane’s mental health in greater detail and preceded the start of treatment.

5.3.14 Diane was noted as saying she “wanted to go home” during triage. Her reluctance to remain for treatment in the department was not highlighted, nor prioritised. There was no mention of mental capacity during assessment, but her clothing and appearance were noted.

5.3.15 A record of Diane’s previous attendance was available on the emergency department records. A check of previous attendance notes would have revealed previous attempts to leave the emergency department before treatment was complete. There was no process to check previous records. The patient notes were of a varying standard. Some were handwritten and others electronic. The handwritten records were more comprehensive than those recorded electronically. The local investigation heard from staff that when the emergency department was very busy, there were not enough computers for staff to make notes.

5.3.16 Diane’s mental health records were held on a computer system to which the emergency department staff had no access. Until the involvement of the liaison mental health service, the emergency department staff were unaware of Diane’s mental health history, including her diagnosis of emotionally unstable personality disorder, anxiety and her history of self-harm and suicidal thoughts.

5.3.17 The assessment of mental state was inconsistent across Diane’s four presentations. The investigation was unable to identify the processes or structure of the assessments, either in the reference case or other similar events observed during the investigation.

5.3.18 The standards for measuring physiological parameters in emergencies such as stroke, heart attack or sepsis are established in care bundles and pathways. These pathways have been successful in reducing mortality. The investigation noted evidence of this structured approach in the treatment of Diane’s physical health. The guidance for a paracetamol overdose, from collecting samples and screening through to overdose management, was considered specific and comprehensive61. The process was applied consistently and was robust. Even when there were omissions in care by an individual, the system offered layered protections.

5.3.19 The investigation found the process for triage and initial assessment completed by emergency department nurses was effective at identifying physical health problems, but lacked structure and consistency when assessing mental state. The investigation also noted the initial assessment did not access any previous medical records.

5.3.20 The Initial Assessment of Emergency Department Patients[62] published by the Royal College of Emergency Medicine (RCEM) sets out a variety of methods for assessing patients on arrival at the emergency department.

5.3.21 Triage is described as “a face to face encounter which should occur within 15 minutes of arrival or registration and should normally require less than five minutes contact” [63].

5.3.22 Each hospital the investigation visited, placed clinical assessment at the front of their process. Emergency department staff explained the aim of this rapid assessment was to ensure the patient did not have a clinical condition that made it unsafe for them to wait in a queuing system but the rapid assessment was not designed to replace an initial assessment. Following triage, the investigation heard from practitioners that an emergency department doctor should assess the patient.

5.3.23 NICE quality standard QS34 states that healthcare professionals should “ensure that people who have self-harmed have an initial assessment after an episode of self-harm that includes physical health, mental state, safeguarding concerns, social circumstances and risk of further self-harm or suicide”. It goes on to define the constituent parts of initial assessment in more detail.

5.3.24 CG16 does refer to an initial assessment in primary care but promotes a process of triage in the emergency department:

• When an individual presents in the emergency department following an episode of self-harm, emergency department staff responsible for triage should urgently establish the likely physical risk, and the person’s emotional and mental state, in an atmosphere of respect and understanding.
• Emergency department staff responsible for triage should take account of the underlying emotional distress, which may not be outwardly exhibited, as well as the severity of injury when making decisions about priority for treatment.
• All people who have self-harmed should be offered a preliminary psychosocial assessment at triage (or at the initial assessment in primary or community settings) following an act of self-harm. Assessment should determine a person’s mental capacity, their willingness to remain for further (psychosocial) assessment, their level of distress and the possible presence of mental illness.

5.3.25 The EBTP refers to the initial assessment as a basic review which comprises:

• a physical assessment; a decision as to whether they need emergency physical care should be taken as a priority
• a personalised risk assessment, including a decision as to the appropriate action needed should the person leave the ED while waiting for review by the liaison mental health team
• observations on behaviour and mental state

“If there is a lack of clarity about whether an urgent or emergency mental health response is needed, staff should call the liaison mental health team for advice.”

The pathway is expanded in a resources document which refers to a parallel process and not a basic review (Figure 5).

5.3.28 The investigation aimed to understand the process for assessment on initial presentation. First contact was variously called triage, initial assessment, basic review and a parallel process. The content and emphasis of those elements were different. The investigation compared the guideline for the assessment and diagnosis of chest pain against the guidelines for self-harm and found the latter offered little detail for health professionals on the process for mental health assessment.

5.3.29 NICE guideline CG16 dates back to 2004 with limited amendment in 2011 and a review in 2016 which found no new evidence. The focus is on a ‘traditional’ triage process, which introduces a concept of an optional preliminary psychosocial assessment for elements that are inherent in the definition of initial assessment found in quality standard QS34. The long-term management of self-harm was added as an additional stand-alone guideline in 2011.

5.3.30 With the increasing presence of liaison mental health services in acute hospitals, the more recent EBTP and the RCEM guidance highlights the advantages of a parallel approach to physical and mental health care[65]. This is further reflected in March 2018 NICE guidance on the service delivery and organisation of emergency medical care[66].

5.3.31 The investigation found there to be different guidance available from various national bodies. The guidance to emergency department staff for initial assessment lacked coherence and consistency.

5.3.32 In complex systems, like the emergency department, regulations and procedures cannot account for every eventuality and staff must demonstrate flexibility to deviate from a planned task[67] to achieve the aim. Over time, this may lead to “the slow steady uncoupling of practice from written procedure”[68] as staff find short-cuts and work-arounds. Behaviour that is “acquired in practice and is seen to work” becomes legitimised and adopted as standard practice.

5.3.33 Guidelines for the treatment of self-harm make up a fraction of the guidance emergency department staff are required to understand and work with. The large number of guidelines and different sources of information in healthcare make it challenging for staff to comply with or be aware of them all[69]. Only two-thirds of respondents to a Royal College of Psychiatrists survey reported having read the self-harm guideline[70].

5.3.34 The Clinical Human Factors Group lists 16 reasons why workers may not comply with rules[71].

Of these, the investigation could identify at least 10 in the reference event and the wider investigation into mental health assessment, including:

• a lack of awareness/understanding of policies and procedures
• misperception of risk
• time pressure/pressure to get the job done
• lack of end-user engagement when policies and procedures are written
• policy and procedure overload
• ambiguous or conflicting messages in the policy/procedure
• lack of training and reinforcement of key policy messages over time
• and a mismatch between the policy and how the job is actually done

5.3.35 Rasmussen identified a phenomenon which he called “drift to danger”[72]. The concept of procedural drift developed[73] and more recently Dekker built on these ideas[74]. Procedural drift is represented in Figure 6 in the top diagram the baseline represents a minimum level of safety that is defined by unambiguous rules, policies and regulation. In day-to-day business, behaviours may drift below the required standard for the reasons outlined above. The drift can be detected and corrected. On the lower graph, the less defined baseline represents a situation where an activity is less prescriptive, and rules are vague and open to interpretation. There can be a significant drift in behaviour that is inherently less safe but is interpreted as ‘within the rules’. In such circumstances, less safe practices are more difficult to detect.

5.3.36 The investigation identified that emergency department staff were exposed to the challenges of lengthy and complicated guidelines about the diagnosis and treatment of numerous conditions. At the same time staff had to interpret a self-harm process that lacked detail and coherence. The lack of detail in the guidance contributed to the variability in initial assessment.

5.4.1 There does not appear to be any consistency in the approach used to assess risk in emergency departments and no standardisation of risk assessment instruments. All seven hospitals visited during the investigation used a different locally-produced risk assessment form.

5.4.2 The EBTP, NICE guideline CG16 and RCEM all recommend a risk assessment by emergency department staff during initial assessment. NICE quality standard QS34 recommends that the initial assessment for those who present having self-harmed includes “…risk of further self‑harm or suicide”.

5.4.3 NICE quality standard QS34 and guideline CG16 also call for an assessment of risk by specialist mental health practitioners during the psychosocial assessment that occurs later in the patient pathway. The investigation highlighted this as an example of common terminology being used by two different groups of healthcare professionals to describe different processes in the same pathway.

5.4.4 The Trust used a locally produced tool to assess and manage risk (Figure 7).

Fig 7 A representation of the locally produced Safeguarding and Risk Management Tool (SMaRT) in use at the time of the reference safety event.

Tick the boxes of factors present - manage as per the highest level of risk identified AND document in care plan

Always consult your Matron/Site Clinical Manager if you are unsure or concerned about an individual’s level of risk

5.4.5 SMaRT was used as a referral from the emergency department to liaison services. Not all emergency department staff were familiar with it. Staff told the investigation that they perceived risk management strategies that required observation of patients as unachievable, due to workload and competing demands.

5.4.6 Diane was categorised as an ‘amber’ risk on three occasions when she had self-harmed, indicating a probable risk (greater than 50% chance) of danger to self or others.

5.4.7 At the bottom of the list of mitigations was the advice “refer to Psychiatric Liaison Service via pager ASAP”. The investigation considered it possible that an individual tasked with completing the form may not have noticed this item placed at the bottom of a ‘list’ of actions that they considered largely unachievable.

5.4.8 The investigation found the SMaRT was poorly understood and served mostly as a ‘tick box’ referral for liaison mental health services. In attempting to assess the severity and likelihood of an outcome of self-harm or harm to others it did not consider the risk of leaving the department.

5.4.9 SMaRT was locally produced to meet a perceived requirement. It was unvalidated but annotated “adapted from mental health triage for use with NICE guidelines on self-harm (NICE CG16 2004)”.

5.4.10 CG16 recommended ‘consideration should be given to introducing the Australian Mental Health Triage Scale, as it is a comprehensive assessment scale that provides an effective process for rating clinical urgency so that patients are seen in a timely manner’. See appendix 2a.

5.4.11 The triage scale was also endorsed by the RCEM as a means of initial assessment. The RCEM mental health toolkit contained a link to the ‘Australian Mental Health Triage’. See appendix 2b.

5.4.12 When describing a full psychosocial assessment, CG16 points to advice in the guidance for the long-term management of self-harm (CG133), which states:

Risk assessment tools and scales are usually checklists that can be completed and scored by a clinician or sometimes the service user depending on the nature of the tool or scale. They are designed to give a crude indication of the level of risk (for example, high or low) of a particular outcome, most often suicide.

Do not use risk assessment tools and scales to predict future suicide or repetition of self-harm.

Do not use risk assessment tools and scales to determine who should and should not be offered treatment or who should be discharged.

Risk assessment tools may be considered to help structure risk assessments as long as they include the areas identified in recommendation 1.3.6.

5.4.13 NICE quality standard QS34 recommended that initial assessment include the risk of further self-harm or suicide. Although dealing with short-term risk, the investigation considered SMaRT, as used by the Trust in the reference case, attempted to link symptoms and behaviours to the likelihood of future harm.

5.4.14 The investigation found there was the potential for misunderstanding in the guidance around interpretation and use of the Australian mental health triage tool.

5.4.19 Research shows poor reliability of tools and scales for predicting self-harm and future suicide despite the recommendations in national guidelines.

5.4.20 Any new guidance should be clear on the purpose and bounds of any assessment and any tools/checklists recommended should be subject to evaluation.

5.5.1 Having identified the risks, the requirement on emergency department staff is to establish a management plan that provides appropriate treatment and keeps the service user safe while in their care.

5.5.2 Royal College of Emergency Medicine (RCEM) guidance on mental health recognises “there is no question that the middle of an ED, whether busy or quiet, can be a very stressful environment for any patient. However, if a person is feeling paranoid, psychotic, distraught or suicidal, the environment can be clearly detrimental, and could potentially escalate symptoms”.

5.5.3 Diane spent between three and 30 hours in the emergency department during her visits. The investigation heard from her carer how Diane spoke of not wanting to be a burden.

5.5.4 In June 2018, the RCEM issued a safety alert about the threat of patients absconding from the emergency department[79].

5.5.5 NICE clinical guideline CG16 states:

‘If a person who has self-harmed has to wait for treatment, he or she should be offered an environment that is safe, supportive and minimises any distress. For many patients, this may be a separate, quiet room with supervision and regular contact with a named member of staff to ensure safety[80]’.

NICE Quality standard 34 for self-harm states:

‘People who have self harmed receive the monitoring they need while in the health care setting, in order to reduce the risk of further self harm.[81]’

Highlighting consideration for the safe environment those who have self-harmed, it also states:

“Examples of environmental risks to people who self harm include but are not limited to: ligature points - open windows - access to sharps - access to medication.[82]”

5.5.6 When assessed as an ‘amber’ risk, Diane should have been kept within eyesight at all times.

5.5.7 There was no recorded plan for observations.

5.5.8 The emergency department had five exits and members of the public, visiting or accompanying other patients, had access to majors and minors nursing areas.

5.5.9 If special observation was required the Trust had an observation plan, however the Trust had only one member of security staff on duty and the lead time for agency staff meant the immediate need could not be met.

5.5.10 There was no dedicated area of safety or nominated treatment bays for those experiencing a mental health crisis. The emergency department had a mental health assessment room, but this was only used for full psychosocial assessment. On her final presentation, Diane was seated in a chair next to the nurse’s station and next to another service user in crisis - so-called ‘cohorting’. This was in the middle of an extremely busy department but was considered the most appropriate solution to keep her under observation.

5.5.11 The investigation heard how those in crisis often want to smoke cigarettes. To do this they were required to leave the building. Without an accompanying friend or relative to monitor their safety, nurses were faced with the dilemma of how to balance this personal need, which often reduced anxiety in the service user, with the threat of absconding or harm.

5.5.12 The investigation consistently heard that the emergency department was the right place to treat the physical aspects of self-harm, but the environment was considered sub-optimal for those experiencing mental health problems.

5.5.13 The investigation visited a modern, purpose-built mental health hospital to compare what was deemed to be a safe environment for patients who are in crisis and admitted to hospital for treatment. The investigation observed measures to minimise opportunities for self-harm and the use of design to minimise confined spaces and improve visibility. Areas were decorated and lighting was used to create a calm environment.

5.5.14 When the investigation visited emergency departments, the assessment rooms were often presented as the safe environment for mental health patients.

5.5.15 The Psychiatric Liaison Accreditation Network (PLAN) is part of the Royal College of Psychiatrists. It defines the accreditation standards for liaison services.

5.5.16 The availability of facilities to conduct high-risk assessments is a benchmark set by PLAN[83]. Located within the emergency department, the psychiatric assessment room should provide a safe, ligature free environment, appropriately decorated to provide a sense of calm, plus specific requirements for access doors, observation and staff safety.

5.5.17 When referring to an appropriate environment for clinical assessment PLAN highlights:

“The use of a curtain around a patient’s bed does not ensure privacy and should only be used rarely, and as a last resort, i.e. if there is significant risk and no safe alternative room, or if it is not physically possible for the patient to be moved to a more private setting[84].”

5.5.18 In a survey of assessment rooms undertaken in 2016[85], 85% of Trusts reported having a psychiatric assessment room; 23% of those respondents met all the criteria to be considered safe.

5.5.19 A common reason for a room being judged unsafe was because it was not a dedicated facility for conducting psychiatric assessments and was used by emergency department staff for other purposes, such as storing medical equipment or as an overflow waiting area for patients.

5.5.20 The NHS Improvement focus on improving patient flow states, “While waiting for assessment and treatment, to reduce their distress, during the assessment itself, patients should have access to a bespoke mental health assessment room[86].”

5.5.21 The investigation observed psychosocial assessment of ambulant service users undertaken in psychiatric assessment room. There were often patients waiting for assessment because there was only one room and/or one mental health professional to conduct assessments.

5.5.22 Assessment rooms were not observed being used as waiting areas for those experiencing a mental health crisis, patients were either in treatment bays, sat in chairs or on trolleys in the majors or minors department, or asked to remain in the main waiting area.

5.5.23 As a result, even when they had been triaged, those who were experiencing a mental health emergency and awaiting assessment by the liaison mental health service waited in the heart of the department.

5.5.24 Although most emergency departments reported having a psychiatric assessment room, only a 23% were considered compliant with accreditation standards.

5.5.25 The investigation observed psychiatric assessment rooms in frequent use but noted this accounts for only a small part of a service user’s stay in the emergency department.

5.5.26 The investigation observed no emergency department that had a dedicated treatment area for those experiencing a mental health emergency as envisaged by the guidelines.

5.5.27 Some vulnerable people discharge themselves from the emergency department because of excessive waiting times involved[87]. The RCEM defines absconding as a patient who ‘leaves the emergency department without informing staff and at risk of harm to themselves or others either through neglect or deliberate means[88]’.

5.5.28 The investigation heard from service users and their carers how waiting in a non-therapeutic environment could increase a person’s anxiety[89].

• “The A&E waiting area is not conducive for people in mental health crisis - it can get very busy and this can make patients feel more anxious.”
• “I’ve recently been in a situation with my son and we were left for hours in a waiting room after he self-harmed. He sat crying and clearly distressed”
• “They were very disrespectful, very rude and misbehaving…we did not want to be there…the environment of A&E and staff made me feel fidgety and anxious…”

5.5.29 When predicting those who might abscond, it lists mental health problems as a key predictor with factors including:

5.5.30 Verbalising a wish to abscond/high suicide risk/previously absconded/quiet/ withdrawn/unaccompanied/distressed/brought to ED against own wishes/has external commitments/stressors/alcohol or drug dependence.

5.5.31 RCEM’s Key recommendations include:

• “A key action to be performed as soon as risk of absconding has been identified is to undertake a capacity assessment. “
• “Emergency departments should have written guidance detailing specific measures which may be activated to prevent absconding.”
• “Emergency departments should have a specific form for detailing a patient’s physical features, if at high risk of absconding.”
• “Emergency departments should have written guidance on when it is appropriate to contact hospital security and the Police Service for patients who abscond.”

5.5.32 The charity Mind offers the following advice on its website about the emergency department: “A&E can be a difficult place to be in. If you can, ask a friend or family member to go along with you for some extra support. You might have to wait a while before a doctor can see you, so it can be helpful to have someone waiting with you.”

5.5.33 The threat of a patient who is experiencing a mental health crisis absconding from the emergency department and then coming to harm is significant.

5.5.34 The investigation also heard about the engagement of the third sector, or charities, in emergency department waiting areas to good effect. Other Trusts use support workers to make the environment less stressful by offering support while keeping them under observation.

5.6.1 Having successfully kept the service user safe in the emergency department, the next element in the care pathway, in accordance with NICE guideline CG16, is a comprehensive psychosocial assessment by a specialist mental health professional.

5.6.2 CG16 states that everyone who has self-harmed should have a comprehensive assessment of needs and risk by a specialist mental health professional.

5.6.3 There is no clear advice to this effect in the earlier part of CG16 which focuses on triage by emergency department physicians.

5.6.4 The investigation identified that the guidelines lacked detail. The investigation also noted a considerable amount of supporting literature on the subject of psychiatric risk assessment, which has been published by the Royal College of Psychiatrists.

5.6.5 The 2018, National Confidential Inquiry into Suicide and Safety in Mental Health study into the assessment of clinical risk in mental health services90 examined which risk assessment tools are currently used by specialist mental health services. It addressed how effectively these tools were being used prior to patient suicide, especially in patients rated as at low or no risk of suicide at their final contact with a mental health professional. Amongst its findings were:

• Risk assessment tools should not focus on prediction
• Risk is not a number, and risk assessment is not a checklist
• Risk assessment is not a stand-alone or one-off process

5.6.6 The evidence based treatment pathway sets out a clear course of action when a liaison mental health team is available.

5.6.7 In the reference case, the Trust response time from referral to psychosocial assessment was stated as two hours. The actual response time varied between 75 minutes and 20 hours. On the occasion when the response was protracted, Diane received treatment which lasted in excess of 16 hours and the liaison mental health service deferred the assessment until Diane was deemed ‘medically fit’.

5.6.8 The guideline is clear that a psychosocial assessment[91] should not be delayed until after medical treatment is complete and the EBTP reinforces that concept. The Treat as One Report found that just over a third of liaison psychiatry assessments had been delayed. The most common reason provided was that liaison services would not attend until the patient was medically fit. The report stresses that this can lead to ‘unacceptable or unknown risk’ being poorly managed[92].

5.6.9 A study[93] which asked, ‘Are hospital services for self-harm getting better?’ compared statistics from 2001/2 and 2010/11 and concluded that “disappointingly, given the introduction of clinical guidelines and policy emphasis, the variations in service provision were as wide as ten years previously with no apparent improvement in key aspects of clinical management.”

5.6.10 A study over a two year period found 58.9% of self-harm patients discharged directly from emergency departments, do not receive a psychiatric assessment[94] It concluded that non-assessed patients may be at greater risk of further self-harm and completed suicide than those who are assessed.

5.6.11 Another study[95], where 53% of individuals presenting to hospital after self-harm were offered psychosocial assessment, also found that individuals who self-injured were least likely to receive an assessment. The study concluded that the rise in self-injury as a method of self-harm, and the link between such methods and suicide, may have important implications for the management of self-harm in hospitals.

5.6.12 In the Treat as One report[96], reviewers considered that 23% of patients should have been referred to liaison mental health services but were not. Of those, 38% were not referred because the treating team did not feel it necessary.

5.6.13 A service users perspective:

“... when I arrived at A&E I was still unconscious, they did x-rays and did the necessary checks, they took good care of me ... they kept me in a cubicle at A&E, whilst waiting there, a mental health doctor came to me see, they were very nice, they spoke to me about how I was feeling and spoke about the different options available to me such as Psychological Therapy, Medication, Talking Therapy…”

5.6.14 There is evidence that the act of the psychosocial assessment itself following self-harm could reduce repetition by up to 40%[97], with patients valuing a positive therapeutic alliance that promotes hope and encouragement[98].

5.6.15 An international study found brief psychological interventions in the emergency department appeared to be effective at reducing suicide[99].

5.6.16 The investigation observed assessments being undertaken in the privacy of the mental health assessment room or at the bedside by the liaison mental health services staff. The professional used their own assessment methodology, which was then recorded in the service user’s mental health notes.

5.6.17 The investigation heard how the Welsh NHS has succeeded in standardising this assessment for use by trained mental health professionals using the WARRN tool (Wales Applied Risk Research Network) and the benefits of commonality between organisations.

5.7.1 The CQC is the independent regulator of health and adult social care services in England.

5.7.2 The Right Here Right Now report (2015) found the quality of care experienced by a person in mental health crisis varied depending on where they lived and when they sought help. It also found implications for safety, particularly in risks associated with self-harm.

5.7.3 The CQC is restricted under the statute to inspecting a single Trust. It has traditionally inspected acute Trusts or mental health Trusts. The most frequent model for liaison services involves a team from the mental health Trust working within an acute Trust. The ability of the regulator to assure the emergency department care pathway for an adult service user experiencing mental health crisis investigation is influenced by the limitations imposed by the law.

5.7.4 As part of the parity of esteem programme, the CQC has recognised this limitation and has a programme which provides for the inclusion of mental health inspectors as members of CQC teams that inspect emergency departments. The CQC has produced a brief guide for inspectors. It includes the following:

• Effective, rapid triage on presentation: staff trained and skilled in identifying mental illness and clear pathways for mental health care from the point of entry to ED?
• Effective risk assessment and risk management through ED: skilled staff available to assess and document risk?
• Clear observation policy: suitable staff available to provide 1:1 for patients at serious risk?
• Good links with the hospital liaison team. Consider skills and leadership of liaison team and evidence of joint working, location of liaison team office.
• Clinical record keeping. Note: liaison team may use different case records to ED - how is this managed?
• Dedicated rooms for people with mental illness that are suitable and safe.
• Waiting areas for patients are appropriately sited and comfortable.
• Service level agreement between acute trust and mental health trust which is reviewed regularly.
• Evidence of joint strategic working between ED and liaison service (policy review, environment etc).
• Board level oversight of mental health issues in Emergency Department.

5.7.5 The investigation considered that the inspection process highlighted some of the challenges when the liaison service is provided by a mental health Trust operating within the emergency department of an acute Trust. The assurances on leadership and governance arrangements were considered essential within a CQC inspection.

5.7.6 The CQC assurance process was noted as a positive incentive to comply with guidelines. The investigation also heard from emergency department staff where it resulted in generation of local processes to meet inspection criteria.

5.7.7 The CQC commented to the investigation that, while there were various models for service provision, their inspections should ensure that emergency department staff have full responsibility for all aspects of patient care until they leave the emergency department, and this was a vital safeguard.

Diane did not come to direct harm during treatment in the emergency department.

• HSIB considered that Diane’s final two presentations at the emergency department represented missed opportunities to intervene and put in place measures that may have helped to reduce her risk of suicide.
• The provision of liaison mental health services is variable across England and there is no consensus on commissioning models.
• Liaison mental health services have a positive influence on managing the safe care of patients in the emergency department. This influence was at its most effective when services had a permanent presence in the emergency department and were part of the multi-disciplinary team.
• The investigation found that when assessing physical health, emergency departments had established process for triage and Initial assessment but when considering mental state they lacked the same rigour – as a result, the outcome was inconsistent.
• Emergency department staff were exposed to the challenges of many long and complicated guidelines about the diagnosis and treatment of numerous conditions, while at the same time having to interpret a self-harm process that lacked detail and coherence. The lack of detail in the guidance contributed to the variability in initial assessment.
• There was the potential for misunderstanding in the self-harm guidance around interpretation and use of the Australian mental health triage tool.
• The national guidance on self-harm to emergency department staff for initial assessment lacked coherence between documents and did not consistently describe a detailed process.