Buck & Ors v Nottinghamshire Healthcare NHS Trust

On 12^th March 2001 the respondents, who were six nursing staff employed by the appellants at Rampton Hospital, were assaulted by a patient, Miss Rachel Agar. They brought proceedings in respect of their injuries. His Honour Judge Inglis tried a preliminary issue dealing with liability and causation. He gave judgment in the respondents’ favour by a judgment dated 21^st February 2006. This is an appeal from that judgment.

I can take the context in which the incident must be placed largely from the judge’s judgment:-

“2.

Rachel Agar was born on 10 October 1974. From her early teens she displayed grossly disturbed behaviour. She passed through a variety of residential settings. In 1995 and 1996 she committed a number of minor criminal offences, and on 28 October 1996 she came to Rampton hospital under section 35 of the Mental health Act 1983, having been sent there by the Justices for assessment. She remained there until after the events with which this case is concerned. Her Mental Health Act status changed from time to time, and from January 2000 she was detained as though admitted to hospital under a hospital order without restriction.

3.

After her arrival she spent some months on the admission ward, creating, as Dr Travers (the consultant psychiatrist under whose care she came in early 2001) put it in his evidence, “mayhem”. She was transferred to Anston Ward and there over time spent a number of periods, including long periods, in seclusion. She displayed a high level of self harm; of assaults on others; and made many false complaints against staff that she herself had been assaulted. She was manipulative and very stressful to look after. In 2000 she was moved back to an admission ward, Alford. In late 2000 it was proposed that she move to Chatsworth Ward. The move happened more suddenly than had been intended, because of the temporary closure of Alford Ward through staff shortages. She moved on 12 January 2001. Dr Travers took over as the consultant responsible for her care from Dr Hamilton about 4 weeks later. There were joint meetings about her condition and her care.

4.

The history, over a number of years, was of very violent and unpredictable behaviour. She has a personality disorder, described sometimes in reports as a psychopathic disorder. There is some reluctance to use that description nowadays, but it is in fact the relevant category of mental illness under the Mental Health Act 1983. There had been debates about whether she should be in hospital at all, because the Mental Health Act criteria for hospital admission under court orders in the case of people with psychopathic disorders include a requirement that the court must be satisfied that treatment will alleviate or prevent deterioration in the condition. The question whether treatment at Rampton could have that effect remained a live issue. Miss Agar’s violent behaviour at Rampton led to her being prosecuted for offences of violence committed by her on staff there. Summarising the position on 21 September 2000 Dr Hamilton said in a letter:

“in this hospital Miss Agar has been extremely violent to staff and has appeared in court on several occasions but has only been convicted of common assault, although some of her attacks were unquestionably murderous in nature. She had a very lengthy period in seclusion which was only terminated with great difficulty about eight months ago”

Between 1 April 1999 and 30 March 2001 there were 570 recorded incidents, including 283 of self harm, and 120 assaults or threatened assaults on staff. During 2000 at various times she had caused 21 staff on Anston Ward to be off sick (though some of these absences may, it seems to me, have been due to stress caused by the pressures of looking after her). On her transfer to Chatsworth Ward on 12 January 2001 Dr Hamilton wrote in her notes:

Attack on staff could be of murderous severity. Risk indications are known and previously recorded:

e.g. - attempting to isolate staff

- complaining of feeling strange

- loss of eye contact

- ‘glassy’ stare to L or R of person talking to her. In the past moves of ward have resulted in a ‘honeymoon’ period. Indications above should be responded to as potentially serious and acted on by additional precautions.

Dr Hamilton remained Miss Agar’s medical officer until 13 February, when Dr Travers took over.”

In paragraph 5 of the judgment the judge summarised the incidents which occurred between the beginning of 2000 and the incident which occurred on 12^th March 2001. I will not set out the detail of those incidents. It is sufficient to say that during 2000 Miss Agar assaulted staff on a number of occasions. In the first months of 2001 she also carried out assaults on staff and threatened to kill staff or made threats to harm fellow patients and staff. Most significantly on 25^th February 2001 she attempted to strangle a staff nurse and then assaulted numerous members of staff. It was in the context of that assault that Dr Travers, who was her RMO, and to whose evidence I shall return in a moment, opined in notes in relation to that day that Miss Agar was “Tilt-positive”. The relevance of that assessment will become apparent hereafter.

The same notes indicated that when members of staff sought to interview Miss Agar they feared for their safety (see p 729 of the bundle). Miss Agar was placed “in seclusion”. This must not be confused with “confinement to her room”, the point with which this appeal is concerned. Seclusion involved Miss Agar being placed in a specially prepared room where a member of staff stayed on duty with her. Her period in seclusion lasted until the morning of 28^th February 2001.

On 1^st March 2001 various discussions took place as to the way in which Miss Agar should be dealt with. The notes of those discussions and a report of Debbie Anne Ford in connection with those discussions recognised the danger that Miss Agar was to herself, patients and members of staff. The note from Debbie Anne Ford included a recommendation that “at no time was Rachel to be escorted outwith the main ward area with less than two staff.” Further notes were “All incidents were documented and staff involved informed of risk of being a potential for future assaults victim.” “To minimise risk of self-injury or injury to others Rachel was supervised by a minimum of two staff when using tools or pens.” “Rachel was to retire to bed at 22.30 each night to reduce risk to self/others when staffing levels reduced.”

It was recognised by the staff that one objective of Miss Agar was, if possible, to get herself sent to prison. It is recorded in a note of Dr Travers that the view was taken at this time that it was in Miss Agar’s interest that she should be supported in her intention to get herself sent to prison. (See note of 2^nd March 2001, p 737).

On 7^th March 2001 Miss Agar again attacked a staff nurse and had to be secluded at 4.30 that evening (see p 737). Later that evening she was apparently settled in her bed and became calm and talkative (see p 737).

It was then on 12^th March 2001 that the incident with which this case is concerned occurred.

The details are set out in the judgment of the judge and I need only refer to certain features. Nurses, including the respondents, took over from the afternoon shift at about 9.40 pm. The normal complement for the Chatsworth Ward, on which Miss Agar was, was five night staff, but on this particular day only four were rostered since one was away attending an ill patient at another hospital. During the afternoon shift Miss Agar had been unsettled. By the time the nightshift came on Miss Agar was angry about certain antibiotics that she had been prescribed for a self-inflicted wound. Miss Agar was in an agitated state and at the trial there was a dispute as to whether there was one or two telephone calls calling for further staff. The judge found there was one call and that the call was not as urgent as certain of the witnesses would have indicated. Miss Agar was taken to bed in accordance with her care plan at 10.30pm. Shortly after that she came to the office asking for painkillers for a self-inflicted wound in her arm. She was offered paracetamol but she said that they made her sick and went back to her room. Within a minute or two she returned, saying she would have paracetamol and it was agreed to mix it with juice for her to have, and she sat down on a chair outside the office window. She sat in an armchair, apparently calm, to be given her juice. Suddenly she launched a violent attack on the fifth respondent. The other five nurse respondents went to her aid and a violent struggle ensued. Further details are unnecessary for this judgment. During the episode Miss Agar shouted words to the effect “do I have to kill someone to be sent back to prison?” All six respondents were injured in the incident in various degrees of seriousness and with long-term consequences for some of them.

By the conclusion of the hearing before Judge Inglis the allegations of negligence were limited to two specific allegations; the first allegation was that Miss Agar should, before the incident and because of her history including what had happened in February/March 2001, have been confined to her room at night, and the second was that the appellants had failed to provide adequate staff numbers. As to the second allegation, the judge found that the appellants had failed to provide adequate staff, but he found that it would not have made any difference if there had been more staff, having regard to the savagery of Miss Agar’s assault on that evening. There is no appeal against that aspect of the judge’s judgment. His finding of negligence in relation to the first allegation is what gives rise to the issue on this appeal.

To the second allegation there is some important background. Sir Richard Tilt carried out a review of security at high security hospitals following the Fallon Inquiry at Ashworth Hospital in 1998. He reported in February 2000. He made a number of recommendations, one of which was that the feasibility of locking “high-risk” patients in their rooms at night should be examined. The Tilt Recommendations were carried into effect by regulations entitled ‘The Safety and Security in Ashworth, Broadmoor and Rampton Hospitals Directions 2000’. These were Directions made by the Secretary of State for Health under the National Health Service Act 1977, which so far as material came into force on 30^th November 2000.

Paragraph 4 provided:-

“In order to promote conditions of safety and security in Ashworth, Broadmoor and Rampton Hospitals, each hospital authority is directed to exercise its functions in connection with the provision of high security psychiatric services in accordance with these directions.”

The important directions for the purposes of this case are paragraphs 30 and 31, which provide, so far as material, as follows:-

“Risk assessments

30.

(1) Each hospital authority shall ensure that as soon as is practicable after this paragraph comes into force each patient has a risk assessment carried out by his clinical team provided he has not had an equivalent assessment in the previous three months, and every newly admitted patient shall have a risk assessment by his clinical team as soon as is practicable after admission to the Hospital

(2)

The clinical team shall use the risk assessment to determine whether the patient presents a high risk of

(a)

immediately harming others

(b)

committing suicide or self harming;

(c)

being assaulted

(d)

escaping; or

(e)

organising action in collaboration with others to subvert security and safety.

. . . .

(4)

If the patient’s clinical team decide that the patient presents a high risk

(a)

of escaping and of harming others or

(b)

or organising action in collaboration with others to subvert security and safety

the patient’s clinical team must consult a member of the security department before finalising a risk management plan for the patient.

31.(1) Each hospital authority shall have a policy on the circumstances in which a patient, considered to be at high risk of matters set out in sub-paragraphs 30(2) following a risk assessment under paragraph 30, can be locked up in his room at night.

(2)

Where in accordance with its policy a hospital authority is minded to include locking up at night in a patient’s risk management plan the patient’s clinical team shall first consider whether

(a)

there are medical grounds (which in the case of a patient assessed as at high risk of committing suicide or self harm will take into account that risk) for not locking the patient up at night and

(b)

(following consultation with a member of the security department) there are other available measures which would be an effective alternative way to reduce the risk.

Concurrently the Secretary of State published guidance. As is apparent the directions were mandatory but the guidance was not, though where hospitals deviated from the guidance they were required to maintain a written record with the reasons for so doing. The guidance relating to paragraphs 30 and 31 provided:-

“An example of a protocol which might be followed in order to meet the requirement for an individual risk assessment of each high security hospital patient is attached as Annex C. The protocol incorporates arrangements for considering whether some high risk patients ought to be locked in their rooms at night.”

The protocol provided by its introduction as follows:-

“This protocol is designed to ensure that the public and the staff and patients in the hospital are protected from harm by addressing systematically the risk of harm to others and the risk of escape. The protocol enables the identification of all patients who present high levels of risk in specific areas and the safe management of risk, including locking patients in their rooms at night.”

The protocol indicated the four main categories of risk, in accordance with the 2000 Directions, “Risk of harm to self (suicide or self injury), risk of harm to others, risk of escape, risk of being assaulted (i.e. high vulnerability).” By paragraph 7 it was provided that:-

“Where following this protocol would suggest a patient should be locked in their room at night but this is not pursued, the reasons for the exception should be recorded.”

By paragraph 10 it was provided that:-

“10.

A decision to lock patients in their rooms at night (see Box 3 and Box 4 of attachment 2) can only be justified if the aim is to maintain the safety of patients, staff and the overall security of the establishment. Locking patients into their rooms at night should be supervised containment and frequent monitoring and review of the patient would be necessary. The local seclusion procedure should be referred to as a model of good practice in this respect, thus ensuring any necessary changes in the patient’s management are made in a timely manner, to address changes in the patient’s clinical presentation.”

The attachment contained a decision tree for risk management of high-risk patient group. Following that tree the indications were that if the high risk was of suicide or self harm “the patients should normally not be locked in their rooms”, but the tree provided for consideration of other boxes, including “high risk of immediate harm to others”, where if the answer was there was a high risk the tree indicated that “these patients should be locked in their rooms at night unless deployment of other measures reduces risk”. This box recommended consideration of management strategies in box 4. Consideration of box 4 “high risk of immediate harm to others” indicated that locking into room should apply “until judged safe to end such locking in – in accordance with seclusion policy”, and also indicated that locking into room could occur “for identified high risk periods only, (e.g. night time)”.

The directions were published in June 2000 and provided that the directions should come into force on “30^th November 2000” except for a paragraph which is not of relevance in this case. Thus by 30^th November 2000 the appellants should have had a policy on the circumstances in which a patient, considered to be at high risk, might be locked up in their room at night. None of the three hospital authorities, Ashworth, Broadmoor or Rampton, introduced a policy in accordance with the Tilt recommendations by 30^th November 2000, and indeed Rampton did not produce such a policy until the year 2002. Mr Herbert before us accepted that by so failing the hospital authorities were in breach of the Directions but he stressed there was no claim, nor could there be, for breach of statutory duty by the six nurses.

Dr Davenport, who was a joint expert who gave evidence at the trial, explained the position at paragraph 3.26 of her first report in the following terms:-

“When the security Directions 2000 were introduced there was considerable concern expressed by clinicians about the practice of confining patients in their room at night as it was considered a retrograde and anti-therapeutic practice more befitting a prison than a hospital. 24 hour care as prescribed by the SHA in 1991 was regarded by most clinicians in high security practice as essential to the provision of a therapeutic environment. Any resistance was chiefly located within a group of staff who belonged to the Prison Officers Association (POA) where practice was traditionally more custodial. At the material time, Rampton had not implemented the Tilt recommendations and did not have a procedure for confining patients to their rooms at night, even if they were regarded as high risk. Even if such a protocol had been implemented, it would have been extremely rare for a woman patient to have been locked in her room at night as a consequence of being identified as high risk with the Tilt protocol. Almost all women identified as high risk would also have been high risk within the second category, i.e. that of committing suicide or self-harming. It is not modern practice to confine to a room at night a person at high risk of self-injury as the risk of a completed suicide attempt is enhanced. This is only considered when the risk of harm to others is much higher than the risk of suicide and the room can be made safe overnight (i.e. by removing all items which could be used in self-injury). During the period when I practised in the high security women’s service at Ashworth and used the Tilt high security protocol to assess women patients, none were confined to their room at night under the protocol because of the high risk of a completed act of suicide.”

Dr Davenport, in her first report, commented on the circumstances of these assaults in the following terms:-

“At the time of this assault Tilt’s protocol had not been implemented at Rampton Hospital and therefore confining Miss Agar in her room at night was not an option.”

At paragraph 8.6 she also expressed the view:-

“In my view confining Miss Agar to her room at night may have reduced the likelihood of this particular assault occurring. This option was not available, as previously stated, on the night in question.”

In the addendum report produced just before the trial on 20^th January 2006, Dr Davenport said as follows:-

“2.3

If the Tilt protocol had been implemented at the material time, there is no doubt that Miss Agar would have been identified as high risk of perpetrating assaults against others. According to the protocol this would have triggered further clinical discussions with the security liaison department and addressed the issue of whether she should be proactively locked in her room at night. On balance, given the severity of assaults perpetrated by Miss Agar in the months prior to indexed assaults on 12^th March, in my view it would have been appropriate for the clinical team to discuss locking Miss Agar in her room at night. The attempted strangulation of the nurse in the clinic room on 25^th February prompted a team discussion which was recorded as “Tilt-positive for risk to self and others”. However this incident also demonstrated many of the difficulties of managing Miss Agar safely when she is locked in as her self-injury continued whilst in seclusion and she required restraint whilst in seclusion to prevent further self-injury. If the Tilt protocol had been triggered for this particular individual in advance of general implementation it would have prompted further team discussion with the security liaison department as to the relevant risks and benefits of continuing to manage her within the confines of the existing risk-manage-care plan. . . .”

At paragraph 2.14 she repeated her view adding this:-

“In view of the severity of the assaults perpetrated in the months prior to the indexed assaults, I believe it would have been reasonable to lock her in her room at night as her behaviour represented risk to others at the very severe end of the spectrum of behaviours normally manifest in women detained in high security.”

Dr Davenport gave oral evidence before the judge. She was asked her view as to what would have been the position if the Tilt protocol had been in force at the time just before the assaults, the subject of this action. She was not always consistent in her answers. She, for example, at one stage said that she thought it would have been a reasonable decision not to confine Miss Agar in her room (see p 637 lines 15-16). Ultimately, however, she was pressed by Mr Cotter to clarify her position and she did that over a series of answers. What Mr Cotter was seeking to do was to feed in the notes of the nursing staff, particularly the note suggesting that Miss Agar should be put into bed by 10.30 to reduce the risk to the reduced number of staff on at night. Dr Davenport then, when pressed by him, gave evidence that “it would have been a reasonable decision, hypothetically, to confine her to her room at night.” At that stage the judge intervened to explain that the case was concerned with an action being brought by people assaulted by Miss Agar and not by Miss Agar. He stated that “the actual question would then be, would it have been unreasonable not to do so,” to which Dr Davenport answered “I agree it would have been unreasonable not to confine her to her room at night if the protocol had been in existence and if adequate weight had been given to the views of the nursing staff around the risk of assault at night.”

It is fair to say that Dr Travers, in his written statement at para 13, p 297, said:-

“On reviewing the nursing records of the behaviour of Rachel on the preceding days and on 12^th March 2001, there is nothing to suggest that Rachel would have been a suitable patient for the locking in her room under the Tilt protocol.”

However, when he came to give evidence under cross-examination by Mr Cotter, in answer to Mr Cotter’s question “As at 27^th February you were aware that because she was Tilt-positive, high risk, that if you were following the guidance and the protocol you should have confined her at night. You understood that as at 27^th February?” To which Dr Travers answered:- “If appropriate alternative ways of managing her weren’t in place.”

Thus it was that there was powerful evidence before the court to the effect that if the Tilt directions had been complied with by the appellants, there would have been a risk assessment, assessing Miss Agar as an exceptionally high risk of posing serious injury to others and that it would have been unreasonable not to take the view that she should be confined in her room at night.

Mr Herbert before us sought to argue that the evidence showed that it would have been reasonable not to confine her at night. He took two points. First he suggested that the evidence of Miss Agar being high risk to herself would have been an indicator against confining her. Secondly, he suggested that the evidence did not show her to be a high risk at night as opposed to generally and that would have been a contra-indication to confining her at night.

Following the decision tree, it is possible to see that a balance has to be struck where a patient is at high risk of harming him or herself. But, where there is a very serious risk to third parties of serious injury, the decision tree would indicate that confinement at night was required and there is no reason to think that Dr Davenport did not have Miss Agar’s potential for harming herself in mind in giving the answers that she did.

So far as night time is concerned, the judge dealt with that in his conclusion that albeit there was no indication that the high risk which Miss Agar posed was enhanced at night, there was equally no evidence that the high risk she posed was reduced at night. Again, there is no reason to think that Dr Davenport, and indeed Dr Travers, in giving the answers they did, did not have the history of Miss Agar and her actions at night or in the day in mind.

If, therefore, the content of the Safety and Security Directions 2000 informed the content of the duty of care owed by the appellants, the case against the appellants, on the evidence, was a very powerful one. Indeed, my understanding of the judge’s judgment is that it was his view that the Directions did inform as to the standard of care, and it was on that basis and that basis alone that he found the appellants liable.

In King v Sussex Ambulance Trust [2002] ICR 1413 at 1419, Hale LJ (as she then was) said this:-

“The starting point is that an ambulance service owes the same duty of care towards its employees as does any other employer. There is no special rule in English law qualifying the obligations of others towards fire-fighters, or presumably police officers, ambulance technicians and others whose occupations in the public service are inherently dangerous: see Ogwo v Taylor [1988] AC 431. Such public servants accept the risks which are inherent in their work, but not the risks which the exercise of reasonable care on the part of those who owe them a duty of care could avoid. An employer owes his employees a duty to take reasonable care to provide safe equipment and a safe system of work, which includes assessing the tasks to be undertaken, training in how to perform those tasks as safely as possible, and supervision in performing them.”

Later at 1420 she said this:-

“If a removal firm cannot remove furniture from a house without exposing its employees to unacceptable risk, then it can and should refuse to do the job. The ambulance service cannot and should not do that. But that does not mean that they can expose their employees to unacceptable risks. They have the same duty to be efficient and up-to-date and careful of their employees’ safety as anyone else. It does mean that what is reasonable may have to be judged in the light of the service’s duties to the public and the resources available to it to perform those duties . . . . .”

In the Court of Appeal decision in Cook v Bradford Community Health NHS [2002] EWCA Civ 1616, Lord Justice Schiemann put the matter in a rather similar way in a case concerned with an employee in a psychiatric hospital being injured by a dangerous patient. The patient in that case was in a “seclusion suite”, (a method of holding patients to which reference has been made in the instant appeal). Miss Cook was injured when a decision was taken by two health care assistants to allow a patient to come out of seclusion while Miss Cook was still standing in the observation area. Schiemann LJ said this at paragraph 17:-

“The present situation as the judge held was one where her [Miss Cook’s] presence was arguably necessary before the patient left the seclusion room but was not necessary afterwards, and the health authority who has the difficult task of looking after these patients should not expose their employees, however well-trained, to needless risks. There is no avoiding exposing employees to risks. Manifestly the closer your dealings are with a patient, the greater the risk. If your function is merely to bring coffee on this particular occasion, there is absolutely no need for you to be close to the patient. So the judge held. It seems that she had effectively fulfilled or could have fulfilled her function (one does not know the detailed finding on that) but for my part I see nothing wrong in the approach which has been adopted by the recorder.”

Mr Herbert in this appeal submitted on behalf of the appellants as follows:-

“The standard of reasonable care is that which is reasonably to be demanded in the circumstances. One of the circumstances in this case was the nature and extent of the duty of care owed by the hospital authorities to Miss Agar and their responsibility in treating Miss Agar’s mental illness.”

No-one would quarrel, as I would see it, with that formulation. However, his submission continued:-

“In relation to the exercise of that judgment or discretion and in relation to decisions made pursuant thereto a court should only find negligence on the part of either the clinicians who made the decision (for which the defendant hospital authority would be vicariously liable) or the defendants themselves if there was a failure to act in accordance with the practice accepted at the time as proper by a responsible body of persons of the same profession or skill:- see Bolam v Friern Hospital Management Committee [1957] 1WLR 582; Phelps v Hillingdon LBC [2001] 2 AC 619 per Lord Clyde at p 672f, and Carty v Croydon LBC [2005] 1 WLR 2312 per Dyson LJ at para 26.”

It is this latter submission which seems to me to be looking at matters from the wrong angle. Clearly the duty that the appellants owed to Miss Agar is of relevance in considering the duty that the appellants owed to their employees, but it does not follow that the duty owed to employees can be tested simply by the question whether what occurred amounted to or did not amount to a breach of duty to the patient, which is what the Bolam test is concerned with. If the appellants can take precautions so as not to expose their employees to needless risks and still not be in breach of their duty to a patient, then it seems to me that they may well be in breach of duty if they fail to take those precautions. The question whether they were in breach of duty will be tested by reference to the principles applicable as between employer and employee, not as between a doctor and his patient.

I turn next to the relevance of the Directions and the Tilt protocol. First it is right that the claimants have not sought to bring actions for breach of statutory duty. Second it is right to stress that the Directions did not impose any absolute duty on any other hospitals to confine patients at night or anything of that nature. Indeed the Directions did not lay down the precise policy which any hospital was bound to have. We were shown the policy that was ultimately introduced by Rampton, headed “Rampton Hospital Policy on Patients’ Confinement in their Rooms at Night”. The aim of the policy was said to be to provide “advice to clinical teams on the circumstances in which a patient assessed as high risk may be locked up in their room at night. It also outlines procedural guidelines that must be complied with where the decision is taken to implement the policy.” Mr Herbert referred us to certain paragraphs which, he would suggest, gave slightly less encouragement to confining patients to their room than the suggested Tilt protocol. For example, paragraph 4.1 emphasised “there are some patients for whom locking in a room would increase not reduce the risk they present. For this reason patients presenting a high risk of committing suicide; and/or presenting a high risk of self harming, must not be locked in their room at night under this policy unless the exception of 4.2 applies”. Paragraph 4.2 provided “Where there is a combination of risks, and on balance the other risks make locking in at night appropriate, patients who present a high risk of suicide and/or present a high risk of self harm may be locked in at night, providing consideration is given to the environment in a way which reduces these risks.”

He also placed emphasis on paragraph 5, where it is provided at 5.1 “locking patients in their rooms at night is a last resort and should only be considered where all other risk reduction strategies have been considered and no suitable alternative exists.”

Mr Herbert accepted that the appellants were in breach of the 2000 regulations by not having a policy as at the time of the incident with which the case is concerned. He would suggest that since the breach is not one for which the appellants could be liable to the claimants, the question whether the appellants were negligent or not must be viewed in the context of a hospital without such a policy. In particular, he would submit that the clinicians who took the decision in this case as at 12^th March 2002 were not negligent in the absence of any policy and that thus the appellants could not be vicariously liable for the decision taken.

I would accept that it may well be that the appellants were not liable in this case vicariously for any failure by the staff of the appellants. I do not however accept that the 2000 Directions, which by March 2002 had not been adhered to by the appellants, had no relevance to the question whether the appellants are liable or in breach of duty to their employees. A breach of regulations can be evidence which can assist the argument (see Spencer v Boots the Chemists [2002] EWCA Civ 1691 at para 13). I would suggest a breach can help to inform the content of a duty. What, as it seems to me, the 2000 Directions demonstrate is that, if they had been implemented, there were methods whereby the appellants could protect other patients and staff from needless risks where high risk patients were concerned without being in breach of any duty to the patient. That is as true of the policy ultimately brought in as it was of the protocol, the subject of the guidance. It is ultimately for the court to decide the content of the duty and whether there has been a breach, but it is in that sense, as it seems to me, that the existence of the Directions and the failure to implement the same can inform as to the content of the duty owed to the claimants.

The judge recognised the difficult and complicated area of patient care in a high security hospital. He recognised the background of therapeutic development of such care over the last fifteen or twenty years, but his conclusion was that Rachel Agar was an “exceptional patient” by the standards of high security hospitals and she posed an “exceptional risk”. If, therefore, there had been a policy related to confining patients in their rooms at night at the time of the incident on 12^th March 2002, if any patient was to be confined, Rachel Agar must have been that patient. If she had been confined then no-one challenges the judge’s conclusion that the attack on the fifth respondent would never have taken place and the incident which caused serious injury to the respondents generally would not have taken place.

In my view, in concluding that there should have been a rigorous risk assessment as the Tilt Directions contemplated, and in taking the view that following such an assessment the appellants should have had a policy contemplating the confinement of Miss Agar in her room at night, the judge was imposing the appropriate standard of care on the appellants in relation to their employees. In holding that there was not a rigorous assessment, which would have contemplated confining Miss Agar to her room at night and in finding that if there had been she would have been confined in her room at night, the judge’s decision cannot be criticised.

I would accordingly dismiss this appeal.