Savage v South Essex Partnership NHS Foundation Trust

Carol Savage met her death on 5 July 2004 by throwing herself in front of a train at Wickford station in Essex. She was 50 years of age and left a husband, two children and a close-knit and caring extended family. At the time of her death she was a patient at Runwell Hospital which was managed and staffed by employees of the defendant. She was detained under Section 3 of the Mental Health Act 1983, for the purposes of treatment and for her own safety. She had suffered from mental illness on an intermittent or fluctuating basis for many years.

This claim is brought by her adult daughter under Section 7(1) of the Human Rights Act 1998, based on Article 2 of the European Convention on Human Rights. The claimant says she is a victim for the purposes of Section 7(7) of the Act and seeks a finding that the defendant Trust was in breach of Article 2, together with just satisfaction.

A preliminary issue in this claim has been considered by the House of Lords and is now reported at [2009] 1 AC 681. Their Lordships considered the nature and scope of the defendant’s obligation under Article 2 in the circumstances of this case. They ruled that where there is a “real and immediate risk” of a detained mental patient such as Mrs Savage committing suicide Article 2 imposes an operational obligation on the hospital authority to do all that can be reasonable expected of it to prevent it. That obligation arises only if members of staff knew or ought to have known that the particular patient presented such a risk. The threshold that the Claimant has to surmount is a high one, higher than, for example, the familiar Bolam test in clinical negligence cases in domestic law. I will have to consider the law more fully later in this judgment. For the moment I must now set out in some detail the facts leading up to her death as I find them to be.

Carol Savage’s history of mental illness went back to the birth of her son in the late 1970’s when she suffered from post-natal depression. There is not a complete record available of all her subsequent episodes of illness. But more recently between October 2001 and April 2002, for something like five and a half months, she had been an in patient at the same Runwell Hospital detained under Section 3 of the Mental Health Act. In the course of this stay she was recorded as being at risk of suicide or self harm, being said to feel she was being persuaded by others to kill herself. On 23 November 2001 she absconded and was found on the A130 main road some distance away, walking between the cars. She told staff when she was taken back to the hospital that she wanted to die, though she later said she was alright. Her mental history and symptoms attracted a diagnosis of paranoid schizophrenia.

She then lived in the community with her husband, holding down a part time job, until her mental state deteriorated to such a state that on 16 March 2004 her husband took her back to Runwell Hospital initially on an informal basis. Her illness took the form of delusions, that she heard voices, was agitated and concerned, and anxious in particular that harm might come to the claimant her daughter (who had just started a new job in Norfolk). Mr Savage and other family witnesses called these her “horrible thoughts”.

When she reached Runwell Hospital on 16 March with her husband Mrs Savage was very ill. She was said to have been pacing up and down for ten days, unable to sleep, and she would not answer any questions asked of her. She was admitted and assessed as requiring level three observations. She was threatening to abscond.

The following day 17 March she was still on close observations. She was described as pacing and not answering questions, just saying (obviously in respect of her daughter whose safety was and remained throughout her time in hospital an obsession with her) “I know that she is safe” and “I want her safe”. The doctor assessed her as a risk of absconsion (the term used in the NHS for absconding) and she was detained initially under section 2 of the Mental Health Act 1983. The following day she was no better and was detained under section 3.

On 19 March she was still very restless and agitated, refusing oral medication and trying to get out of the ward. She smashed a glass window and had to be restrained by two members of staff. She was put on 15 minute observations and consideration was given to her transfer to a Psychiatric Intensive Care Unit at the earliest opportunity, but no bed was available. The decision was taken to transfer her on 20 March to the Cygnet Clinic in Beckton, East London, a private establishment which had a secure facility. She was plainly still very ill at this stage.

A risk assessment was carried out on that date which assessed the risk of suicide as low and the risk of absconding as high. She stayed at the Cygnet Clinic until 13 April. While there she was risk assessed by a Consultant Psychiatrist on 1 April who said it was difficult to assess her mental state, describing her as “not hostile nor is she suicidal. No signs at all of depression or mania.” He assessed her as a medium risk of absconsion, a low risk of self harm and suicide but said it was too early for her to leave the unit.

She was re-assessed on 11 April on a score sheet which included a zero score for “non accidental self injury”. She was transferred to the Basildon Hospital Belhus unit (a secure unit) on 13 April. Her discharge summary from Beckton diagnosed her as suffering from paranoid schizophrenia. It set out a full account of her presentation at the Cygnet unit and included the information that she had acted bizarrely, seemingly responding to hallucinations which were telling her to jump out of the window and at one stage she had had to be restrained while attempting to climb through a window. She had done the same again on the 22 March. This discharge report and, it seems reasonable to assume, the other assessments made at the Cygnet unit must in due course have reached the staff at Basildon hospital run by the defendant when she was transferred to that unit.

On arrival at Basildon she was immediately assessed by a nurse who stated that the primary identified risks were absconding and sudden outbursts of aggression, hence the move to Cygnet where it was reported that she had become significantly more settled. She received a mental health assessment later that day. The resultant risk assessment indicated that there was a low apparent risk of suicide and a significant risk of severe self neglect. On the second page there are boxes ticked which state that there was a risk history of absconding as well as current warning signs. There is no mention of a risk of absconding on the face of this document until it was reviewed on 18 May by which time the absconsion risk was stated to be significant.

The next day 14 April she was seen by the SHO and said she had been admitted because people wanted to kill her in the community because of her daughter who was in Norfolk. When asked if she felt like self harming she said it would be better if she took tablets and disappeared. She felt worthless and that her husband did not love her. Mr Savage was an assiduous visitor to his wife on a daily basis and could not have done more to care for her, and therefore this remark must have been a further manifestation of her delusional state. Over the days that followed she remained isolated, reported that everyone was “after her” but by 22 April was deemed suitable to be transferred to Grangewater ward, an open ward, and was adjudged fit to go home that weekend on leave with her husband. Up to this point there is not nor could there be any serious criticism of her care by the Trust.

Although Mr Savage expected to take her back to Grangewater ward on the Sunday evening he was told he could not because there were no beds available and she therefore stayed on at home on the Monday Tuesday and Wednesday, not returning until the Thursday 29 April. He said she seemed to go downhill as a result of this extended and enforced leave period. When he returned her to hospital it was to Runwell Hospital in Chalkwell ward, an open ward.

She was seen there by a Senior House Officer Dr Abrar who thought that she had settled a little bit compared with before (meaning mid March on her first admission) but she was still complaining of people being after her to kill her, saying she was “all confused”. Her mood was frightened and confused, and she was quoted as saying “I would harm myself if I could” and “would take some tablets”. A diagnosis of paranoid schizophrenia was again suggested and the plan was to admit her with 30 minute observations and to be seen by the Consultant Dr Ague prior to any leave.

There is no record of the instruction for level two 30 minute observations ever having been amended nor is there any record sheet of those observations, which should have been kept in accordance with the Trust’s policy, and which had been kept during her 2001/2 stay. The then current policy guidance was that level 2 observations should be at intervals of no more than 15 minutes. I will have to consider these policies below.

On 30 April Dr Ague, the Consultant Psychiatrist and responsible medical officer or RMO for Mrs Savage, reviewed her position at a meeting which Mr Savage attended. Mr Savage had earlier reported that while on leave Mrs Savage remained paranoid, expressing concerns that someone would harm her, but was suggesting day leave. At the review Dr Ague found her to be very anxious and agitated and agreed section 17 leave daily for a maximum of six hours in the care of Mr Savage or Mrs Savage’s mother, the occasions of leave to be agreed by nursing staff in consultation with the medical team. Thereafter she enjoyed leave daily from 1 May, being returned to the ward by her husband at night. For ten days or so this state of affairs continued with Mrs Savage continuing to express fears that someone was going to harm her.

At the ward round with Dr Ague on 11 May Mrs Savage said that she felt better but appeared very low in mood as she kept standing up and was walking about in the ward. Dr Ague formed the plan that she should have no leave that week and could only have it the following week if her mood had lifted and if she had seen Dr Abrar prior to leave. At Mr Savage’s suggestion she was prescribed Mirtazapine at a low dose, the significance of which will have to be considered below. Thereafter her mood continued to be low and she remained un-communicative and afraid someone would harm her and harm her daughter.

On 16 May she was observed to be leaving the ward with her suitcase and agreed to come back when staff stopped her. She said she thought she was not wanted on the ward and was hearing voices saying violent things threatening to beat her up.

On 18 May she was tearful and expressing ideas that her family might come to harm and that she would not be safe on the ward. Later that day she was escorted back to the ward after having again packed her case and attempted to leave. She said that everybody was telling her to leave because they did not want her on the ward and attempts were made to reassure her. At an attempted one to one session that night she said she was not ready talk to her named nurse, Nurse Puleng. The same nurse on that day made a handwritten amendment to the earlier (Basildon) risk assessment form dated 13 April to reflect this incident, writing on it the date and “Abscontion risk (2)”, which applying the scale on that form indicated it was a significant issue and a risk plan was to be drawn up. None ever was. She did this no doubt because she would have found no documented risk assessment on her file generated on or after her arrival on Chalkwell ward.

On 27 May the care co-ordinator Michael Manze (a social worker) purported to complete a risk assessment. It appears to assess her as displaying a low risk of suicide and self harm and a risk history of absconding saying “she will try and leave the ward due to the persecutory nature of the ideas she is having”. Mr Bishop with realistic candour does not now suggest this form was of any value at all as a means of assessing and alerting others to the type and level of risk Mrs Savage presented. It was incomplete, inaccurate and worse than useless in my view.

On 28 May she attended the music club for an hour in the morning but still expressed feelings that people were trying to hurt her and her family and was restless and uncomfortable. Later that day after her husband’s visit she was observed leaving the ward and walking up towards the perimeter of the hospital grounds, making no attempt to disguise the fact that she was leaving the hospital. When brought back by staff she said she did not know where she was going and was reluctant to come back but also expressed fears concerning her family.

The following day she started back on day leave and this pattern continued over the days that followed, through which she remained worried and agitated about her own safety and that of her daughter and family. She participated in some activities of a group nature and in general appeared to be enjoying the home leave with her husband. But on any view she was still plainly seriously ill. Mr Savage told Dr Ague on 15 June that she was able to do tasks and jobs at home which helped her to be occupied and she appeared to enjoy them. When not kept busy at home she sat as she did on the ward and was bothered by “the thoughts”.

The next day she attempted to leave the ward at 2130 and got outside the door but was intercepted by staff and encouraged to return. She said that she wanted to go home. This was the fourth time she had tried to abscond from Chalkwell ward and her sixth attempt overall.

By now her RMO Dr Ague, who had examined Mrs Savage on 30 April, 6, 11 and 25 May and 1 and 15 June had gone on annual leave from which he had not returned by the time of her death. Mrs Savage was effectively left under the supervision of Dr Abrar and continued to enjoy leave at home either with her mother or husband on a daily basis. She was seen on 22 June by Dr Abrar on a ward round. Dr Abrar was a very junior trainee in psychiatric terms, though in my judgment an intelligent and conscientious doctor, having started at this post in February of that year. Mrs Savage was still showing clear symptoms of paranoid schizophrenia. She was still convinced that something horrible was going to happen and was paranoid about people following her and taking her thoughts out of her head. At the same time she was asking Dr Abrar to give her home leave to visit her daughter over the weekend as it was her birthday. Dr Abrar withheld all leave until the return of Dr Ague telling her she was not in a condition to visit. She said in evidence that she was erring on the side of caution because of her junior status, but at all events home leave was stopped altogether until 29 June when her condition remained much as before. During this period a second opinion was sought as to possible changes to her anti-psychotic medication regime.

On 29 June Dr Abrar saw her again and noted her condition continued to be ambivalent, she was rubbing her hands and pacing up and down. She asked for weekend leave but Dr Abrar refused this due to her state of mind. She did however grant home leave for 30 June (her wedding anniversary) and that was reported to have “gone well” by both Mr and Mrs Savage, though the support worker described her on return as “unsmiling and making rocking movements.”

The next day Mr Savage requested weekend leave again but Dr Abrar felt able only to grant day leave which was taken on 3 July. When he returned her to the ward he reported her state as “not too bad”; it was at about this time that she showed her husband a piece of A4 paper on both sides of which she had written out repeatedly “no harm will come to my family”.

On 4 July her husband visited her at her hospital and she was reported to have managed a smile and had not voiced any concerns. Though her mood was described as settled she had been observed pacing the corridor and asking staff about the safety of her family. The staff re-assured her. The nursing assessment that day was that she had taken her meals and “poses no management problems at the moment”. That evening it was said that she remained settled and slept well.

The following day was her last day at the hospital. At lunchtime she was described as seeking re-assurance about her family’s safety. It was a fine sunny day and she spent long periods outside the ward on the veranda with other patients. The support worker Gill Prigg saw her sitting with another patient called Carol whom she had befriended. When Ms Prigg went home at the end of her shift Mrs Savage said she would see her tomorrow. Senior Support Worker Marion Munro saw her at about 4 o’clock in the afternoon when Mrs Munro was talking to another patient who was quite hysterical and was trying to calm her down, suggesting that she went for a walk. Mrs Savage who was also on the veranda outside offered to take the patient for a walk but the patient declined and Mrs Savage left the veranda area. Mrs Munroe did not know where she went to but had quite a good view of the ground of the hospital from the veranda and did not see her leave via the garden exit. She was not sure of the time but believed it was about 1615 when she last saw Mrs Savage. She described her mood as more chatty in a way that was out of character and nice. She could not remember telling Mrs Savage’s mother on a later occasion that Carol was crying a lot that day. She recalled Mrs Savage as being tearful and upset on the 4^th but not the 5^th.

At about 5 o’clock on the afternoon of 5 July Mrs Savage was killed by a train at Wickford railway station. It would have taken her about 45 -50 minutes to cover the two and a bit miles to the station from the hospital.

A large part of the claimant’s case consists of allegations that the hospital staff failed to follow the Trust’s own stated policy particularly relating to risk assessment for patients detained under the Mental Health Act and the observations to be kept on such patients. I should first summarise the relevant features of these policies.

First as to risk assessment, the ward manager Mark Gillett accepted that a good risk assessment was absolutely fundamental to the management of a patient in Mrs Savage’s position. The defendant’s clinical policy CLP28 had the stated aim of outlining their approach to the minimisation and prevention of adverse events resulting from risks associated with clinical care. Included in the factors specific to that were the risk of self harm and of absconding. Contributory factors highlighted in connection with suicide were or included a failure to monitor patients, supervise staff and obtain past records as well as poor communication between staff and a previous history of absconding. To meet these risks there was a requirement for a systematic assessment of clinical risk by qualified staff and Section 3 of the policy set out what that involved. It included an assessment of the patient’s history, previous suicidal behaviour, poor compliance to treatment, mental state, evidence of delusion/hallucinations of a persecutory nature and evidence of suicidal ideation or plans.

This assessment was required to be documented and detailed in an individual care plan which had to be identified and documented both on the medical and nursing records and should answer questions which included how serious the risk was, whether it was specific or general, how immediate it was and how volatile it was and what specific treatment and management plan could best reduce it.

Staffs were also told in the policy that the risk assessment process was an ongoing matter and reviews should take place on a daily basis and be evaluated weekly at the patient’s ward round. Changes in the patient’s behaviour or condition should trigger a re-assessment as should further information from relatives notes or other professionals.

As I have already found there was no documented risk assessment for Mrs Savage worthy of the name from the time she came to Chalkwell ward, nor any risk management or care plan.

Another key feature of the management of patients in Mrs Savage’s condition is the formal observation policy CLP8 and the guidance notes relating to it. Observation is a prescribed intervention when a patient poses a risk to herself or others and is described in the policy as an essential part of risk management.

In accordance with national standards the defendant’s policy identified four levels of observation. The lowest was level one, called general observation. Level two was intermittent observation with the frequency specified subject to a maximum of 15 minute intervals. Level three was continuous observation “within eyesight” and level four within arms length. The policy noted that risk of absconding on its own was not a reason for a prescription of increased observation. There were prescribed procedural guidelines CLPG8 for engagement and formal observation. This required a member of staff to act in the role of allocated nurse for each shift for the particular patient to be accountable and responsible for ensuring positive engagement with the patient “in a positive therapeutic way”.

General observation is described as the minimum acceptable level for all in patients and requires the whereabouts of the patient to be known to staff, but not that the patient be kept within sight. Level two requires the patient’s location to be periodically checked. Levels three and four were never engaged in Mrs Savage’s case. All observations other than level one should be recorded on a log, and there are examples of these from the 2001-2 treatment period.

When decisions regarding observations are made a risk assessment is required which should include an interview with the patient and carers and a careful study of the patient’s history by means of ratified risk assessment tools and the patient’s thoughts feelings and wishes about suicide should be considered. The patient’s notes are described as a vital source of information.

These decisions must be kept under review. They must also be documented on a prescribed form called the “observation care plan” and where levels two and above are engaged a specified form of observation record sheet.

If a patient’s observation level is reduced from any of levels two three and four this should be discussed with the RMO or his designated deputy and the process should be documented in the management plan and communicated to all members of the clinical team. The decision must be recorded and signed in the medical and nursing notes.

It is easy to see these documents as tiresome management-driven bureaucracy, as I fear Dr Ague did, as to an extent did I on first introduction to them, but as the case progressed I saw more clearly what an important role they could play in circumstances such as these. They not only focus attention on these two critically important areas in the care of someone in Mrs Savage’s condition but ensure that relevant assessments and instructions are communicated and shared between all members of the multi disciplinary team.

As to psychiatric issues, Dr Walsh called by the claimant believed the diagnosis of paranoid schizophrenia was incorrect and misleading. She believes the right diagnosis was Schizo-Affective Disorder under which the patient displays simultaneously symptoms of depressed mood and delusionary symptoms. She believes the clinical team did not fully appreciate her depressive disorder. In her written report she was more forgiving of this misdiagnosis, saying it was very difficult, than she proved in the witness box.

She conducted an extremely thorough review of all medical records including those from the GP relating to earlier hospital admissions. She noted that in September 1996 when she was suffering from episodes of paranoid psychosis a doctor said that “there also seems to have been quite a marked affective mood component at times”, and in 1998 a doctor from the Runwell Hospital wrote to the GP in similar terms concluding “her illness appears to be a Schizo-Affective type of illness triggered by a particular stress”.

When she was discharged from Runwell Hospital in April 2002 her medication included Mirtazapine, which I accept is primarily an anti depressant. Dr Ague, who strenuously denied that she had ever merited a diagnosis of depressive disorder at any stage, had the care of Mrs Savage as an outpatient when she returned home following that discharge and he appears to have continued the prescription of Mirtazapine, as Dr Abrar was to note on her admission to Runwell in March 2004. Despite the note that it was plainly intended that she should continue on this drug it was not given to her until mid May when either Mrs Savage or her husband asked for it and it was initiated at a low dose, increased to 45mg by the time of her death.

The significance of this says Dr Walsh, if she is right, is that a significant part of her symptomatology was never treated in a proper way. Mirtazapine should have been continued throughout and if it did not serve to reduce the disorder a different anti depressant should have been tried for about 4-6 weeks. If that had not worked she could then have been called treatment resistant and alternative strategies might have included Lithium and, if no success was achieved by that, ECT.

As to the treatment Mrs Savage received Dr Walsh considers that the history taking and the recording of treatment were both inadequate. The psychiatric history appears to have been unknown to those that were treating her and in particular Drs Ague and Abrar were unaware of what happened in the 2001 admission in terms of the incident on the A130 road. Although she accepts that in such a ward, as she put it, “you will have a lot of patients reporting suicidal ideation” and she did not think that such remarks triggered a new assessment of risk every time a patient makes such a comment she would expect to see a consultant assessing the risk on the occasions when Mrs Savage said it if he believed she had not ever said it before. Certainly following episodes of absconding she would expect a doctor to investigate why the patient had done this and consider the level of observations.

As to the level of observations themselves there is simply no documentation on this topic at all beyond Dr Abrar’s assessment on Mrs Savage’s return to Chalkwell ward on 29 April that she be on 30 minute observations i.e. level two. There is no note or log in the prescribed form kept of those observations and for the rest of her stay it appears to have been assumed by all concerned that she was at the lowest level namely general observations. No document shows this adjustment being made. Mr Gillett the ward manager thought the fact that she had been granted home leave indicated that there must have been an antecedent decision to reduce her status to general observations and Dr Ague for his part said that “tacitly”, as he put it, he reduced her level to general observations when he allowed weekend leave on 8 June. He “presumed” that she had remained on level two up to that point. That seems to me highly unlikely, indeed I find that she was not at any stage treated as anything other than a level one patient through May and June till her death. The standard of nursing records at this hospital was generally as I find much superior to the standard of the medical notes and I cannot believe that not a single day’s observation records, if she had been on Level 2 between the end of April and beginning of June would have survived. I find that the position was that de facto the staff assumed she was on general observations and that the inexperienced Dr Abrar’s intentionally and understandably cautious recommendation of 30 minute observations was simply ignored.

Dr Walsh’s view is that after each episode of absconding between 16 May and 16 June (there were four) the level of observations should have been reviewed by a doctor who would have then been in a position to assess the seriousness of these episodes; that was the minimum she would expect. I see force in this criticism. Mrs Savage was not easy to question but if she was not forthcoming a higher level of observation would clearly have been called for. Her motivation in absconding could not simply be assumed, from the fact that she took a case with her, to be no more than a desire to go home; on two occasions she was also saying that nobody in the ward wanted her, and there should have been some inquiry into why she was trying to abscond. If that drew a blank a precautionary approach was necessary.

Dr Walsh is also critical of the amount of leave Mrs Savage was granted while acknowledging the therapeutic value of home leave for a patient in her condition. There is no doubt that the five day episode of leave between the 23 and 29 April seems to have upset Mrs Savage, as she was expecting to go home after two days. Thereafter, in my judgment, the doctors were under understandable pressure from the family to grant leave and careful decisions were taken as to the quantity and timing of that leave. I was not greatly impressed by Dr Walsh’s criticisms of this aspect of Mrs Savage’s care. The leave itself I am sure caused her no harm and probably did her good. Her husband watched over her like a hawk and incorporated useful and pleasurable activities for her such as she could perform. As he put it she needed relief from the monotony of life in an acute psychiatric ward. I do not consider that the granting of leave even if it could have been called generous played a part in exacerbating her mental state and I doubt that as Dr Walsh contended, it made staff more comfortable about giving her freedom when she was in the ward. The problem ran deeper than that.

Though at one stage she appeared to advocate a position whereby Mrs Savage should have been returned to a locked ward she did not finally hold to that view. In my view that would have been very destructive of the welfare of Mrs Savage given the type of symptoms she had particularly her persecutory delusions. Her view is that had the medical staff properly investigated Mrs Savage’s position and the risk she posed both of absconding and of self harm there should have been more frequent observations at either 15 or at most 30 minutes. This should have been the regime, as I understand her evidence, throughout her detention, certainly from 29 April onwards.

Dr Turner was the defendant’s expert on psychiatric issues. I thought him a measured and careful witness. His view was that she was suffering from paranoid schizophrenia with associated depression as occurs he believes with up to 25% of those who suffer from that illness. If there were unaddressed symptoms of depression in his view they would have increased the risk of suicide only if they were very prominent symptoms such as self blame, morbid guilt, a sense of loss and a sense that the patient is “a bad person”. Those symptoms do not seem to be prominent in Mrs Savage’s case. I do not consider that, whatever label was attached to them, the depressive symptoms which were undoubtedly present significantly increased the risks present due to her schizophrenia, which I am satisfied dominated her mental problems.

Dr Turner accepts there was a real and immediate risk of absconding in this case and that once she was placed in an open ward it was more important to consider it and its likely consequences. This is a key issue and an important and I think proper assessment by him. He agreed it was fundamental to assess what the consequences of absconding might be. On the facts of this case, despite the fact that Mrs Savage presented symtoms of impulsivity, paranoia, delusional fears and was worried that she herself might be the cause of the threat to her family and daughter, he did not accept that it was simplistic to assume that her attempts to leave the ward in May and June had the benign purpose of going home. He thought that was a logical assumption. The fact that she had packed her bag and on one occasion had sat outside the ward he thought could be seen as evidence of her wanting to go home and merely an aberration of her current mental state. Though he said in evidence that potentially a better interrogation of the patient’s motivation might have revealed more, the process of such examination can be distressing (and all agree that Mrs Savage did not find it easy to elaborate on her thinking) and he said it was a common experience that the observation of behaviour was sometimes more useful that questioning. I prefer Dr Walsh’s views and those of Professor Rogers on this point.

As to the risk of suicide it was Dr Turner’s view that this was not real and immediate or, as was put to him, obvious. Though he accepted that there is an increased risk for all patients with these symptoms and 10% or so are likely to commit suicide statements of suicidal ideation he said are very common, there were no previous incidents of deliberate self harm, cutting, overdoses or the like in this long illness. He believed that her mood state was properly monitored and in the days and weeks leading to her death was stable.

Professor Rogers was the nursing expert called by the claimant and was professor of nursing at the University of Glamorgan. He did not have wide clinical experience of nursing in a ward such as that I am considering in this case, his last nursing post ending in 1992 when he had worked in a private hospital specialising in care for patients with acquired brain injury, before which he had experience in medium secure units. In some areas of his over long report he made serious criticisms of the nursing staff’s approach to recording the administration of drugs, which Mr Bishop exposed as unfounded and for which he apologised. As to risk assessments he could not see that these had been undertaken judged by the documents in the case. He described Mr Manze’s form completed on 27 May as incorrect incomplete and woefully inadequate. There should have been a risk assessment based on a through examination of her history and discussion with her family and agreed with all leading to a risk management plan specifying the risk, the conditions in which they would occur, the warning signs and the actions to be taken. I agree with those criticisms

He thought the risk of absconding became a high risk once Mrs Savage had attempted to abscond and thought those absconsions should have prompted the nurse in charge and the doctors to discuss and review her case in the light of what she had done.

As to the risk of suicide he assessed this as medium to high and considers that the risk was never properly assessed. He relied on the previous suicide attempt in November 2001, the reports of self injury when she was in the PICU (I find the evidence for this tenuous), her two mentions of suicide and her absconding behaviour. While he accepted that it was not uncommon to hear thoughts of suicide being expressed by patients with her condition questions had to be asked as to why she said it and whether she meant it. If she is a patient who does not repeatedly say such things and is reserved then it is extremely concerning. If she will not provide information about this she has to be managed cautiously by means which include increased levels of observation and a review of medication. He thought the two risks of absconding and suicide were linked because she was most at risk of suicide when not in the ward.

As to levels of observation unless the team was absolutely sure her motive in absconding was merely to go home one would consider 30 minute observations though in his written material he thought they should be at least 15 minutes (which happens to accord with the Trust guidelines as we now know).

He accepted that such a requirement would preclude home leave but he was not deterred by that, indeed his opinion was that it was inappropriate for Mrs Savage to have any leave at all from 29 April onwards. To the extent that his evidence was to the effect that, once the real and immediate risk is recognised, then safety comes first even if it may have counter-therapeutic effects, he seems to me to be prefiguring Lord Rodger’s view which I set out below.

Dr Carter was called by the defendant and had very considerable experience of nursing at all levels. His view was that she was not a real and immediate risk of suicide. She had had copious opportunities to harm herself and had not taken them. Even when in 2001 she had absconded and wandered on the A130 she did not have appeared to have taken active suicidal steps such as throwing herself in the path of a car. I cannot find as a fact that she was not making a real suicidal attempt on that occasion, as opposed to merely having some sort of trouble crossing the road. Expressions of wanting to die do not indicate that the patient is imminently about to kill herself.

As to documentation he could not see any compliance with the Trust’s own policy about risk assessment and levels of observation and the documents that should be generated. He said that this was unfortunate but did not mean that these matters were not discussed and it was implicit in her being granted home leave in his opinion that she had been placed on general observations.

As to the note keeping he described the medical notes as having many entries which were “telegraphic in their brevity” and there was an absence of narrative. I agree with that assessment. But there is on my judgment no reason to doubt the evidence of Dr Abrar, who though inexperienced at the time I found a careful witness, that Dr Ague’s ward rounds lasted most of the morning and between 30 and 45 minutes were spent on each patient and when she examined Mrs Savage herself that would have taken her 45-50 minutes. So far as the nursing notes were concerned he found that these daily entries gave an account of her behaviour and progress and describes these as “good sensible observations” and in general I agree with that assessment. Therefore while there are no proper written risk assessments at any stage from 29 April onwards there was in his judgment good quality information to be derived from the copious entries in those notes as to how she was progressing.

In general I found him as a witness to be over-indulgent towards the Trust’s position in this case. He placed too much reliance on the fact that Mrs Savage had had copious opportunities to kill herself before the 5^th July, which is not really a valid answer to a claim such as this, and I was unimpressed with aspects of his written report, particularly the statement that he could “see no evidence of suicidal ideation” despite the fact that he had been supplied with the clearest evidence that there was.

The absence of adequate documentation as to risk assessment, reviews of risk assessment, a care plan and a scheme of observations is not of itself indicative of the defendant’s liability in this claim. The Article 2 obligation is not to keep good records of clinical steps taken. Reliable and credible evidence from the members of staff responsible for the care of Mrs Savage over the relevant period might be capable of answering the allegations levelled against the Trust.

Though mental health care is a multi-disciplinary or team exercise two key members of the team in this case, on any view, were the Consultant Psychiatrist Dr Claudio Ague, Mrs Savage’s responsible medical officer from the time of her final admission to the hospital, and Mark Gillett a registered mental nurse and ward manager of Chalkwell ward at the relevant time.

Dealing first with Mr Gillett, he cut a sorry figure in the witness box. He was the staff member with overall responsibility for patients and staff on the ward and accepted that he was responsible for ensuring implementation of Trust policies and procedures. When confronted with those policies he was visibly uncomfortable. The defendant’s embarrassment was compounded by the fact that he (and those representing the defendant) had failed to notice that the policies relevant to observations which were put before the court, very much at the heart of much of his evidence, and on which he was properly cross-examined at some length, were not in fact those current at the time.

He agreed that good risk assessment was fundamental to the care of a patient in Mrs Savage’s position that key events should be recorded and risk assessments had to be documented to be of value. He described a daily record of decisions on levels of observation as extremely important and agreed that there were no documents at all of that nature in this case, though he maintained that discussions would definitely have taken place about observations but he could not say between whom or when. Though in his witness statement he had put forward the Michael Manze document of 27 May as something to be relied on he agreed that critical events such as her expressions of suicidal ideation ought to have been on it and were not nor was past absconding. He could not answer why there was no risk assessment carried out by any member of hospital staff (Mr Manze was a social worker not based on the ward). The risk management plan of 28 May was not a plan for her maintenance in hospital, he accepted. Confronted with this document, which as I have said in my view was entirely worthless, he went no further than to concede it was “not a perfect risk assessment” and “it could be better”. For his own part he was unaware of the incident of November 2001, or could not recall it, and though he was unable to recall being told about her suicidal ideation he said that they could not take every word of each patient seriously or they would have all their patients on continuous observations. Such statements were something that they heard all the time.

He did not appear to be aware that Mrs Savage’s fears included fear of harm to herself as well as to her family (though the nursing notes show her as expressing that very fear on more than twenty occasions). He described the expectations of weekly one to one sessions between the patients and the named nurse as difficult to put into practice. He though that attempts to access Mrs Savage’s mental state were the staff’s responsibility.

Though he maintained that he did not view Mrs Savage as a real and immediate risk of suicide overall he did not strike me as a manager with a clear picture of Mrs Savage’s problems and risks or an appropriate grip on the procedures and policies to be adopted. In a sense it is possible to be sympathetic with him in that second ground. The complete policies of the defendant for the management of patients are 2-3 inches thick and can only be described as a gold standard. None of the staff who gave evidence seem to have had any specific training in them. They do give the air of being the product of managers removed from the front line. But the fact is the defendants formulated and promulgated them and cannot therefore be surprised to be held to them when events such as this occur. And as the case progressed I saw more and more clearly their utility and good sense.

As for Dr Ague he had been a Consultant Psychiatrist since about 1964 and was plainly nearing the end of what was no doubt a long and distinguished career. His notes were laconic and illegible (even at times by him) and of little value but I accept that he did carry out regular ward rounds with his patients one of whom was Mrs Savage and took appropriate time to do so.

As to the question of depression, he had himself prescribed Metrazapine for Mrs Savage when she was an out patient between 2002 and 2004 and did so because she had been prescribed it before and seemed reasonably well. He was unable to explain why it was stopped when she came into Runwell hospital. It was re-started on 11 May after a ward round on which either she or her husband or both had asked for it. He supposed she wanted it because she could not sleep although she was also being prescribed a hypnotic at the time. He was adamant that it was not prescribed for depression and he would not have done so.

He could not remember himself reducing the level of observations from the 30 minutes at which Dr Abrar had set them on 30 April and he did not, he accepted, make any positive decision to reduce them. Under the policy that must be a medical decision and certainly in my judgment the very inexperienced Dr Abrar would not have taken it on her shoulders. I believe he regarded the policies and guidance of the defendant as in effect bureaucracy.

He was alarmingly ignorant of the previous history. Even though the notes relating to her previous admission were on site and available he did not know about the episode on the A130 in November 2001, but thought he knew she had been “wandering about in a bizarre way”. He did not know that she had said that she wanted to die. He may not have been aware of the expression of suicidal ideation on 14 April. He said that Mrs Savage did not volunteer information easily, an experience that other witnesses including the family found, but he did not himself ask her directly about thoughts of self harm of suicide. He did not read any risk assessments and very rarely read nursing notes (though had he done so in my judgment in this case they were of much higher quality than his own in terms of the information they conveyed). He said it was not his habit to write about levels of observation but thought that he must have taken a decision to reduce the level 2 observation set by Dr Abrar “tacitly” when he allowed her weekend leave on 8 June and he presumed she had been on level 2 observations up to that time (i.e. for five weeks or so). His evidence, and now the Trust’s case before the court, was that every time he saw the patient on his weekly rounds he was conducting a risk assessment, though he accepted that in none of his notes is it possible to discern this. He said his risk assessment was his description of her mental health in those notes together with his own “medical behaviours”, for example reducing the level or frequency of leave. In all of his notes there is nothing spelt out about the risk of absconding, though he accepted that it was necessary to consider the consequences of that, or any opinions by him on the risk of self harm. In evidence he was relatively dismissive. She had made he said no suicide attempt in the past and historic factors he said were probably the most important, though he appears not to have fully informed himself about these.

Crucially he agreed, when asked by Ms Richards about the vital question about the consequences of absconding, that Mrs Savage at the relevant time was impulsive, felt persecuted and was terrified, treatment-resistant, was experiencing unpredictable psychotic thoughts and he agreed with the proposition that “anything could happen at any moment” so far as she was concerned.

There is nothing reassuring in his evidence, nor anything which enables me to say that, despite the manifest absence of proper documentation, he was and is justified in saying that she did not present a real and immediate risk of suicide. Interestingly his own view was and is that you cannot stop people committing suicide, but given the above description of Mrs Savage’s psyche as it appears from the notes what happened to her was exactly what could have been predicted.

Finally Dr Abrar. She recorded the expressions of self harm on 29 April and had to prompt her but succeeded in eliciting them. It was possible that she asked her the same sorts of questions when she spoke with her on later occasions, because she said that is now her habit and if she did so today and received a negative response she would record that fact in her note, but she did not do so at that time. She too said that every time she saw a patient she was undertaking a risk assessment but her extreme inexperience at this time does not mean that any significant reliance can be placed on that in her case. She agreed there should have been a review of her risk assessment every time she attempted to abscond. She thought she would not have taken the decision to down-grade the levels of observation and may not herself have been told about the various attempts of absconding in May and June.

In its decision in this case the House of Lords considered the application of the so called Osman/Keenan test defining the duty of the State under Article 2 to take steps to prevent persons killing themselves, specifically in the context in which Mrs Savage was namely a detained patient in a mental hospital.

Lord Rodger considered the Strasbourg and domestic cases on this issue and held that the operational duty under Article 2 was capable of extending to such persons. He observed that persons in custody were vulnerable and authorities were under a duty to protect them, as was plain from the many decisions bought on behalf of prisoners under this Article. He also considered the cases involving military conscripts committing suicide and the duty towards hospital patients and persons suffering from mental illness. Having completed this review he expressed himself as satisfied that the Trust was under a general obligation by virtue of Article 2 to take precautions to prevent suicides among detained patients at this hospital and concluded (paragraph 66) :-

“In my view it is abundantly clear that, where there is a real and immediate risk of a patient committing suicide, article 2 imposes an operational obligation on the medical authorities to do all that can reasonable be expected of them to prevent it.”

and he added (at paragraph 72):-

“The operational obligation arises only if members of staff know or ought to know that a particular patient presents a “real and immediate risk” of suicide. In these circumstances article 2 requires them to all that can reasonably be expected to prevent the patient from committing suicide. If they fail to do this not only will they and the health authorities be liable in negligence but there will also be a violation of the operational obligation under Article 2 to protect the patient’s life. This is comparable to the position in Osman’s case 29 EHRR 245 and Keenan’s case 33 EHRR 913.”

Baroness Hale (at paragraph 97) agreed that :-

“…the positive protection obligation under Article 2 is generally an obligation to have proper systems in place. But in some circumstances an operational duty to protect a particular individual is triggered. The latter duty is not engaged by ordinary medical negligence alone. [Counsel for the claimant] is not arguing for any less demanding test than the test derived from Osman’s case….and Keenan’s case…. He accepts that this is a different and more stringent test than ordinary medical negligence. For my part I have little doubt that it is right in principle to apply the approach adopted in Osman’s case and Keenan’s case to patients detained in hospital under the Mental Health Act 1983 as it applies to persons detained under other powers in other institutions.”

It is plain therefore that there are two stages in the exercise that I face. The first is to decide whether the defendant had the requisite knowledge, actual or constructive, of a “real and immediate risk to the life” of Carol Savage from self harm and the second is whether the defendant failed to do all that could reasonably have been expected of it to avoid or prevent that risk.

The test depends not only on what the relevant authority knew but also what it ought to have known (see Van Colle v Chief Constable of the Hertfordshire Police [2009]1 AC 225 at 32 ). The relevant knowledge is what they knew or ought to have known at the time and the court must warn itself against the dangers of hindsight. The authorities are clear that there is a high threshold to be crossed before the test is satisfied. Lord Rodger stressed (paragraph 41) that the House had been at pains to stress that the threshold of real and immediate risk to life for triggering the duty is high; that it will be harder to establish than “mere negligence” (paragraph 66); and the test is different from and more stringent than ordinary medical negligence (paragraph 97). In Van Colle Lord Bingham at paragraph 30 said

“The test of real an immediate risk is one not easily satisfied, the threshold being high, and I would for my part accept that a court should not likely find that a public authority has violated one of an individual’s fundamental rights or freedoms, thereby ruling, as such a finding necessarily does, that the United Kingdom has violated an important international convention”

Lord Hope in Van Colle at paragraph 67 and 68 said in three places that the threat must be “imminent”. He added that the test was in clear terms and provides an objective test which requires no further explanation. While therefore the original source for the interpretation of the test must be the ECtHR cases themselves I consider that the views of their Lordships in these two recent domestic decisions, explaining what the European jurisprudence amounts to, are important and influential on my decision.

Once the test is triggered by the requisite knowledge on the part of the local authority, and consideration turns to whether they had done all that could have reasonably be expected of them, unsurprisingly the ECtHR has held that this is a question that can only be answered in the light of all the circumstances of any particular case – see Osman and Opuz v Turkey [2009]27BHRC159 at 130. More specifically in Savage Lord Rodger said (at 42) that this second limb duty does allow for competing considerations but -

“… the immediacy of the danger to life means that, for the time being, there is in practice little room for considering other, more general matters concerning his treatment. There will be time enough for them, if and when the danger to life has been overcome. In the meantime, the authorities’ duty is to try to prevent the suicide”.

Finally, at this second stage the test for causation is not the English “but for” test, but a looser one; the claimant does not have to show to show that had the trust acted appropriately there would probably have been no death, but merely that she has “lost a substantial chance of this” – Van Colle, per Lord Brown of Eaton-under Heywood at paragraph 138.

It is agreed that my task is to approach this question on the basis that the ECtHR has approached similar cases in the past and apply the principles those cases establish to the facts I have found. That Court’s general approach seems to me to be to set out the facts of the case fairly fully, as I hope I have done above, state the test and then simply state its finding that violation of the Article is or is not established. To that end I have been referred to a number of the Court’s decisions in Article 2 cases, all of which I have considered, albeit none of them exactly covers this case in the sense of relating to a person detained under mental health legislation for reasons both of therapy and protection of the patient herself, as opposed to being under a loss of liberty for other reasons. I hope I will be excused from not setting out the facts of these cases. What matters is the overall impression of what the Court’s approach to issues of this kind has been

At the forefront of his argument Mr Bishop refers me to examples of four cases where no violation has been found, starting with Osman itself, a case whose facts were described as “comparatively extreme” and “very strong” by Lord Brown in Van Colle (at 115 and 125), and Keenan, where though there was a finding that the risk was real, the prisoner’s periods of apparent normalcy precluded a finding that it was immediate. That case featured a complete absence of any entry in his medical notes for 12 days before his suicide and no reference to him in the occurrence book for 5 days before.

He also relies on two domestic decisions. First in Van Colle itself Lord Bingham, having held that the Court of Appeal’s decision could not stand, made ten specific findings of fact at paragraph 36 as to what the detective in question knew or ought to have known at the time and whether he ought to have concluded they showed there was a real and immediate risk to the intended witness. The approach is most instructive, but the decision itself is not, so utterly different was the whole context of the case. Secondly Rabone (cited below) where the claim was rejected on the ground that the operational obligation did not extend to mental patients who were not detained under the 1983 Act., but Simon J nevertheless went on to consider whether the test was satisfied, and held that the risk on the facts of that case was real but was not immediate.

For her part Ms Richards puts forward Ataman v Turkey (Appeal No 46252/99) involving a conscript with mental health problems who was assigned guard duties with a loaded firearm. She places particular reliance on Edwards v UK [2002] ECHR 46477/99 where the applicants’ son was in custody and another prisoner with a known propensity to violence and who was behaving bizarrely was put into the same cell and kicked him to death. The main feature of interest for my purposes is that the finding that the risk was or ought to have been known about was plainly reached by aggregating a number of failings at different times by different persons and bodies who failed to communicate and share knowledge about the attacker.

As to knowledge of the risk, actual or constructive, in this case I must start by reminding myself of the dangers of hindsight, which are particularly strong in a case such as this which culminates in such a dramatic and tragic outcome which can all too easily colour a fact finder’s view of what led up to it. I find the defendant through its staff either had or ought to have had knowledge on the evidence available to it of the following main salient features of Mrs Savage’s position:-

There is no doubt that she presented a real and immediate risk of absconding. Dr Turner rightly accepted this. As to whether that also indicated a similar risk of suicide, I am also of the view that it did. There was little or no risk of that while she was on the ward, or at home with her family. But once she was out in the world on her own such was her psychotic state of mind it truly was the case that anything could happen at any moment and the risk of suicide must be assessed as both real and immediate.

As to whether the defendant did all it could reasonably have been expected to do the answer to that must be that it did not. At the least there was a real prospect or substantial chance that had she been made subject to level two observations at 15 or even 30 minute intervals she would not have slipped away unnoticed in the way she did on the 5 July. Mr Bishop points out that she was on 15 minute observations in November 2001 when she escaped onto the A130, but the answer to that is that the claimant does not have to prove that her mother would probably have survived had they been in place and observed in July 2004. Mrs Savage was probably not so determined or skilful an absconder as to require her detention in a locked ward, but that would have been a matter for proper assessment by her team led by the RMO. Had that been necessary, then even if it would have run the risk of reinforcing her paranoid thoughts her safety would have been the overriding concern. But in my judgment all that was required to give her a real prospect or substantial chance of survival was the imposition of a raised level of observations, which would not have been an unreasonable or unduly onerous step to require of the defendant in the light of the evidence in this case.

Section (7) of the Human Rights Act 1998 states that :-

“For the purposes of this Section, a person is a victim of an unlawful act only if he would be a victim for the purposes of Article 34 of the convention if proceedings were brought in the European Court of human rights in respect of that act”

Article 34 itself merely states:-

“The court may receive applications from any person… claiming to be the victim of a violation by one of the High contracting Parties of the rights set forth in the convention…”

The question of victim status, therefore, can only be settled by looking at other decisions of the ECtHR.

The claimant says she is an indirect victim and points to a number of other cases in which relatives have been accorded the necessary standing to bring a claim based on an Article 2 violation. These have extended to siblings – Renolde v France [2008] ECHR 5608/05 at Para 69; parents – Edwards v UK [2002] ECHR 46477/99; the father mother and son of the deceased – Cats v Ukraine [2008] ECHR 29971/04 at 94; and a nephew – Yasa v Turkey [1998] ECHR 22495/93 at 63-66. The last of these claimed to be acting “as a person who is himself affected and not as his uncle’s representative” and the court held that as such he “could legitimately claim to be a victim of an act as tragic as the murder of his uncle”.

Mr Bishop for the defendant argues that Lord Scott’s analysis in Savage is right, notwithstanding the fact that standing was not an issue before their Lordships for decision. I prefer Miss Richards’ argument to the contrary. She points out that the cases referred to by Simon J in Rabone v Pennine Care NHS Trust [2009] LS Law Med 117-8 are cases in which victim status has been thrown in doubt by virtue of settlement of other proceedings, which was a feature of Rabone itself. Mr Bishop seeks to argue from those cases to say that in this case, though there have been no other proceedings, let alone any settlement of them, where as here there was an available domestic civil remedy which Mr Savage for valid reasons decided not to pursue, the position is similar to that prevailing in cases of compromised alternative claims. He argues that to qualify as victims applicants must either be representing the estate of the deceased or directly affected by the violations.

Miss Richards’ answer is that the concept of the victim under the convention is an autonomous one and is decided “irrespective of domestic concept such as those concerning an interest or capacity to act” - Micallef v Malta [Application number 17056/06].

It seems clear to me that the claimant is entitled to bring this claim as a victim. The deceased was her mother to whom she was close and, one might add, much of the deceased’s final illness centred around a deluded but sincere concern for the safety of the claimant. Though of course the claimant herself bears no shred of responsibility for her mother’s death, her own role in those deluded beliefs must be in her mind. I have no hesitation in regarding her as someone entitled to be treated as a victim for the purposes of this action.

In the light of all the above the claimant is entitled to a finding and declaration that the defendant has violated its positive obligation to protect the life of Carol Savage under Article 2 of the ECHR.

As to just satisfaction I have to determine whether to award compensation at all and if so the amount. Section 8(4) of the Human Rights Act requires me when deciding these questions to take into account the principles applied by the ECtHR in relation to the ward of compensation under Article 41 of the Convention. In R(Greenfield) v Secretary of State for the Home Department[2005]1 WLR 673 the House of Lords held that domestic courts should generally follow the approach of the ECtHR to this question and not aim to be significantly more or less generous in their awards. This was followed in R (Pennington) v Parole Board [2010] EWHC 78 (Admin).

That said it is far from simple to discern the principles which the ECtHR applies. The diligence of counsel has provided me with two schedules. Miss Richards’ schedule runs from the beginning of 2008 and sets out over 17 pages cases where that court has awarded compensation for non-pecuniary loss. The range is wide and seems to run from about €5,000 - €60,000. She suggests £30,000 would be appropriate in this case. Mr Bishop reminds me that in Van Colle the Court of Appeal (whose decision was not in the event varied by the House of Lords in this respect) reduced the estate’s compensation from £15,000 to £10,000 and the individual compensation to each parent from £17,500 to £7,500. I bear in mind that the claimant has openly stated that she has not brought this action for financial reward, as well as the fact that there is no element to be included for the so-called adjectival duty under article 2, there having been a full inquest which properly investigated this death. The amount I grant under this head, and I think it is right to make an award, can never compensate her for the loss of her mother and can only be a symbolic acknowledgment that the defendant ought properly to give her some compensation to reflect her loss. I assess the figure for just satisfaction purposes at £10,000.