STEPHANIE DANIELS · Prevention of Future Deaths report

On 26 March 2012 I commenced an investigation into the death of STEPHANIE DANIELS, aged 32. The investigation concluded at the end of an 19 day jury inquest on 29 November 2013.

The cause of death was found to be:
1a Asphyxia due to combination of ligature strangulation and obstruction of the airway

The conclusion of the inquest was that:
The deceased killed herself whilst the balance of her mind was disturbed by suffering from schizophrenia and auditory hallucinations and that her death was contributed to by neglect

The jury answered some additional specific questions as follows:

Upon the deceased’s admission to the Safire unit on 22nd March 2012, what level of observations should she have been subject to according to the policy in force at the time and her behaviour, history and presentation?
Continuous – within eye sight

Should she have remained upon that level of observations throughout her stay on the unit until the disclosure about the surrender of the ligature and the lighter on the morning of 24th March 2012?
Yes

Was the deceased admitted to Safire unit before or after the handover from the late to night shift took place on 22nd March 2012?
We were unable to determine this

Was a noose surrendered by the deceased on 23rd March 2012 but not recorded by the mental health staff in Amigos records?
No

Had the deceased actually been subject to any ‘discreet 1:1’ continuous within sight observations from the time she surrendered the ligature and the lighter on the morning of the 24th March 2012 to the start of the late shift at about 13.30hrs?
No

Was the nurse in charge of the early shift present and did they participate in the handover of the deceased’s care to the late shift at about 13.30hrs on 24th March 2013?
Present: Yes
Participated: No

Was the nurse in charge of the late shift present and did they participate in the handover of the deceased’s care to the late shift at about 13.30hrs on 24th March 2013?

Present: The nurse in charge arrived late to the handover.
Participated: No

What information was verbally handed over from the early shift to the late shift from about 13.30 hrs on 24th March 2012?
Chaotic
Came in via CRHT
On general observations, risk of self harm, scars on arms
Ligature handed in to staff

1. Stephanie Daniels (“the deceased”) was born on 16 April 1979 and was 32 years of age when she died on 24 March 2012. This was only some 4 months after I had conducted a lengthy and detailed Article 2 inquest into the self inflicted death of a detained patient who had absconded from a different unit but run by MHSC. I had written a comprehensive Rule 43 report letter (Re: Feisal King, deceased) and a number of similar themes and issues that emerged in that case were repeated in this one. Namely, communication, record keeping, handover, clinical leadership, the physical clerking in process and bed availability. In addition, I had identified significant failings in the Serious Untoward Incident Panel Investigation Report (“SUI report”) into that death.

2. The deceased had a troubled and unsettled childhood and was placed with foster parents and whilst there was abused. At the age of 13 she took her first overdose and was seen by a child Psychiatrist and was in contact with mental health services for some time. Not long after this she first reported hearing “voices” in her head. She had a baby when aged 16 and with whom by 2012 she enjoyed supervised contact about once a fortnight. Unfortunately, apart from her mental health problems she also abused illicit drugs, particularly heroin and cocaine. She suffered physical health problems primarily of a respiratory nature and had recurrent chest infections. She was also obese.

3. In 2001 she came into contact with the criminal justice system and was remanded into custody at HMP Styal. She repeatedly self harmed and was reported to have made one serious attempt to kill herself. She used various items as ligatures. Her first contact with MHSC was in September 2003 and she had contact with mental health services from then right up until her death. Stephanie was diagnosed with suffering from mental disorder when she was in her early twenties.

4. From 2004 she spent long periods in hospital either as a voluntary patient or compulsorily detained under the Mental Health Act 1983, as amended (“MHA”). She also had a significant history of numerous attempts at suicide and self harm. She had a diagnoses of schizophrenia, and an emotionally unstable personality disorder plus substance dependence. In addition she was being treated by the Community Drugs Team and was prescribed methadone. She had been treated with a wide range of medication including anti-psychotics, mood stabilizers and anti-depressants.

5. The pattern of her illness was that between hospital admissions she seemed to function with a reasonable quality of life for periods of time. However, these episodes were relatively short. The nature of her illness was one of a relapsing and remitting condition. Most of these relapses appeared to occur without any obvious precipitant. When unwell she presented with auditory hallucinations telling her that she was a bad person and that she should kill herself. These voices made her feel distressed and agitated. She was often acutely suicidal and regularly attempted self harm by overdosing or cutting herself. Whilst an inpatient she would often require one to one nursing to try to prevent self harm and her self harm attempts had included the use of a ligature. Sadly, she had a history of using forms of ligature whilst on the ward and also in her 24 hour a day supported community accommodation.

6. Her most recent admission to hospital was between 29 July and 5 December 2011 and she was formally detained under the MHA. Her poor physical health was exacerbated by recreational drug use and a chaotic life style. She had been assessed as suffering from drug resistant schizophrenia and therefore treated with Clozapine, which is only licensed for that use. Treatment with such a drug requires regular blood tests and monitoring of heart function. It was noted that she was developing abnormal heart rhythms and this medication had to be stopped. During her various admissions a number of different anti-psychotic drugs were tried but she was eventually discharged on a drug known as Quetiapine, plus a number of other medications including mood stabilisers, anti-depressants and substance misuse treatment.

7. Following her discharge from hospital she was regularly reviewed by her CPN and she was moved to different accommodation at a place called Clifton House on 9 January 2012 because of the increased level of supervision she required. This provided 24 hour residential care for adults with significant mental health problems. She was also seen by her Psychiatrist at outpatient clinic appointments. Initially she was positive about the future but by the end of January 2012 the CPN began to notice signs of deterioration in her mental state.

8. At the beginning of February 2012 she began to self harm again. On 20 February she disclosed that she was having increasing suicidal thoughts. She was taken to A&E and reviewed but seen the following day by her CPN when she reported low mood, intrusive thoughts and voices asking her to kill herself and her 15 year old daughter in a suicide pact by jumping from a bridge. Due to these increased risks she was referred to the CRHT. She was seen and reviewed once again by her Psychiatrist who increased her medication. This resulted in some improvement in her condition and she now felt able to resist the voices, especially with the help of the staff at Clifton House and she consequently did not want a hospital admission.

9. However, on 6 March 2012 a member of staff at Clifton House reported that she had attempted to strangle herself by using a ligature and she was taken to the Manchester Royal Infirmary for assessment, following which she was discharged back to Clifton House with levels of supervision increased. The deceased was anxious to avoid a hospital admission and thought she could cope with the help afforded by the extra supervision. Despite this on 12 March 2012 she rang the CRHT from a bridge saying that she intended to jump. They alerted the Police who were able to intervene and take her back to Clifton House where yet again they offered to provide additional support.

10. Her Psychiatrist reviewed her again on 14 March and she presented as agitated and depressed with escalating suicidal ideas. This resulted in a CRHT team meeting at which it was agreed that she now presented simply too high a risk to manage in the community and she required an urgent inpatient admission. Unfortunately, no inpatient bed was available to her, although she was given the highest priority for a bed when one became available. In the meantime another medication, namely Sodium Valproate, was introduced and her risks were reviewed and managed. Her medication also included a drug called MXL slow release. This is an opiate drug to replace methadone. Her AMIGOS electronic medical records (“AMIGOS”) had repeatedly noted that she was at risk of postural hyphaoxia and a number of her medications would have sedative side effects.

11. On the 22 March she phoned the CRHT from Clifton House telling them she was barricading herself in her room and attempting to hang herself. They in turn alerted the staff at Clifton House who ensured her safety and a bed was found at the Safire Short Stay Assessment unit (“Safire”) based at Park House Psychiatric Unit located in the grounds of North Manchester General Hospital but run by MHSC. There was a delay of 8 days in securing her admission which should have occurred as MHSC should have found her a bed, private or NHS, in any event. This assurance was given after the Feisal King inquest Rule 43 letter and that the CCG understood to be the position. She agreed to the admission and to receive treatment and so compulsory detention under the MHA was regarded as unnecessary. She arrived at about 21.00 hours and was commenced on what is known as general observations on that ward which was every 30 minutes, but the records suggest only from 21.30 hours at the earliest.

12. The relevant MHSC observation policy then in force defined general observations as being once every hour. A greater frequency than that (for example 1 in 30 or 1 in 15) was defined as intermittent observations and there was a requirement to complete a particular form of written record. The next level of observation was defined as continuous within sight at all times and the highest level was continuous within arm’s length at all times. Safire ward had been operating a general observation policy of 1 in 30 and the records kept of that were not in accordance with the MHSC policy then in force. The policy also defined the sort of behaviour or presentation that would justify each level of observation.

13. The nurse in charge of the shift maintained that the deceased had been admitted to the ward before he arrived and took the hand over and was therefore a patient already on the ward and had been assessed by his colleagues. The documentary evidence did not support this contention but the jury were unable to determine whether she was admitted before or after the handover.

14. I instructed two independent court appointed expert witnesses (a Professor of Mental Health Nursing and a Consultant Forensic Psychiatrist) to give evidence about the standard of her mental health nursing care and her psychiatric management. They both agreed that the deceased presented as a very difficult patient to deal with and that it would have been a challenge for any mental health team. The only criticism they made for pre-admission care was the delay in finding her a bed when she plainly was very ill and in need of admission.

15. Safire usually catered for up to 8 patients but could accommodate 10 and their usual complement of staff was 2 registered mental health nurses (“RMN”) and 2 support workers (“SW”). At weekends and evenings the RMN’s would have additional responsibilities of answering a crisis line and also trying to secure beds for urgent patients. These duties took up a lot of time. The shift pattern involved an early shift starting at 07.15, then a late shift starting at 13.30 and finally a night shift starting at 20.45. There would be a hand over from one shift to another and someone from the incoming shift would make written record of what they considered the main highlights. There was no requirement for any of the patients records to be read and considered, even for new or recent admissions.

16. Some time after 21.30 hours on 22 March 2012 the on call junior doctor attended the ward and prescribed the deceased her regular medications. There was no record made in AMIGOS of why he was called, by whom or when or if he had actually read any of the deceased’s records. There was no record of seeing the deceased. When giving evidence he accepted he should have done so and the absence of any record supported that fact that he did not. He did not clerk the patient in and did not notice that she had not been clerked in order to pass on the responsibility to a colleague. The MHSC protocol was that a patient must be clerked in (which involves taking a history, reviewing the records, assessing and prescribing appropriate medication, completing a physical assessment and a treatment/management risk plan) within 6 hours and in any event within 24 hours.

17. The nurse in charge of the night shift on 22 March 2012 recollected having had some contact with the deceased during a previous admission on a different ward. He had a few brief interactions with her but began to complete nursing records at about 00.50 on 23 March 2012. He told the court that he did read the last 24 hours’ records and considered at the time the level of observations (1 in 30) as appropriate. It seems that overnight there were no incidents or developments of note save that it was recorded by him at 01.20 hours that the deceased would not cooperate with something called a risk follow up assessment.

18. The following morning she was assessed and it seems that she continued to have thoughts of self harm/suicide. She received what is known as PRN (as and when required) medication and this seemed to help. This had actually been recorded in error in the medication chart as the injectable form of Haloperidol whereas it was the oral form. When giving evidence the Nurse who made the error admitted it. This had been prescribed by another junior Doctor who had attended the ward. There was no record made in AMIGOS of why he was called, by whom or when or if he had actually read any of the deceased’s records. Nor was there any record of him seeing the deceased. There was no review of her observation regime which remained the same. When giving evidence he said that his usual practice was to read the recent records, but he accepted that had he done so he should have noted that he deceased had not been clerked in and he would have reviewed her observations status and the management plan. He also accepted that had he properly checked the records he would have noted the risk of postural hypoxia and the recent cardiac history. This would be bound to be a consideration of two further medications which can have sedative and cardiac side effects.

19. The Consultant under whose name the deceased had been admitted never knew that the deceased had been actually admitted as a patient and therefore had not taken any steps to check or review the deceased’s condition. No steps were taken to inform him by any staff member on Safire ward.

20. The nurse in charge of the late shift on 23 March 2012 recorded a part of the handover information given to him that the deceased had surrendered a ligature that morning. The jury actually found that no ligature had been surrendered. However, it did not occur to him that in the light of this observations should have been reviewed by the previous shift as well by himself. When giving evidence he agreed that he should have done so. On the handover to the night shift he related the history of a noose being handed in and this was recorded, but once again it did not occur to the nurse in charge of the new shift to enquire about the deceased’s observations being reviewed nor do so herself. She made a similar concession when giving evidence.

21. On the morning of 24 March, a Support Worker on the unit spoke to the deceased, who disclosed that she had hidden a ligature in her bra and that she had a lighter and she planned to burn her room down, and she had a male voice in her head telling her to kill herself. She was persuaded to surrender both but said “I’ll just make another if you take it away”. This was brought to the attention of the nurse in charge of the shift straight away and she supervised a “risk follow up entry” being recorded in AMIGOS.

22. PRN medication was administered and the nurse in charge advised. Once again oral Haloperidol was given but not recorded at all. The nurse admitted the error when giving evidence. The ligature (which could also be used as a weapon) and the lighter which were surrendered by the deceased were then apparently left in the office. The deceased returned to the office several times to ask for a cigarette thereafter but no consideration was given to searching the deceased or her room. Her level of observations remained the same. The SW said that the nurse in charge said that they should commence with what was described as “discreet 1 to 1 continuous within sight observations”. No clinical record was made of this and no ongoing observations records were kept. This would not have been in accordance with the MHSC observation policy. The SW was not asked for a statement by the SUI investigation panel although he made a very significant clinical risk follow up entry.

23. The nurse in charge never mentioned this in any statement she made after the incident and nor when she was interviewed by the Chair of the SUI investigation. However, when giving evidence she accepted that she had not seen a copy of the statement the SW had made to the Police and the first time she knew what he was going to say was when he got into the witness box. She gave evidence immediately thereafter and agreed that the substance of what the SW said was correct but the jury did not find this happened as a fact.

24. There was a significant difference in recollection between the nurses in charge of the early and late shifts concerning the information handed over between the shifts on the 24 March. She maintained that fact the deceased was on “Discreet continuous observations” was handed over and that they would need to be reviewed. The late shift staff and in particular the nurse in charge denied this. The jury found that the “Discreet continuous observations” had not been handed over although they did find that the surrender of a ligature had been mentioned. This itself should have prompted a review and according to the MHSC observation policy at least continuous within sight observations should have been commenced.

25. It is recorded that the deceased was seen at 16.00 as part of the general observations check but one of the staff thinks she may have spoken to the deceased at about 16.15 in the garden. At any rate at about 16.30 she could not be found in the main areas of the unit when the next general observations were due and she was then located in a locked toilet and shower room with a ligature tied around her neck and apparently unconscious. The alarm was raised and assistance summoned.

26. Unfortunately, a number of issues then arose. There was a large degree of panic and confusion. A specific ligature cutting tool which was meant to be on the ward could not be found; the ligature could then not be cut until some scissors were found; an oxygen cylinder had a missing part and could not be used; staff believed that the suction machine could not be plugged in because it did not reach the nearest socket; what is known as the “crash team” (the emergency cardiac arrest team) were not contacted. Basic life support measures were initiated once the ligature had been cut and a 999 call was made. Records have established that this was actually made at 16.40 hours and 50 seconds and the call lasted for some 3 minutes and 40 seconds. A further call was made at 16.48 which lasted for 42 seconds. The NWAS (Northwest Ambulance Service) personnel arrived at 16.50 and took over care of the patient and carried on CPR. The NWAS defibrillator was used and this showed that the deceased had no shockable rhythm.

27. They left at 17.03 and she was conveyed to the A&E department in the main part of the hospital site a short distance away, arriving at 17.06 and handing over to triage at 17.10. Efforts continued to resuscitate the deceased but these proved unsuccessful and she was pronounced dead at 17.25 hours. I authorised a Forensic Post Mortem examination during which it was discovered that the deceased had been wearing a partial upper denture and this had been found lodged in her windpipe. I obtained evidence from experts in both ambulance service care and emergency medical treatment who said that the actions of both the NWAS and the treating emergency Doctors could not be criticised, nor would it have been apparent that she had an obstruction in the airway. Furthermore, whilst there were failings in the immediate emergency response when she was found these would not have been contributory to or causative of death.

28. Following the death the Police carried out an investigation and examined the deceased’s room. They did not discover any other ligatures nor any torn sheeting. It could not be established from where or when the ligature that the deceased actually used came from but it had the appearance of being from NHS hospital sheeting. They were provided with copies of the relevant medical and hand over records. The original hand over records could not be produced by MHSC at the hearing and had either been mislaid or lost. However, the court had good quality photocopies to work from.

29. No one could explain where the surrendered ligature had gone and there remained the possibility that the deceased could have recovered that ligature and used it or may have had another one, but no search of her room or her person was made.

30. MHSC initiated quickly after the incident a Serious Untoward Incident Panel Investigation Report (“SUI report”) which was chaired by the Trust’s Chief Nurse and had four other members. These included a senior Consultant Psychiatrist, his Specialist Registrar plus two other senior health professionals. They had access to all the records and obtained statements from a number of the staff involved and spoke to several of them. The SUI report was completed by the end of June 2012. The Chair of the panel gave evidence and accepted that there were significant failures and omissions in the report produced. For example, they had not obtained a statement from the Nurse in charge of the night shift on 22 March; nor a statement from the Nurse in charge of the late shift on 23 March who records the handing in that morning of a ligature; nor a statement from the SW on duty on the early shift who the deceased surrendered the ligature and lighter to and made a very significant risk follow up entry; nor had they found out why there was delay in the deceased getting a bed. They failed to note or record the medication recording errors, despite the fact that if an injection of Haloperidol was administered as the records suggest, then the Trust’s Rapid Tranquilisation Policy should have been followed. That required the deceased to have close observations for up to 2 hours post administration but there was no record of that being done.

31. The Police, as part of my investigation did obtain a statement from the SW on the early shift. It revealed that his recollection was that some form of “discreet” 1 to 1 continuous within eyesight observations was initiated by the nurse in charge, on the basis that she said that the deceased had a history of reacting adversely to increased observations. When challenged to substantiate this in giving evidence she could only say that is what she understood from her other nursing colleague but he did not give the same account when he gave evidence.

32. The SUI report concluded that from admission a greater level of observations than 1 in 30 but less than continuous within sight on night shift was the appropriate level up to the morning of 24 March. Both expert witnesses are firmly of the view that continuous within sight should have been the observation regime from the time of admission. Importantly, that the deceased’s death was both predictable and preventable. The SUI panel did agree that the deceased had not been clerked in at all despite some six shifts of staff being involved. Nor did she have any psychiatric input. They concluded that from the morning of 24 March continuous within arms length observations would have been appropriate although both expert witnesses said the risks could have been managed with the lesser within eyesight status.

33. Finally, during the course of the inquest it became apparent that another patient who was seriously unwell had her observations increased to the intermittent 1 in 15 level from the late afternoon of the 22 March level and the records suggested that this was done whilst of Safire ward. She was then moved to another ward and her observation levels remained the same for several days but there were gaps in the observation records. Her observation levels were increased to continuous within sight but again there were gaps in the records where nothing was recorded. This raises the possibility that important observations were not being done or certainly not recorded. Whether this is an isolated issue or represents something more systemic would be for MHSC to investigate.

1. Internal NHS SUI Investigation v Independent Investigation

I am concerned that a death such as this of either a detained or voluntary patient (where Article 2 is arguably engaged) requires a prompt, thorough and robust investigation to be completed as soon as possible. Deficiencies in systems, protocols, policies, record keeping and individual actions need to be identified quickly and remedial action taken. This cannot wait for an inquest which may not take place for many months. This was not the first case of a poor or incomplete SUI investigation. The court was aware of and invited submissions about the recent case of R (Antoniou) v Central and North West London NHS Foundation Trust and others [2013] EWHC 3055 (Admin).

In this case there were significant errors and omissions in the SUI investigation. Important witnesses were not interviewed. The delay in finding the deceased a bed was not a central issue and no specific findings were made about it. The medication recording errors had not been noted and had the deceased been injected with PRN Haloperidol for severe agitation, then she should have been subject to physical observations for a continuous period of time immediately afterwards, as well as other steps in compliance with the Trust’s Rapid Tranquilisation Policy. It is accepted that SUI investigations are important and hope to learn lessons quickly to be implemented. Whilst the current law indicated that it is not a requirement for there to be an independent investigation at that stage, it is a matter of concern that very significant failures in the investigative process have occurred. The Trust investigation did not reveal at all the allegation of the commencement of discreet continuous observations on the morning of 24th March.

This is not the first time that the Trust SUI investigations have been found to be flawed and I have experience of other Trusts’ investigations also being significantly flawed. In conclusion in this sort of case I am concerned that without appropriately speedy and thorough independent investigation commissioned by the NHS Trust involved, flawed SUI investigation reports may continue to be produced. This is a policy decision for the NHS but I strongly urge consideration of this.

2. Handover
All the evidence showed that the handover of information between nursing and clinical colleagues was a vital piece in the jigsaw of care. MHSC have introduced a new policy but I have a concern that simple issues may be overlooked. For new patients being admitted or transferred, there is no requirement for the nurse in charge to review their recent records. MHSC have produced new or updated policies/protocols but experience has shown that what may be delivered on paper is not being done in practice. Consequently, I am concerned that without appropriate audit and clinical/nursing leadership this may prove to be ineffective.

3. Bed Availability
MHSC say that following the death of the deceased, a new policy has been introduced so that there is no waiting time at all for the allocation of a bed in the case of a patient who is deemed clinically to require one. A bed will be found somewhere which will be appropriate to their needs. As I understood the evidence from the CCG in the case this should have occurred in any event. However, other NHS Trusts nationwide who do not have such a policy, may have patients whose delayed admission means that they are not having the appropriate nursing and clinical input, as well as medication review. In turn this means their condition may continue to deteriorate and when effective care does start, the patient may well be more ill than they should be. I am concerned that the importance of this is recognised not only by MHSC but nationally for all other NHS mental health trusts.

4. Clerking In
The failure to properly clerk in the patient is a matter of serious concern, especially as many such patients will have physical health problems. MHSC had clear policies requiring the clerking in of a patient, but these were simply not adhered to. It is very common for patients with mental health problems to have associated physical conditions which require appropriate monitoring and treatment. It seems that despite the existence of appropriate policies, in practice these were not being complied with. Whether or not any new or different policy or auditing of compliance is the way to achieve uniformity is a matter for MHSC. I repeat what I have said earlier. MHSC have produced new or updated policies/protocols but experience has shown that what may be delivered on paper is not being done in practice. Consequently, I am concerned that without appropriate audit and clinical/nursing leadership this may prove to be ineffective.

5. Supervision of Junior Medical Staff
I am concerned about the lack of appropriate clinical supervision and guidance for junior medical staff. Two junior doctors were asked to attend the ward and made no appropriate clinical records of the reason for their attendance, reviewing the records, seeing the patient and explaining any clinical decision to prescribe medication. It is appreciated that they are busy with a number of duties but it is a matter of concern that they did not undertake basic clinical recording duties for a patient who clearly should have been seen. They did not notice that the patient had not been clerked in. Medication was being prescribed without adequate consideration of the relevant clinical history. They did not notice the named Consultant in charge of the patient was unaware of the admission. Appropriate clinical supervision would be expected to ensure an appropriate standard of performance. I understand that supervision may be delegated by the North West Deanery to the relevant NHS Trust but there has to be some basic accepted levels of interaction, communication and supervision between the junior Doctors and their Consultants to ensure an appropriate standard and continuity of care. This may be a joint responsibility between the Deanery and the NHS trust involved.

6. Prescribing of Medication by Junior Medical Staff
I am concerned by the circumstances in this case where medication came to be prescribed. There is an overlap of my concerns about supervision and my observations at paragraph 5 above should be regarded as repeated here. Both junior doctors had no recollection of attending the ward, speaking to the staff or seeing the patient. They simply prescribed the medication. They had no recollection of reviewing the deceased’s records and understood that was essential when considering prescribing any medication, and in particular PRN rapid tranquillisation. In this case, the patient was already taking a number of drugs which had sedative effects. Two further medications were introduced that have similar properties and that also could potentially affect heart function.

7. Mechanism by which the Consultant in charge of the patient would learn of the patient’s admission
It is of concern that there was apparently no simple method of ensuring that the Consultant in whose name the patient was admitted became aware of the admission and could therefore ensure appropriate clinical leadership and review was undertaken. It would seem that there could be a number of simple solutions for this problem.

8. Performing and recording observations on other patients
I was concerned about the discovery of incomplete written observations for another patient where there are significant gaps in the records and may illustrate a systemic problem because the patient was transferred to a different ward. This was only discovered during the course of the inquest and was brought to the attention of MHSC so that they could carry out their own investigations.