Commissioner of Police for the Metropolis, R (on the application of) v Coroner Southern District of Greater London & Ors

MR JUSTICE DAVIS: On Friday 15 September 2000 Scott Robbins, a man of 27 years of age, was arrested on suspicion of the theft of two bicycles. Another man, Gary Rowley, was also arrested at the same time in connection with the same alleged offence. The two were taken to Croydon Police Station at around 12.15 in the afternoon. Following interviews, Mr Robbins was detained in a cell at the police station over the weekend, the intention by then being to return him to Camborne in Cornwall where, as it had been discovered, a warrant not backed for bail had previously been issued by the Camborne magistrates. At around 5.40 pm on Saturday 16 September 2000 Mr Robbins was found hanging in his cell from a ligature, constructed out of laces from his trainers and his belt, which had been attached to a light designed to be set flush, or nearly flush, in the ceiling of the cell. All efforts to resuscitate him failed and he was certified dead by 7.00 pm.

An inquest was in due course convened before a coroner and a jury, commencing on 12 November 2002. On 19 November 2002, at the conclusion of all the evidence, the coroner was invited to withdraw any finding of neglect from the jury. The coroner ruled against that submission. He proceeded to sum up to the jury. On 20 November 2002 the jury, by a majority verdict, brought in a verdict as recorded in the Inquisition that "Scott Cecil Robbins killed himself, contributed to by system neglect".

Realistically, there could be no challenge by anyone to the view that this was suicide, but the Commissioner for the Metropolitan Police was aggrieved at the finding and verdict of system neglect, and aggrieved by the ruling permitting that to be left to the jury. On 17 January 2003 he submitted a claim form, the coroner being named as respondent, seeking to quash the verdict in so far as it included the words "contributed to by system neglect". He is supported in that claim by four individual police officers as interested parties. He is opposed by the family, and most particularly the mother, of Mr Robbins as interested parties.

Permission was granted by Pitchford J on 21 February 2003. At the substantive hearing before me the claimant has appeared by Mr Michael Wood QC and Mr Daniels; the four police officers as interested parties by Ms Sibson; the respondent, being the coroner, by Mr Whitfield QC; and the family by Mr Simblet. I should add that Mr Whitfield made clear that in making submissions on behalf of the coroner he did so to assist neither one party nor the other but in an attempt to assist the court, and also in an attempt to set out the coroner's own position. I would like to acknowledge the conspicuous fairness and thoroughness of Mr Whitfield's submissions, which have greatly assisted me. I would also like to acknowledge the thoroughness of the submissions of all other counsel concerned, not least because they were appearing before a judge not over-familiar with the coronal system.

I turn to the background facts. Mr Robbins was born on 12 September 1973. He was brought up in Cornwall. He did not do at all well at school. He was, it appears, dyslexic although this does not seem to have been fully addressed at the time. On leaving school at the age of 16 he seems to have led an uncertain and fractured existence. Over the years he was convicted for a number of criminal offences, being convicted on no fewer than 21 occasions for some 79 offences. These included offences of burglary, driving offences and offences of violence against the person. Four of his sentences involved terms of imprisonment, the longest being one of three and a half years in total, imposed in 1997. At no time in connection with these offences was there any incident of self-harm or attempted self-harm on the part of Mr Robbins, whether in police custody or in prison. In addition, he had had a relationship from which there was a child. In September 2000, as it happens, he was in a relationship with a young woman in Brighton. Mr Robbins, it would appear, although he had been brought up in Cornwall, had relatively recently removed himself to the Brighton area.

In the context of his previous convictions, the alleged offence for which he was on this occasion arrested, being the theft of two bicycles, was relatively trivial. He in fact was, with Mr Rowley, arrested at East Croydon railway station, they being on a train from Brighton to London. He and Mr Rowley were then taken to Croydon Police Station. The two arresting constables, Police Constable Brooks and Police Constable Day, noticed nothing at all untoward about Mr Robbins and no concerns of any kind about his mental state or mood were raised.

At Croydon Police Station, Police Sergeant Meek, who was assisting Police Sergeant Thomas, booked in Mr Robbins and Mr Rowley. This was at 12.15 pm. As part of the custody record, Police Sergeant Meek filled in a prescribed form, 57M, relating to Mr Robbins. This recorded the following against the various standard questions set out on the form. Part A, question 1: "'There is help that police must give to people with reading problems.' 'Do you need this help?'" Answer: "Yes" (by way of a tick in the box against the word "Yes"). Question 2: "'There is special help that police must give to people with learning difficulties or learning disabilities (mental handicap).'" Neither box Yes or No is ticked, but someone has written in "reading/writing". Question 3: "'The police must give special help to people who went to a special school.' 'Do you need this special help?'" The answer against that is ticked as "No". Question 4: "'The police must also give special help to people who have a mental health problem or who suffer from mental illness.' 'Do you need this special help?'" The answer there was "No". Question 5: "'Do you need help for any other reason?'" The answer there was "No". Under Part B, which related to need for medical attention, the answer "No" was given to the question: "Are you suffering from any medical condition, illness or injury?" The answer "No" was given to the question: "Are you receiving treatment for this condition, illness or injury?" The answer "No" was given to the question: "Are you taking any medication?" Question 4 reads as follows: "Have you ever tried to harm yourself?" The answer to that was ticked as "Yes", and then there was recorded: "Five years ago -- bad relationship". The form was then signed by Mr Robbins and the relevant custody officer.

A custody officer's powers to remove a suspect's personal items or clothing is governed by section 54 of the Police and Criminal Evidence Act 1984, and specifically section 54 subsections (3), (4) and (5). In addition, matters are prescribed in Code C of the Police Codes of Practice. Paragraph 4.2 of Code C, in the relevant respects, reads as follows:

It was common ground before me that there is a continuing duty to consider removing clothing or items from a detainee imposed on each custody officer under whose care he comes throughout his detention.

In the light of Mr Robbins' answer about the previous incident of self-harm, Police Sergeant Meek decided that Mr Robbins should see the forensic medical examiner. On this date this happened to be Dr Farhan, who, as it happened, is a doctor approved under section 12 of the Mental Health Act 1983 and who had been a forensic medical examiner for over six years. Dr Farhan interviewed Mr Robbins at around 1.40 pm. He had seen the form 57M prepared by Police Sergeant Meek. He questioned Mr Robbins, amongst other things, about the previous incident of self-harm. This had, it appeared, involved an overdose of tablets at a time when Mr Robbins was homeless, in around 1994 or 1995. Dr Farhan formed the view that the previous attempt had not been serious, that the causes of it were now past and that the risk of self-harm had passed. There was, he considered, no continuing risk. The report, Form 83, which he prepared and which formed part of the custody record, reads in the relevant respects as follows. The examination on the form is recorded as beginning at 13.40 hrs and concluding at 13.51 hrs. Against the typed entry as to "Recommendations: Fit to be detained", the answer was "Yes; "Fit to be interviewed at this time", and again the answer was "Yes", although written in the margin in, I apprehend, Dr Farhan's writing are the words "Advise Appropriate Adult". Just below that are written the manuscript words "Cannot read properly. Dyslexic". Then against the heading "Medical findings/advice to police" are written the words "Drug Abuser -- Heroin. Also Dyslexic -- cannot read properly. Looks well -- alert. Has no medical complaints. No psychiatric features though appears a bit slow." Then under the heading "Management whilst in custody" are written the manuscript words "Routine check" and the form is then signed.

It can thus be seen that Dr Farhan made no particular recommendation at all in respect of Mr Robbins' continued detention, whether as to extra observation or the like, and specifically he made no recommendation to the police that they should remove items of clothing or other personal items as being potential ligatures.

At around 2.55 pm Police Sergeant Kenward took over responsibility as custody officer. Nothing was observed in Mr Robbins' demeanour to give any cause for concern or any cause to reassess the previous decisions made. Regular hourly checks of Mr Robbins in his cell, in accordance with standard practice, were continued. At 6.38 pm Mr Robbins was interviewed by two police officers, being Police Constables Brown and Collings. The duty solicitor, Mr Crome, was present on behalf of Mr Robbins. Mr Robbins was further detained. Neither of the two interviewing police officers nor Mr Crome noticed anything untoward about Mr Robbins' demeanour at that time and nothing gave rise to any concern on their part about his mood or mental state. At the end of that interview, Mr Robbins was informed by Police Constable Brown that he would not be released but would be detained and then taken back to Cornwall in due course. By this time the Croydon police had discovered, via the police computer system, the the existence of the warrant not backed for bail issued by the Camborne magistrates. Neither Police Constable Brown nor Mr Crome considered Mr Robbins to react adversely to this news. A little later Mr Robbins was told that he would not be taken back to Camborne until after the weekend. This was for practical reasons of transport. Again, no obvious reaction was noted.

During that evening Mr Robbins was observed coughing up blood. Another forensic medical examiner, a Dr Scerri, was called. This was at around 8.13 pm. Dr Scerri considered that there had been a ruptured blood vessel. Dr Scerri had no concerns as to self-harm arising out of his examination of Mr Robbins and made no recommendations himself, whether as to removal of clothing or otherwise, about Mr Robbins' continued detention. He recorded him as fit to be detained with a comment "Call FME if coughing of blood persists". At around 10.45 pm Police Sergeant Darbyshire took over as custody sergeant. No concerns as to Mr Robbins were noted overnight, which was uneventful. He slept in a cell in which, as I was told, there was a bed, blankets and a pillow with pillow-case, but no sheets.

Shortly after 7.00 in the morning, Police Sergeant Thomas took over from Police Sergeant Darbyshire. Mr Robbins ate a full breakfast at around 9.00 am. After that, at around 10.11 am, he called his girlfriend on the telephone, she being in Brighton. He was heard by Police Sergeant Thomas to tell her that he loved her, that he was missing her and that he could not wait to get back to see her again. No indication of depression or indication of any sign of self-harm was noted. Evidence, in fact, was given at the inquest that Mr Robbins throughout had been considered a "model prisoner". At around 2.10 pm Mr Robbins took lunch. At around 3.00 pm Police Sergeant Kenward took over as custody sergeant. Again, no concerns about Mr Robbins or his mood or behaviour had been noticed. The last check of him in his cell was at around 5.00 pm. At 5.40 pm an officer went to his cell in connection with the evening meal and there Mr Robbins was found hanging. All attempts to revive him failed.

As I have said, the inquest was conducted over some six days. The family of Mr Robbins were represented by counsel (not Mr Simblet). It is plain that counsel for the family was most assiduous in exploring all avenues of evidence as to how Mr Robbins had come by his death. Ms Sibson appeared on behalf of the police officers, Police Sergeants Thomas, Meek, Darbyshire and Kenward. Mr Daniels appeared for the Commissioner. Amongst other witnesses, all the relevant police officers gave evidence, as did Doctors Farhan and Scerri.

On 19 November 2002, at the conclusion of the evidence, submissions, which were reduced to writing, were made as to whether the evidence was such as to permit a finding of neglect or system neglect to be left to the jury. Counsel for the family in fact submitted that there should be left to the jury, amongst other possible verdicts, the following verdict "The deceased died as a result of a violation of his right to life contrary to Article 2 of the European Convention on Human Rights". Speaking for myself, I find it most surprising that, as the law currently stands at all events, it can be said that a jury at an inquest could ever be invited to consider so arcane and so quasi-legal, as well as so emotive, a verdict. But that was what was suggested. In addition, however, submissions were made as to whether, amongst other possible verdicts, verdicts of suicide contributed to by "neglect" or by "system neglect" should be left to the jury. Counsel for the family submitted that they should be, amongst other things submitting that there could be a free-standing verdict of "system neglect". Mr Daniels submitted that no verdict involving neglect, given the evidence, could properly be left to the jury at all. Ms Sibson also submitted that no verdict involving neglect, whether system neglect (so-called) or neglectalone, should be left.

The coroner was, of course, required to bear in mind, amongst other things, the Coroners Rules 1984. Rule 36(1) reads as follows:

On the morning of 20 November 2002 the coroner handed down his proposed ruling in writing. In the relevant respects it reads as follows:

Going back to the ruling, the coroner goes on:

This form of procedure is well established at inquests. It reflects the principle of the case of R v Galbraith [1981] 1 WLR 1039 which is so familiar in the criminal context. In the course of Galbraith, Lord Lane CJ said this (at page 1042 of the report):

There are, of course, certain differences for a coroner's court. For one thing, a criminal court is operating to the criminal standard, whereas a coroner's court very often is not. For another, a submission of no case to answer at a criminal trial is almost always made at the close of the prosecution case. Such an application, or the raising of such a point by the judge himself, at the end of the defence evidence is rare; whereas at an inquest such a submission is ordinarily (and perhaps reflecting the inquisitorial approach intended to be adopted) made at the conclusion of the totality of the evidence.

Having delivered his ruling, the coroner proceeded to sum up to the jury. The summing-up was detailed and thorough. At pages 65-68 in the bundle appears the coroner's detailed direction as to neglect and system neglect. Having reminded the jury that neglect was not the same as negligence and having reminded them that they were not to identify any individuals, he set out what, on the balance of probabilities, there was to be assessed. He directed them that there had to be a gross failure or failures to provide or procure basic medical attention for someone in a dependent position who could not provide it for himself, and gave detailed directions as to that. He went on to say this:

He then went on to give further directions. He said that the failure or failures could be in the system or could be on the part of individuals, either singly or taken together. He said that a series of shortcomings could amount to neglect, and he gave some examples. He then went on to say this:

It is plain from the way in which the matter was summed up to the jury in this context that the coroner was distinguishing between neglect on the part of an individual or individuals on the one hand and system neglect, as he called it, on the other hand. It is further clear that the coroner left it to the jury that they could return a verdict incorporating reference to neglect, that is to say, by reference to individuals (although of course none could be named), or alternatively a verdict incorporating reference to system neglect.

It is, I think, of some importance to note that no complaint or criticism with regard to the coroner's charge to the jury is made by the claimant in the judicial review claim form. Indeed, no criticism of it is made in the submissions on behalf of the family as interested parties before me. The attack of the claimant is on the coroner's previous ruling that there was sufficient evidence to justify a finding of neglect being left to the jury.

In his written argument and at some stages in his oral argument, as I understood it, Mr Simblet seemed to be submitting that the very fact that the jury had found that there was system neglect of itself demonstrated that there was sufficient evidence to justify that finding. Put like that, I cannot accept that proposition; and, in fairness, I should record that Mr Simblet ultimately, as I understood him, did not seek to maintain that particular stance. Whether the totality of the evidence adduced is such as to be capable of entitling a jury to find a particular verdict is a matter of law to be decided by the coroner. If it is not in law so sufficient, then a subsequent finding by the jury cannot displace that. That position, after a certain amount of debate, is now well established in the criminal law and practice. It should be, and in my judgment is, the same in coronal law and practice. At the same time, I must and I do bear in mind Lord Lane's observation that borderline cases are to be left to the discretion of the individual judge or, as in the present context, the individual coroner.

The modern starting-point for the approach to be adopted in such cases, where neglect is a potential issue at an inquest, can be taken as the Court of Appeal decision in the case of R v HM Coroner for North Humberside and Scunthorpe ex parte Jamieson [1995] QB 1. That case involved a prisoner who hanged himself in a prison hospital. The facts there were different to the present case, one such fact being that, in contrast to the present case, it was acknowledged inex parte Jamieson that the prisoner both was and was appreciated to be a suicide risk. The coroner in that case in effect withdrew from the jury the possibility of a verdict of suicide aggravated, or in circumstances brought about, by lack of care. The Court of Appeal, upholding a decision of the Divisional Court, held that even if the prison authorities had given the deceased the opportunity to take his own life, the circumstances could not support a verdict that neglect had contributed to his suicide and the coroner's ruling had therefore not constituted misdirection.

In giving the judgment of the court, Sir Thomas Bingham MR undertook an extensive review of the relevant statutory materials and reported decisions. The overall conclusions are stated as general propositions, being fourteen in number:

Mr Wood, supported by Ms Sibson, submitted that that authority in essence set out the following approach to be adopted in cases of this kind, which potentially might involve neglect. First, there must be a basis for finding that the deceased's condition was known, or should have been known, to be such that action was necessary. Secondly, there must be a basis for finding that the fact that the action was not taken amounts to a "gross" failure. Thirdly, there must be a basis for finding that the failure to take action had a clear and direct causal connection between the conduct said to amount to the neglect and the cause of the death. Mr Whitfield in essence agreed with that formulation. But he supplemented it by pointing out that, first, the failure to be demonstrated is not civil negligence but neglect in the sense of failure to provide care; secondly, that the adjective "gross" imports a value judgment and, provided all other elements are present, would typically be a jury issue (a point Mr Simblet also stressed); thirdly, that the causal connection did not require that the neglect, if any, was the sole cause: rather, it extended to contributory factors which had in causative terms to be of some substance, in the sense that they had to be more than trivial, but which could co-exist with other factors which could also be regarded as contributory. I agree with that formulation of Mr Wood and with the supplementary observations of Mr Whitfield.

Nevertheless, certain other points arise out of the ex parte Jamieson case, as it seems to me. I queried with counsel the rationale for the clearly stated requirement that the failure must be "gross". In the context of civil negligence proceedings (which are, it must be stressed, quite distinct from coronal proceedings) it has been said that to style the alleged negligence in any given case as "gross negligence" adds nothing save a pejorative epithet; although there may, of course, be instances where such an epithet actually has real legal relevance: for example, in the application in some cases of the doctrine of novus actus interveniens. Gross negligence also, of course, has a real part to play in, for example, certain kinds of manslaughter cases. But what is the rationale for the requirement in this jurisdiction? Ms Sibson submitted that the rationale is to link the finding of neglect to any finding as to causation (which would reflect, by analogy, the increasing trend in modern civil negligence cases not to regard questions of causation as independent from questions of liability). Mr Whitfield suggested that it may be a policy point designed to ensure that a jury at an inquest is not troubled with relatively trivial matters and designed to emphasise the need for cogency in evidence which is to be adduced by way of proof of this point. Be that as it may, and whatever the rationale for the requirement is, there is no doubt that it is a requirement: as laid down inex parte Jamieson.

A further point arising fromex parte Jamieson which gives cause for reflection is the requirement in proposition (6) that the verdict be factual, expressing no judgment or opinion. That accords with what is laid down by rules 36 and 42 of the Coroners Rules. But I must say that if a jury, when considering a verdict which may include a finding of neglect, in answering the question as to how the deceased came by his death is to do so by considering whether a gross failure has been identified, it is rather difficult, to my mind at least, to see how a judgment or opinion has not in effect, even if not "expressed", been stated and even if, as will always be the case under the present system, no persons or individuals are specifically identified in the verdict. The decision inex parte Jamieson emphasises that what is to be considered is neglect, not negligence: that being entirely consonant with rule 42. Similar sentiments have been enunciated in numerous cases decided both before and after theex parte Jamieson decision. The coronal system, it is to be noted, is such that those who may be said to be negligent may well not have been legally represented themselves at the inquest and will not have been able to rebut any such allegation in the way that they might under the procedures available in an ordinary civil negligence trial: a point stressed by Sir Thomas Bingham in his fourth proposition in ex parte Jamieson. A lawyer, I think, could understand all that. But, in actuality and at all events in terms of public perception, one may wonder how the line is to be drawn between neglect which involves a "gross failure" on the one hand and negligence on the other hand. Would a member of the public attending such an inquest, or a member of the public reading a report about such an inquest in the newspaper the day after, be able to appreciate such a distinction? Indeed, to take the example that Sir Thomas Bingham posed at proposition (11), if a finding of "neglect" can be appropriate in a case where a prison warder observed a prisoner in a cell preparing to hang a noose around his neck but passed on without any attempt to intervene, it can be said that such a finding of neglect would be regarded by many people as tantamount at the least to an attribution of individual blame, and probably also of gross negligence on the part of the prison officer concerned: even if so much has not been made "express" in the verdict, and even though the verdict as such cannot be so framed.

Mr Whitfield, while submitting, as I have said, that neglect in the coronal sense will not always involve negligence in the tort sense, frankly accepted that it may well be that, in the great majority of cases in practice, where there is gross neglect established there is the framework for saying that negligence has been established. It is the case that the theory behind inquests, and the legal framework applicable to inquests under the present rules, are inquisitorial and involve a neutral investigation by way of factual enquiry into how a deceased came by his death. The practice, however, may be different, or at all events may be becoming so. Witnesses are meant to be examined, not cross-examined: but counsel before me candidly told me that nowadays at inquests witnesses are frequently cross-examined, sometimes vigorously so. Inquests are meant to be inquisitorial: counsel before me, however, told me that as often as not inquests can effectively be adversarial, sometimes highly so. Blame of persons or attribution of liability is designed to form no part of the verdict at an inquest: but the motive of some interested parties at some inquests may well be (and even though for totally understandable human and personal reasons) precisely to seek to establish and attribute blame. Yet further, I do not think it over-cynical to wonder if, in at least some inquests, some of those participating may have a, even if not the, motive of seeking to establish neglect with a view to enhancing prospects in any subsequent civil proceedings or in settlement of any potential civil claims (and in saying that, I wish to stress most emphatically that the present is not such a case). At all events, in many cases, whatever the motive may be, it is not, I think, unreal to think that a verdict of neglect may at least have that as a consequence.

Against those observations, I turn to the case of Keenan v United Kingdom [2001] 33 EHRR 38. In that case a prisoner serving a sentence of four months' imprisonment at Exeter hanged himself while in prison. He had a history of mental illness, with a number of previous suicide threats and a number of recorded instances of deliberate self-harm. While in prison he on occasion behaved bizarrely. He complained of "cracking up". On one occasion a cellmate saw him fashion a noose from a bedsheet and reported that to the prison staff. A prison psychiatric reports noted evidence of aggression and paranoia. Shortly before he was due to be released he was sent to the segregation block. There he behaved bizarrely, sometimes saying, then denying, that he felt suicidal. He spent a period of some days in the segregation block. There were daily visits by doctors. On 15 May 1993 he was found hanging in his cell. In due course a jury at an inquest recorded a verdict of misadventure and the cause of death as asphyxiation by hanging. In the judgment of the European Court of Human Rights there was no violation of Article 2 of the European Convention on Human Rights, although a violation of Article 3 was found.

In the course of its judgment, the European Court of Human Rights said this:

In the course of giving his concurring opinion, Sir Stephen Sedley (who was sitting as an ad hoc judge of the court) said this:

I come then to the recent decision of the Court of Appeal in R (Amin) v Secretary of State for the Home Department; R (Middleton) v West Somerset Coroner [2003] QB 51, [2002] EWCA Civ 390. In that case the deceased, Mr Middleton, was found hanged while in prison. His family alleged that the prison knew that he was a suicide risk and should have put him on suicide watch. The coroner ruled that, because of the provisions of rule 42, a verdict of neglect could not be returned. The jury found that the deceased had killed himself while the balance of his mind was disturbed. A rider handed in by the jury indicated a failing of the prison service in its duty of care to the deceased, which rider the coroner refused to append to the inquisition. In the judgment of the Court of Appeal, delivered by Lord Woolf CJ, there was a detailed consideration of the statutory background and of the ex parte Jamieson decision. This was then said in the case of Middleton:

I was told that the application of this decision in Middleton to that in ex parte Jamieson, taken in conjunction with the provisions of rule 42, has in practice caused uncertainty to coroners and to those lawyers who involve themselves in inquests. All counsel before me expressed some uncertainty as to the precise reach of the decision in Middleton. I was told, however, that it was this case that has given rise to the concept of "system neglect", as now advanced at some inquests as capable of forming part of a verdict. In the experience of counsel appearing before me, and based also on their researches into the authorities, "system neglect" had not previously manifested itself as a potential part of a verdict to be left to the jury: although it may be that in a rather different context one can find elements of such a concept in section 8 of the Coroner's Act 1988, to which I was referred by Mr Whitfield.

In the latest edition of Jervis on Coroners at paragraph 12.94, a very blunt critique of the decision inMiddleton is advanced. It is even there suggested that the decision was reached per incuriam:although in my view that cannot be so. I was also told that permission to appeal from the decision of the Court of Appeal in the Middleton case has been given, and that the case is due to be heard at the end of the year by the House of Lords. The ambit of that appeal is not known to me, although I imagine that the relationship and compatibility of rule 42 with the European Convention on Human Rights may be the subject of discussion, amongst other matters.

At all events, a number of propositions can, I think, be extracted from Middleton which is, of course, binding on me. First, the imposition in the interim period of the Human Rights Act 1998 does not require thatex parte Jamieson is to be regarded as overruled. On the contrary, ex parte Jamieson continues to apply in general to inquests. That is explicitly stated in Middleton. Secondly, when, however, it is necessary, in order to "vindicate" Article 2, to give a verdict of neglect it is permissible to do so. Thirdly, if so, by reference to the facts of the particular inquest being conducted, where a finding can serve to reduce the risk of repetition of the circumstances giving rise to the risk of death, that may be done. It is to be noted that the Court of Appeal in Middleton seems to have thought that it was for the coroner and not for the jury to decide whether such a finding would reduce the repetition of the circumstances giving rise to the death.

It seems, from the reasoning and manner of expression in Middleton, that a distinction is indeed being drawn between defects in the system on the one hand as compared to individual acts of neglect or individual failings on the other hand. At all events, I gather that this is the view now taken in many coroners' courts; and it is the reason why, in this particular case, the coroner was being invited by the advocates to consider a direction to the jury on the basis that the jury could, on the appropriate findings, include a finding by way of contribution to the death either of "neglect" or of "system neglect". I would, however, observe that, as I read Middleton, nowhere is it explicitly stated (unless it be by reference to the "finding of system neglect" in paragraph 91) that that should be done by way of direction as to separate verdicts to a jury. The Court of Appeal for the most part refers simply to a verdict of "neglect" where appropriate "defects in the system" had been identified.

In the present case, at all events (in a summing-up of which, I repeat, neither the claimant nor the family make any criticism) the coroner instructed the jury in terms that would have suggested to the jury that they could either return a verdict of neglect, on the basis of individuals' failings, or alternatively a verdict of system neglect, if they identified a failure in the system itself. I should add that the coroner addressed the jury as to the basis on which they should approach their verdict generally in this context by reference to the three essential formulations set out inex parte Jamieson. Accordingly, as Pitchford J pointed out in granting permission, the claimant can have no complaint under that head.

Against all that, I turn to the facts of this case. It is to be noted that in making his ruling the coroner did not actually identify any factors capable of giving rise to a finding of neglect by the jury. Nor, in the course of his summing-up to the jury, did he specifically marshall the matters said to be capable of giving rise to such a finding. Further, in giving their verdict the jury, although courteously prompted by the coroner, did not add any rider or other indication as to what was the system neglect which was part of their verdict. All this prompted Mr Wood to make the forensic point that this was so just because, in truth, as he submitted, there was no sufficient evidence which could be identified by the coroner to warrant such a finding.

What then was the evidence capable of entitling the jury to find that the state agents (to use the phrase in Middleton, which would extend in this context to police officers and forensic medical examiners retained by the police) had or ought to have had knowledge of Mr Robbins' condition such that action was necessary? As I see it, on the evidence the only indication was the attempt at self-harm some five years before, as reported to Police Sergeant Meek and as duly recorded by him on form 57M. Police Sergeant Meek, in consequence, asked Dr Farhan, as the forensic medical examiner, to examine Mr Robbins. That examination, in the assessment of Dr Farhan, revealed nothing untoward at all. Nor did the later examination of Dr Scerri reveal anything. Nothing that the various police officers then or thereafter observed gave rise to any concern. Mr Crome, the duty solicitor, had no concerns when he attended on Mr Robbins.

Mr Whitfield, who in the course of his written submissions put forward every point that might be taken as going towards this particular issue and as to the possibility of a finding of neglect, pointed to a degree of inconsistency between the original account of the self-harm incident in around 1995 and what was actually said to Dr Farhan; and he also pointed to the denial, on one interpretation of what was said, of a history of self-harm in Mr Robbins' conversations with Dr Scerri. In addition, there was some evidence from Mr Rowley, who was in his cell, to the effect that he overheard a doctor or solicitor say (in the presence, he thought, of a police sergeant) words to the effect, "He's unstable". Mr Rowley did not, as he accepted, tell anyone of any concerns that he may have had about Mr Robbins. To the police, as I have said, Mr Robbins was a model prisoner.

In my view, and taking on board all the points made by Mr Whitfield and Mr Simblet, it simply is not possible to see how there was, or should have been known to be, a real and immediate risk of Mr Robbins seeking to take his own life. The points adverted to by Mr Whitfield, whether taken individually or collectively, in my view fall a long way short of beginning to show that.

The position on the facts here is entirely different from that in Keenan and Middleton. I think it striking that the coroner here, in the course of his ruling as to whether a violation of Article 2; or a verdict of system neglect should be left as a free-standing verdict, ruled that that could not be done; and he ruled that the position was quite different to that in Keenan. In my judgment the coroner was plainly right so to rule. One only has to look at Keenan to see how much more indicative the facts were of possible neglect, and one can at least see the basis for an argument under Article 2 of the Convention by reference to the facts of that case. But the court in Keenan rejected the argument that there was any violation of Article 2; and that must, a fortiori, be the conclusion here, given the evidence in the present case, taken at its very highest in favour of the arguments of the family. That in itself seems to me to be a strong pointer against the availability of a finding of "system neglect". For while, as I see it, a verdict of system neglect is by no means to be taken as co-extensive with a finding of a violation of Article 2, nevertheless the rationale for a verdict of system neglect does, in view of the statements in Middleton, at least have a connection with Article 2, whereby Article 2 is to be "vindicated".

It is in my view in fact extremely difficult to see just what the "system neglect" here could be said to be. A good deal of evidence at the inquest, at all events, seems to have been devoted to the training of the police at Croydon Police Station. In this context a booklet on deliberate self-harm had been published by the Home Office. A copy was at the Croydon Police Station but, of the four custody officers involved, only one had read it and at least three of them had not received any training as to it. At least one officer in the course of his oral evidence expressed resentment at not having been trained appropriately on those aspects.

The booklet is entitled "Deliberate Self Harm & Suicide by Detained Persons" and is designed to be an information book for police custody staff. At the opening page this is said:

Under the rubric "Who"? there are set out a number of potential indicators of those who might be vulnerable to the risk of deliberate self-harm. A series of factors is put forward. These, for example, amongst others include the mentally ill, drug and alcohol abusers, people with a history of family problems or relationship difficulties, people facing uncertainty about their future, people remanded or who have had bail refused unexpectedly, people charged with further offences, people suffering recent break-up of a relationship. I think there was some justification in Mr Wood's observation that the criteria put forward would extend to a very significant proportion of all persons detained in a police cell. Other detailed information is put forward in that booklet, which can fairly be said to be of a most helpful kind, although equally I think Mr Wood was also justified in pointing out that much of what is contained in it is really little more than common sense. However, I should make clear that Mr Wood expressly accepted that there had been a lack of training in this regard in this particular case.

But Mr Wood's simple point as to that was that, even accepting that that was so, there was nothing here to give rise to any gross failure in the system. Moreover, he submitted, it was not as if any lack of training here caused the police to take inappropriate steps in the light of the recorded indication by Mr Robbins of the previous attempt at self-harm some five years ago. On the contrary, the police summonsed Dr Farhan, the forensic medical examiner, and his assessment was that there was no cause for concern. He made no recommendations with regard to Mr Robbins' continued detention: which, as it happened, reflected precisely the view of all others who observed Mr Robbins both then and thereafter.

I can see no effective answer to those submissions.

Mr Simblet, at one stage in his argument, suggested that the jury could have concluded that there was neglect on the part of Dr Farhan, and from that alone could have concluded that there was "system neglect". But in my view such an isolated error, if there was any error at all, could not possibly amount to "defects in the system" as contemplated in Middleton as entitling a verdict of system neglect to be given. In any case, I do not think the evidence, taking it at its highest in favour of the family's arguments, could show anything more than an individual error of judgment on the part of Dr Farhan at the most; and not even that, as I see it, since it is very hard to see how Dr Farhan did not act reasonably in forming the conclusion that he did.

Mr Simblet in the event withdrew the suggestion that an individual failing could constitute "system neglect" but he suggested that the concatenation of failings identified by Mr Whitfield could entitle a jury to find "system neglect". He submitted that Mr Robbins was, or should have been, identified as at a risk of self-harm and that it should have been appreciated that further action was necessary. But, as I have indicated, I do not think there was evidence to support that. Some mention was also made of the light fitting, which was designed, in fact, to fit flush to the ceiling, and it is said that that in itself was source of danger. I am not quite sure what was meant to be made of that particular point. It seems to me that it leads nowhere if a prisoner has not been identified as being a risk, and at all events it cannot in itself be relied upon as evidence of system neglect.

The real emphasis was on the lack of training of the officers but again, in causation terms, that deficiency had no causative impact, even if such a failing in training could be described as gross, which I cannot myself accept. The incidents involving Dr Farhan and Dr Scerri could not, in my view, taken on their own, have come within the rubric of "system neglect". I should add that I was taken by counsel through the relevant passages of the transcripts of the oral examinations of the various relevant witnesses. None of this went so far as to give an evidential basis for a finding of gross failure.

The issue of alleged gross failure is linked to the issue of whether there was, in Sir Thomas Bingham's words, a "clear and direct causal connection" between the conduct established as gross failure and the cause of death. Here too, as it seems to me, and reflecting really what I have already said, the evidence taken at its highest was insufficient to entitle a jury to conclude that that had been made out. The conceded lack of training of the police officers does not provide any such causal connection, given the fact that a qualified medical officer was called and that that doctor made no recommendation, and had no reason to make any recommendation, for any special precautions with regard to Mr Robbins. Three of the police officers, when examined at the inquest, gave evidence to the effect that they could not, even after the event, say how they would have behaved any differently. One, it is true, gave evidence to the effect that in the light of his experiences on this day he always now removes ligatures from all detainees. That, I think, is perhaps a wholly understandable human response, after the event, to the trauma arising from what that police officer must have experienced (albeit a response which might raise questions under Code C): but it is hardly supportive of any causal connection to the death in this case.

Although the language of Sir Thomas Bingham as to causation would seem to be very clear and very emphatic as to what is, after all, in essence, generically a causative enquiry under rule 36, Mr Simblet cited to me a number of authorities which put possibly differing interpretations on such words. Thus, in the case of R v HM Coroner for Coventry ex parte Chief Constable of Staffordshire Police (2000) 164 JP 665 Tomlinson J gave a detailed exposition on the point and said this at page 676C-E:

This passage was approved by Penry-Davey J in the Divisional Court decision in the case of R v HM Coroner for Inner West London ex parte Scott [2001] 62 BMLR 222, [2001] EWHC Admin 105. That is one way of putting it, and certainly it emphasises that the neglect need not be the sole or even predominant cause of death: even if it can be said that that formulation does not readily accord with Sir Thomas Bingham's exact words, as observed by Jackson J in paragraph 77 of his judgment in R (Dawson) v HM Coroner for Kingston-upon-Hull [2001] EWHC Admin 352. It may be that, as stated by Sullivan J in the Divisional Court decision in N v Coroner for the City of Liverpool [2001] EWHC Admin 922, it is important not to apply Sir Thomas Bingham's words in an "over literal manner". But however it is put, at least some causational link must be capable of being established; and in my judgment, accepting Mr Wood's submission, the totality of the evidence (taken at its highest) was not capable of justifying a finding of any such causational link.

In sum, therefore, as I see it, there was not here the evidence capable of making good any of the three essential elements as laid down in ex parte Jamieson. As I have mentioned, in summing up to the jury the coroner gave the specific directions as to the possibility of a finding of neglect by reference to individual failings or of a finding of "system neglect". Given this clearly delineated alternative, and given the actual finding reached by the jury, Mr Whitfield conceded that the jury must have rejected the view that such individual failings, if any, as had been potentially identified on the evidence as capable of amounting to "neglect", had been rejected as justifying a finding of neglect. Consequently he conceded that were I to find -- as I have found -- that the totality of the evidence adduced was insufficient to justify the jury, properly directed, in returning a separate verdict incorporating "system neglect", it would follow that, subject to the exercise of my discretion as to the grant of any remedy, this verdict could not stand.

But I would not rest my conclusion simply on that concession, not least because there perhaps may be doubts as to whether a verdict incorporating "system neglect" as an alternative to "neglect" simpliciter is appropriate. In my judgment, and in the light of what I have already said, the evidence here could not have justified leaving to the jury a verdict incorporating neglect at all. In my judgment, that is so whether one applies the ex parte Jamieson test or even if, as was at one stage suggested, one applies some kind of lower test or threshold if Middleton so indicates that. Either way, as it seems to me, that position simply could not have been established on the evidence as adduced at this inquest.

In reaching this conclusion, I have borne in mind that the inquiry here is essentially factual. Considerations of whether an alleged failure is "gross" is a value judgment which ordinarily is to be assessed by the jury. In the ordinary way, causation questions are pre-eminently factual questions. I accept Mr Whitfield's submissions as to that, much emphasised also by Mr Simblet, and I also accept Mr Simblett's emphatically made point that questions of weight and credibility are jury matters, and that the jury's function is not to be usurped. But the question here is whether, at the close of the evidence, the evidence, in its totality and taken at its highest in favour of the family's arguments, could justify a verdict of neglect or of system neglect being left to the jury. In my view the evidence was not sufficient for that purpose; nor, in my view, was this sufficiently a borderline case which can and should be left to the discretion of the coroner.

The remaining question is what remedy, if any, should be granted. There are three possibilities: first, to quash the verdict recorded in the inquisition in so far as it includes the words "contributed to by system neglect"; secondly, to quash the verdict and direct a fresh inquest; thirdly, not to grant a remedy and to leave the verdict as recorded in the inquisition as it stands. A fourth suggestion by Mr Simblet that I should simply quash the reference to "system" is in the circumstances clearly untenable.

In my judgment the right course here is to quash the verdict to the extent that it includes the words "contributed to by system neglect". It cannot be just in my view to withhold the grant of a remedy altogether and to leave the verdict unchanged. That would leave on the record what some at least would be prepared to construe as a criticism of the police and also perhaps would involve in the public perception a possible stigma for individuals concerned in this matter. Nor do I think it necessary or desirable for there to be a fresh inquest. In a case where the court has decided that a potential verdict was wrongly withheld from the jury, the court may well have a strong initial predisposition to order a fresh inquest. But in the present case, as I have found, there was left to a jury a verdict which was not properly open to the jury, which is altogether a different matter. When I asked Mr Simblet what a fresh inquest, relating to a death which occurred over two and a half years ago, could achieve, I received no very clear answer. Certainly Micawberism is no justification for such a course. I have borne in mind the moving statement of Mr Robbins' mother, dated 19 June 2003, which indicates that she would wish there to be another inquest rather than there being a quashing of the part of the verdict relating to system neglect: but, even so, I have come to the conclusion that the right and just course is to quash that part of the verdict.

I grant relief on this claim accordingly.

MR DANIELS: My Lord, costs. There is a skeleton argument on behalf of the claimant, extraordinarily short. I did not presume upon your Lordship's judgment in this case, so paragraphs 1-3 can be safely ignored. Paragraphs 4 and 5 relate to costs against a coroner. It is not in any contention sought to take costs against any of the interested parties in this case.

MR JUSTICE DAVIS: What is the norm in cases of this kind? Where a judge makes a mistake in a summing-up, the judge is not ordered to pay the costs.

MR DANIELS: The judge, of course, would not have the right to appear, or would appear, in the Court of Appeal or make any representations to the court.

MR JUSTICE DAVIS: Mr Whitfield came along to try and help.

MR DANIELS: Indeed.

MR JUSTICE DAVIS: He did not take up an adversarial approach at all.

MR DANIELS: I do not contest that at all. The position is set out very usefully in the (?) Touche case.

MR JUSTICE DAVIS: Just tell me what that says.

MR DANIELS: In short, it says this. The circumstances have changed. There is a case calledex parte (?) Hay which accorded with what your Lordship has just said in respect of, particularly, where a coroner appeared on a more amicus basis between parties. In that particular case Simon Brown LJ found that he thought it "a strange distinction" that a coroner could appear and contest a case and also be on an amicus basis. With the greatest respect to my learned friend Mr Whitfield, he was extraordinarily careful to make his submissions as fair as possible, and I certainly do not seek to suggest to say he did not do so, but it is also fair to say that he did put, in very capable terms, the contrary argument, which he supported both in his skeleton argument and in his oral submissions, that there was sufficient evidence to go to the jury and therefore the coroner's decision was correct.

MR JUSTICE DAVIS: So what order did he say as to costs is appropriate?

MR DANIELS: Costs as appropriate.

MR JUSTICE DAVIS: He said the coroner can pay your costs?

MR DANIELS: The coroner should pay our costs.

MR JUSTICE DAVIS: And he can be left to recover them from his funding in due course?

MR DANIELS: Indeed. The difference between ex parte Hay and Touche fundamentally appears to be that in between the two authorities the method by which a coroner would fund such a judicial review, or fight against a judicial review, had changed and that the coroner is now supported by the local authority as opposed to having to fish in his own pocket for it.

MR JUSTICE DAVIS: Fine. Miss Sibson, do you want to say anything about costs at this stage?

MS SIBSON: I have no application to make.

MR JUSTICE DAVIS: No order as to your costs. Mr Whitfield, what do you say about costs?

MR WHITFIELD: My Lord, certainly the case of Touche -- and you may be assisted if I take you to it very quickly.

MR JUSTICE DAVIS: Yes?

MR WHITFIELD: It is tab 14, passages at paragraphs 50-59. The relevant passages are 55-59.

MR JUSTICE DAVIS: Yes.

MR WHITFIELD: My Lord, you were kind enough to express the view that, on instructions, I came along to help to the court as much as I did to defend the position of the coroner who was, in turn, naturally helped by the argument which he has heard. He has been here the whole time.

A factor in this case, you may think, is that the submissions made on behalf of the coroner did not materially add to the length of the hearing. They would have had to have been made in any event. Each case depends upon its facts. You have discretion to order costs against the coroner, you have discretion not to order costs against the coroner, in the circumstances of the particular case. It is very much for your decision.

MR JUSTICE DAVIS: So your submission is no order as to costs.

MR WHITFIELD: I would respectfully say that it is in this case.

MR JUSTICE DAVIS: Mr Simblet, what do you say. Are you in the same position as Ms Sibson, in effect?

MR SIMBLET: Yes.

MR JUSTICE DAVIS: Obviously, you are going to have yours from public funding or whichever assessment order ....

MR SIMBLET: In fact, my Lord, I am not publicly funded, so I shall be making no application one way or the other.

MR JUSTICE DAVIS: Thank you very much. Is there anything which you wish to say reply, Mr Daniels, to Mr Whitfield?

MR DANIELS: I think it is very clearly set out in Touche, and I do not seek to add anything to what has been said, other than simply saying this. I accept what my learned friend says about his position but on the basis of Touche it is apparent that that does not meet the argument. It may have met the argument previously but Touche says it does not.

MR JUSTICE DAVIS: Although the Touche case indicates that ordinarily the court may well order a respondent coroner to pay costs, I think the circumstances of this case are rather unusual and I think the fair and just order as to costs is that there be no order as to costs.

MR SIMBLET: My Lord, the only application I do have, is from counsel in a terribly invidious position.

MR JUSTICE DAVIS: Do not worry about that.

MR SIMBLET: But I ought as matter of professional propriety ask for permission to appeal. I do not have very much in support at the moment of the submission, as that it would have a real prospect of success. Obviously it is a legal judgment that your Lordship has made that differs from the legal judgment made by the coroner below and by the court below. To that extent it is a question of law and one in which the Court of Appeals might divine that there is a real prospect of success. I have been accused of Micawberism in terms of remedy. I do not want to be making the same mistake.

MR JUSTICE DAVIS: Not really, Mr Simblet.

COUNSEL: But nevertheless your Lordship is aware of the complexity and conflicting features of a great many authorities to which your Lordship has been referred. It is not a straightforward issue as to what is involved in neglect or system neglect at the moment, and it would be unfortunate if the family were unable to have the matter looked at if in fact -- as we know, a difference is made if, as we know is going to happen, there are going to be further cases discussing those issues, and it would be unfortunate if they were shut out from pursuing an appeal if the law were to change.

MR JUSTICE DAVIS: As it happens, I am sorry my decision took so long but it did occur to me while I was preparing this judgment that this case might be taken as a good benchmark in the House of Lords as to what is happening in coroner's courts at the moment, in the light of Jamieson and Middleton, and it did occur to me to spell it out so the House of Lords knows what is going on as an example, but the decision itself is simply on the evidence. I think really, Mr Simblett, in those circumstances I am not going to grant permission to appeal. If you can interest the Court of Appeal, good luck to you.

Is there anything else? May I thank all counsel who have been involved.