England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE DIVISIONAL COURT
LORD JUSTICE KENNEDY & MR JUSTICE MORRISON

		Royal Courts of Justice Strand, London, WC2A 2LL Wednesday 21st March 2001

LORD JUSTICE SIMON BROWN:

1. On 6 February 1999 Laura Touche gave birth to twins, delivered by caesarean section. On 15 February 1999, tragically, she died. She was only 31. She died from a cerebral haemorrhage, the result of severe hypertension, possibly secondary to eclampsia. The medical evidence suggests that had her blood pressure been monitored in the immediate post-operative phase her death would probably have been avoided.
2. The critical issue raised in these proceedings is whether such a death is natural or unnatural - whether, in particular, an inquest must be held into it pursuant to s.8(1) of the Coroner's Act 1988 which requires such an inquest "where ... there is reasonable cause to suspect that the deceased ... has died ... an unnatural death".
3. It is the coroner's contention that Mrs Touche died a natural death. Her husband contends the contrary. He is anxious for an inquest. The Divisional Court (Kennedy LJ and Morrison J) on 22 June 2000 accepted Mr Touche's argument and directed that an inquest be held. The coroner now appeals to this Court.
4. Perhaps not surprisingly the case has attracted some attention: the facts, after all, are heart-rending. The issue raised, however, is essentially one of law and its resolution cannot depend on sympathy.
5. With that brief introduction let me turn at once to set out such further facts as need be stated.

The facts

6. The deceased was delivered of healthy twins at about 10.25 p.m. on 6 February 1999 by caesarean section under spinal anaesthetic at the Portland Hospital in London. Her pregnancy and labour had been uncomplicated. Following delivery her blood pressure was noted to be 120/60 which was within normal bounds and at around 11 p.m. she was transferred to the postnatal ward. She was complaining of headache. The next note of her blood pressure was at 1.35 a.m. when it was recorded to be 190/100. By then her headache was severe and she was clearly unwell. Only at this stage did treatment begin and her blood pressure start to be taken regularly until finally it fell to normal limits. By then, however, it was too late. At 5.15 a.m. she was suffering a left-sided hemiplegia. At 6.15 a.m. she was transferred to the Middlesex Hospital and from there to the National Hospital for Neurology and Neurosurgery at Queen Square where eight days later, on 15 February, she died.
7. A hospital post-mortem examination, carried out by Professor Scaravilli on 18 February 1999, recorded the cause of death as:

"1a. Brain swelling and tonsillar herniation

b. Intra cerebral haemorrhage

2. Recent pregnancy"

8. It was some months before Mr Touche's investigations into the circumstances of his wife's death led him to seek an inquest. On 28 July 1999 he wrote to the coroner referring to the 2½ hour period between 11 p.m. and 1.30 a.m. when it appeared that Mrs Touche's blood pressure had not been monitored. On 26 August 1999 his solicitors wrote, alleging that "a basic, fundamental failure to record blood pressure readings ... vitiated any opportunity to avoid the catastrophic events which lead to Mrs Touche's death". On 15 September 1999 the solicitors wrote again saying:

"The Portland Hospital have already confirmed in writing to our client that a protocol does not exist to reflect the level of monitoring that should be given following a caesarean section. We have expert evidence to the effect that every NHS hospital in the country has a protocol in place for the care of patients in the post-operative phase in order to maintain standards within the hospital and ensure an appropriate level of patient care. It is disturbing that a private hospital with this reputation chooses not to adopt such a protocol."

9. On 31 August 1999 the solicitors obtained a report from Dr Bogod, an experienced consultant anaesthetist with a particular interest in obstetric anaesthesia. He was very critical of the lack of records relating to the periods during and after surgery. In particular he found the failure to monitor or record vital signs, including blood pressure, at a time when Mrs Touche was receiving pain relief "astonishing" and described the level of neglect as "starkly apparent".
10. The coroner himself took the trouble to obtain a report from Professor Rubin whose particular interest is in the medical aspects of pregnancy. His report of 29 February 2000 pointed out that maternal death in the United Kingdom is now "very rare" and described Mrs Touche's death as "extraordinary" because he has "looked after countless numbers of pregnant and post-partum women who have blood pressure in the range recorded in Mrs Touche [and] none has ever had a stroke".
11. The final report to which I must refer was obtained by Mr Touche's solicitors on 15 May 2000 by way of preliminary opinion from Dr Williams who runs a high-risk obstetrics service at the Chelsea and Westminster Hospital and who has a particular interest in pre-eclampsia. His essential conclusions were first that the failure to undertake blood pressure readings during the post-operative period involved "sub-standard practice", and second that the deceased's severe hypertension was responsible for her cerebral haemorrhage and that "it is likely that more prompt identification and treatment of her hypertension would have prevented her cerebral haemorrhage".
12. In short, the evidence as a whole provides clear grounds for suspecting that the Portland Hospital failed to monitor the deceased's blood pressure as it should have done in the critical post-operative phase and that such failure was an effective cause of her death in that, but for it, she would probably (or at least possibly) not have suffered cerebral haemorrhage and died.

The Coroners Act 1988

13. S.8(1) provides:

"Where a coroner is informed that the body of a person (the deceased) is lying within his district and there is reasonable cause to suspect that the deceased -

(a) has died a violent or unnatural death;

(b) has died a sudden death of which the cause is unknown; or

(c) has died in prison or in such a place or in such circumstances as to require an inquest under any other Act, then, whether the cause of death arose within his district or not, the coroner shall as soon as practicable hold an inquest into the death of the deceased either with, or, subject to sub- section (3) below, without a jury."

14. S.8(3) provides so far as relevant that the inquest must be held with a jury "if it appears to a coroner ... (d) that the death occurred in circumstances the continuance or possible recurrence of which is prejudicial to the health or safety of the public or any section of the public".

The issue arising

15. The central question to address is whether, in the light of the facts already summarised, there is reasonable cause to suspect that Mrs Touche died an unnatural death. I pose it in the present tense (and earlier summarised the evidence as it now stands) because the coroner's stance has remained the same throughout: whilst he "entirely accept[s] that ... the post operative monitoring would appear wholly inadequate", he does not regard the death as having been unnatural. He deposes as follows:

"I asked myself whether this was a case in which the defects and human fault complained of lifted the case out of the category of natural and into a category of unnatural death and, applying my commonsense as a coroner, I concluded that it did not."

16. In the ordinary way, of course, it is for the coroner to decide whether there is reasonable cause to suspect that a particular death is unnatural, and his decision will not be challengeable unless it is Wednesbury unreasonable or involves a self-misdirection in law. The facts here having now substantially crystallised, however, the point has been reached where really there can only be one correct answer to the central question and that answer must necessarily depend on what is meant in s.8(1) by "an unnatural death". Clearly there is reasonable cause to suspect that the circumstances of Mrs Touche's death are those indicated by the evidence already summarised. Assuming they are, is it properly to be regarded as an unnatural death? If so, subject to a single qualification to which I shall shortly return, the coroner has no alternative but to hold an inquest. If, however, the death is not to be regarded as unnatural he has no power to hold an inquest.

The ruling authority

17. The correct approach to take to the question whether there is reasonable cause to suspect that a deceased has died an unnatural death was decided by this court in R v Poplar Coroner ex parte Thomas [1993] QB 610. The deceased in that case died aged 17 from a severe attack of asthma. The evidence suggested that she would not have died had the ambulance which was called arrived promptly rather than after a thirty-three minute delay. The late arrival of the ambulance notwithstanding, this court concluded that the coroner had been entitled to regard the death as natural and so not hold an inquest. The leading judgment was given by Dillon LJ who at page 628 said this:

"Whether Miss Thomas' death was natural or unnatural must therefore depend on what was the cause of death. At this point, I remind myself of the observations of Lord Salmon in Alphacell Limited v Woodward [1972] AC 824, 847 where he said:

'I consider ... that what or who has caused a certain event to occur is essentially a practical question of fact which can best be answered by ordinary common sense rather than by abstract metaphysical theory.'

Lord Salmon repeated what he had there said in his speech in McGhee v National Coal Board [1973] 1 WLR 1, 11 ..."

18. Dillon LJ then considered five possible explanations for the delay including "(v) the ambulance came late because the ambulance crew were inefficient and the management was slack", and continued:

"I do not suggest that any of these scenarios fits the facts of Miss Thomas' case. I do not know what the cause of delay was. But in each of these scenarios common sense indicates that what caused the patient's death was, on Lord Salmon's test in Alphacell Limited v Woodward [1972] AC 824, 847, the asthmatic attack, not the congestion of the traffic, the bursting of the water main, the malfunction of the computer or the inefficiency of the ambulance service. But the asthmatic attack is a natural cause of death, and the death is not, in my judgment, turned into an unnatural death by any of the facts suggested in any of the alternative scenarios. ... The coroner ... was saying that, even when all the other evidence is taken into account, the cause of death was still the asthmatic attack and the death was not an unnatural death. That is also my view for the reasons I have endeavoured to give."

19. Farquharson LJ agreed with Dillon LJ's reasoning.
20. As the third member of the court, I agreed with the outcome but "reach[ed] that conclusion with more hesitation than Dillon and Farquharson LJJ and by a rather different route". At page 630 I said:

"I agree ... that the question whether or not a death is natural or unnatural depends ultimately on the view one takes as to the cause of death. But I do not find the question of causation in this context susceptible of quite the same sort of robust approach that the House of Lords advocated in a very different context in cases such as McGhee v National Coal Board [1973] 1WLR1. The question arising there was: can the court properly infer, in the absence of a provable direct link, that one particular state of affairs caused or contributed to another. In those cases the possibility of there being more than one cause was immaterial. ... The question posed in the present context is surely therefore different: given that all the important facts are known to the coroner, what view should he take of causes that may well be secondary but are not self-evidently irrelevant? As in litigation why should he not sometimes find a death to be the result of two causes, either one of which could serve to make it unnatural."

21. A little later I indicated that I for my part would have regarded the death as an unnatural one "if the late arrival of the ambulance had constituted a more extreme failure of the service", adding that "by failure I mean culpable human failure on the part of those responsible for providing a reasonably efficient emergency service". I concluded:

"... it seems to me necessary to recognise that cases may well arise in which human fault can and properly should be found to turn what would otherwise be a natural death into an unnatural one, and one into which therefore an inquest should be held."

The Divisional Court's judgment

22. In their judgment below the Divisional Court said that they were "unable to detect any conflict between" my judgment and that of the other two members of the court in Thomas, and that leading counsel appearing for the coroner before them (not Mr Burnett QC) had "realistically accepted that on occasions a coroner may have to find there was more than one cause of death. That possibility was simply not canvassed by Dillon LJ". Then followed the first of two critical passages in the judgment:

"So where, as in this case, a patient is in hospital suffering from a condition which if not monitored and treated in a routine way will result in death, and, for whatever reason, monitoring and treatment is omitted, then, as it seems to us, the coroner must hold an inquest unless he can say that there are no grounds for suspecting that the omission was an effective cause of death. That seems to us to be the conclusion to which one is led by a careful analysis of Thomas."

23. Later in the judgment, having reviewed the evidence, and set out the coroner's own conclusion, comes this passage:

" ... we would prefer to see the coroner asking himself a question along the lines indicated earlier in this judgment, namely whether there are any grounds for suspecting that the wholly inadequate post operative monitoring and the consequential loss of the opportunity to provide timely treatment was an effective cause of death. If the coroner had approached the matter in that way it seems to us that his conclusion must have been different. ... In dealing with the statutory test omission can be as important as commission, and that, as it seems to us, is what ... the coroner failed properly to recognise and to evaluate."

24. There is one further paragraph in the judgment of the Divisional Court to which I should refer before turning to the central arguments advanced on appeal. Under the heading "Other Matters" appears this:

"Nothing in this judgment is concerned with what may in due course be the appropriate verdict, so we have not found it necessary to consider 'lack of care' or two of the decisions to which we were referred (R v Southwark Coroner ex parte Hicks [1987] 1 WLR 1624 and R v North Humberside Coroner ex parte Jamieson [1995] QB 1)".

The arguments on appeal

25. Mr Burnett QC for the coroner submits, first, that the Divisional Court was wrong to have found no conflict between my judgment and that of the other Lords Justices in Thomas and wrong, therefore, to have applied my approach rather than that of Dillon LJ. On Dillon LJ's approach, submits Mr Burnett, the coroner was clearly right, or at the very least entitled, to conclude that Mrs Touche died a natural death. Secondly he argues that the coroner in fact directed himself in accordance with my approach rather than the majority approach in Thomas and yet nevertheless properly came to the conclusion that this was a natural death. The Divisional Court, in other words, went further even than I had gone in Thomas. On their judgment, submits Mr Burnett, there would have to be an inquest every time a death takes place in hospital which might have been avoided but for a failure to provide some routine monitoring or treatment. Indeed, he suggests, coroners in future will be required at the outset, when first notified of a hospital death, to embark upon a detailed examination of the facts to see what if any routine treatment should have been provided. The Divisional Court, he suggests, has blurred the clear line established in Thomas; confusion and uncertainty now reign.
26. Mr Havers QC for Mr Touche resists that argument at all points and submits that the judgment of the Divisional Court is correct for the reasons they gave. Alternatively he seeks to uphold the decision on the much narrower ground that it would be open to the coroner (or jury) in this case to return a verdict that the death was caused or contributed to by neglect, in which event, as is common ground, there would certainly have to be an inquest. The verdict which Mr Havers contemplates is that the deceased died from natural causes to which neglect contributed - see R v North Humberside ex parte Jamieson [1995] QB1, 25. Such a verdict (which hereafter, for convenience, I shall call simply a "neglect" verdict) is the preferred modern version of what in R v Southwark Coroner ex parte Hicks [1987] 1 WLR 1624 was called "lack of care". Dillon LJ in Thomas expressly accepted that "another instance of deaths which are unnatural but not violent is where persons die from 'lack of care' in the narrow and somewhat technical sense in which that term was interpreted by the Divisional Court in [Hicks]"
27. As already noted, the Divisional Court in the present case expressly disavowed any such basis for their decision and this narrower argument was only introduced into the appeal at a late stage, a respondent's notice being settled for the purpose only during the hearing before us. It is nevertheless convenient to take it first.

Neglect

28. Sir Thomas Bingham MR, giving the judgment of the court in Jamieson, conducted a wide-ranging review of all relevant statutory and judicial authority (including not least a number of earlier cases concerned with lack of care verdicts) and stated fourteen general conclusions as to the essential nature of the coroner's jurisdiction. It is a landmark decision in coronial law, given in the context of a prisoner who had hanged himself in a prison hospital cell. Mr Burnett usefully distilled such of those conclusions as affect the present appeal (in particular conclusions 7 - 12) into the following propositions: 1. self-neglect is a gross failure to take adequate nourishment or liquid or to obtain basic medical attention or adequate shelter or warmth; 2. neglect is the obverse of self-neglect; 3. neglect means a gross failure to provide or procure basic medical attention for someone in a dependent position (for example because of illness) who cannot provide it for himself; 4. the need for the basic medical attention must be obvious; 5. the crucial consideration is what the condition of the dependent person appeared to be; 6. neglect can rarely if ever be an appropriate verdict on its own but it may be factually accurate to say that it contributed to a death; 7. neither neglect nor self-neglect should ever form part of a verdict unless a clear and direct causal connection is established between the conduct so described and the cause of death. I did not understand Mr Havers to dissent from this analysis.
29. It follows from this that the critical questions now to be asked under this head are whether, on the evidence presently available, there is reason to suspect, first, that there was a gross failure by the Portland Hospital to provide Mrs Touche (indisputably a dependent in their care) with basic medical attention, and, second, that her need for such attention was obvious at the time.
30. Mr Burnett submits that in addressing these questions the court is not concerned with considerations of fine judgment such as are generally in play in medical negligence actions. The concept of "neglect" involves failure which is, as he puts it, plain as a pikestaff - note the words of emphasis in the Jamieson formulation: "gross", "basic", "obvious". The hospital's conduct here, he submits, cannot properly be stigmatised as involving a gross failure to meet an obvious basic need.
31. In submitting the contrary, Mr Havers relies in part on Dr Bogod's characterisation of the hospital's failure to monitor Mrs Touche's blood pressure as "astonishing", and its level of neglect as "starkly apparent"; in part on the coroner's own recognition that the post-operative monitoring was "wholly inadequate"; and in part on the acknowledged rarity of maternal death in the U.K. Such a death is simply not to be expected nowadays and its very occurrence, submits Mr Havers, points strongly to a failure of care. There is, the evidence suggests, a basic need for routine blood pressure monitoring in the immediate post-operative phase following a caesarean section under spinal anaesthetic. NHS hospitals apparently meet that need; the Portland Hospital does not.
32. I find Mr Havers' argument on this part of the case compelling. That, of course, is not to say that if an inquest is now held the coroner (or jury) will be bound to qualify the inevitable verdict of death from natural causes by a reference to "neglect". That would inevitably depend upon the evidence as it emerges and the coroner's (or jury's) evaluation of it in the light of appropriate legal directions (or self-directions) based on Jamieson. Still less, let me make plain at this point, am I indicating any view upon the merits or prospects of success of a very substantial damages claim which apparently Mr Touche has outstanding against the Portland Hospital. Those proceedings are entirely separate from these and everything I have said is without prejudice to them. Who knows what evidence the hospital may have? Rather it is to conclude no more than that upon such material as is presently vailable to the coroner he could not properly decide otherwise than that there is reasonable cause to suspect that Mrs Touche's death was (a) at least contributed to by "neglect" (narrowly defined as by Jamieson) and thus (b) unnatural (as would necessarily follow from Thomas).

The jurisdiction issue

33. Given this conclusion on the issue of "neglect", an inquest would ordinarily be required. It is now that I come to the qualification I referred to in paragraph 16 above. The coroner's jurisdiction to hold an inquest under s.8 depends upon his being informed of the presence of a body within his district. The coroner in the present case was so informed by his officer on 16 February 1999, the day following Mrs Touche's death. But the information then provided to him, from a doctor at the National Hospital at Queens Square, was that the deceased:

"Gave birth to twins by caesarean on 6.2.99 at Portland Hospital. Collapsed three hours later. Admitted to National Hospital on 7.2.99. Exam indicated spontaneous brain haemorrhage unconnected with surgical procedure. ... No evidence of neglect nor complaint by family. No PM required."

34. Unsurprisingly in the light of that report, the coroner did not consider it appropriate to hold an inquest. The deceased in the event was cremated on 22 February, the procedures set out in the Cremation Regulations 1930 being duly observed. These included the completion of a series of prescribed forms by which Mr Touche as the applicant and two doctors acting respectively as the medical attendant and the medical referee certified amongst other things that they had no reasonable cause to suspect that the deceased died an unnatural death or from "privation or neglect", and that there was no reason for any further inquiry or examination. Had the coroner, of course, decided at the outset to hold an inquest, the body would have remained in his charge.
35. It was not until after the Divisional Court's judgment that the coroner became aware of Mrs Touche's cremation. The jurisdictional question, therefore, has arisen only in the course of this appeal. It arises because a coroner cannot decide to hold an inquest under s.8 unless at the time of his decision a body remains in existence. It presents, however, no insuperable problem. The solution lies in s.15 of the Coroner's Act 1988 which, so far as material, provides:

"(1) Where a coroner has reason to believe -

(a) that a death has occurred in or near his district in such circumstances that an inquest ought to be held; and

(b) that owing to the destruction of the body by fire or otherwise ... an inquest cannot be held except in pursuance of this section, he may report the facts to the Secretary of State.

(2) Where a report is made under sub-section (1) above, the Secretary of State may, if he thinks it desirable to do so, direct a coroner ... to hold an inquest into the death."

36. Whilst it is clear that the coroner's original decision cannot be impugned, it seems to me that on the information subsequently brought to his attention he should have concluded that an inquest ought after all to be held. Mr Burnett has helpfully told us that if this is indeed the court's view, then the coroner will readily report the facts to the Secretary of State under the provisions of s.15 so that, whilst obviously we cannot dictate the Secretary of State's decision in the matter, an inquest appears likely to be directed.

The need for a jury

37. Assuming there is now to be an inquest, will it be held with a jury? Although the point was not touched on by the Divisional Court nor addressed at length before us, Mr Burnett, I think, acknowledges that a jury would probably have to be summoned: it would seem difficult on the material presently available to regard this death as having occurred otherwise than "in circumstances the continuance or possible recurrence of which is prejudicial to the health ... of ... [a] section of the public" within the meaning of s.8(3)(d) of the Act. The point needs no elaboration. On the authority of Thomas, however, this (perhaps somewhat surprisingly) cannot affect the question whether an inquest need be held at all - see Dillon LJ's judgment at p.629D-G.

The wider argument

38. I have not thus far addressed the wider point which lies at the heart of this appeal: were the Divisional Court right to hold as they did that, whenever a death takes place in hospital and a failure to provide "routine" treatment is a cause (even a secondary cause) of death, the death is unnatural? It is this holding which so concerns the coroner and, Mr Burnett says, other coroners too. It would result, he suggests, in a very significant increase in the number of inquests to be held. Had the Divisional Court founded their judgments simply on the possibility of a "neglect" verdict, we are told, the coroner would probably not have appealed. Despite my conclusion on the issue of neglect, therefore, I must address the wider point.
39. The first question arising, of course, is whether the Divisional Court's judgment is consistent with the Court of Appeal's judgment in Thomas, which in turn raises the question whether my judgment in that case was reconcilable with those of the majority. With regard to both these questions it is, I fear, necessary to cite from another judgment of mine, this time in the Divisional Court in R v Coroner for Western District of East Sussex ex parte Homberg, Robert and Manners (1994) 158 JP 357, 370:

"Although I myself would have been disposed to include within the proper scope of such a verdict [neglect] the death of someone seriously ill or injured who would have been saved by medical care but for wholly unreasonable delay in the arrival of the emergency services, such a view is obviously inconsistent with the majority decision of the Court of Appeal in [Thomas]. That is not to say, owever, that a lack of care verdict, whether freestanding or in terms of aggravating some other cause of death, would offend ex parte Thomas. On the contrary, Dillon LJ's judgment clearly recognises the legitimate continuance of such verdicts whenever properly founded on the facts. I would therefore accept Mr Fitzgerald's submission that ex parte Thomas must be confined to the s.8(1)(a) context in which it arose; essentially it decides no more than that a broad common sense view must be taken when deciding the bald question whether a death is unnatural so as to determine whether to hold an inquest. Whereas, however, for that purpose one shuts ones mind to all but the dominant cause of death, once an inquest is held, the duty to inquire into 'how the deceased came by his death' requires one then to take a broader view and investigate not merely the dominant but also (in Jervis's language) any 'acts or omissions which are directly responsible for the death.'"

40. A little later on I referred to my own judgment in Thomas as "not a dissenting judgment but clearly expressing a minority view".
41. As is plain from that passage, I was then regarding my view in Thomas as incompatible with the majority view. Revisiting the cases, I have to say that that remains my understanding. Subject only to "neglect" cases (a category which I suspect the majority in Thomas would have drawn even more narrowly than Jamieson does), Dillon LJ was, I believe, inviting the broadest view to be taken of causation so as simply to determine "the dominant cause of death" as I called it in Homberg.
42. I accordingly find it puzzling not merely that the Divisional Court in the present case found no conflict within the Thomas judgments but also that the editors of both Halsbury's Laws and Halsbury's Statutes cite Thomas as authority for the proposition that "cases may well arise in which human fault can and properly should be found to turn what would otherwise be a natural death into an unnatural one, and one into which therefore an inquest should be held", a proposition of mine in which I had thought, and still think, I was differing from the majority view.
43. How then ought this court now to proceed? The doctrine of precedent clearly suggests that the majority view in Thomas should be applied and the Divisional Court's reasoning in the present case accordingly rejected. I nevertheless question whether many today would find the majority view in Thomas (certainly in the way I understand it) entirely satisfactory. Consider, for example, the very real doubt now thrown upon the usefulness of Lord Salmon's dictum in Alphacell - a dictum central to Dillon LJ's reasoning - as to causation being simply a matter of ordinary common sense, by Lord Hoffmann's illuminating lecture "Common Sense and Causing Loss" (delivered to the Chancery Bar Association on 15 June 1999). As Lord Hoffmann explains, to get to the right answer on an issue of causation it is necessary first to identify the question and in formulating the question it is necessary to look at the rule of law which requires it to be asked. What policy underlies it? When deciding, therefore, whether or not for s.8(1)(a) purposes a death is unnatural, one should be considering why Parliament has included this category of deaths amongst those into which an inquest must be held. What is it about unnatural deaths that calls for an inquest? Is there not a powerful case for saying that an inquest should be held whenever a wholly unexpected death, albeit from natural causes, results from some culpable human failure? (Or, more strictly, whenever the coroner has reasonable grounds to suspect that such is the case.) Such deaths prompt understandable public concern and surely no small part of the coroner's function is to carry out an appropriate investigation to allay such concern.
44. Is that not indeed the approach which the editors of the respective Halsbury's series appear to derive from Thomas? As we now learn, moreover, it appears consistent too with the approach coroners up and down the country take to certain rare deaths such as those from Legionnaires Disease. In Thomas we were given to understand that inquests are held into these deaths because "it is regarded, on a broad view, as unnatural that a person should die of an extremely rare disease". Now we are told that such inquests are held because "the disease is caused by the mechanical spraying of infected water into the atmosphere. This act is unnatural and may be unlawful and the disease is seen as occurring unnaturally - unlike hypertension which occurs very naturally indeed. The holding of an inquest in such cases has nothing to do with the fact that the death may be rare".
45. "The mechanical spraying of infected water into the atmosphere" I take to refer to the effect of inadequately maintained airconditioning systems. Quite why that is said to be unnatural whereas inadequate monitoring which allows hypertension (itself, of course, "very natural") to develop into death from cerebral haemorrhage is said to be natural, I have some difficulty in understanding.
46. Given our conclusion on the narrow point - that inquests should in any event be held into cases like this because of the possibility of a "neglect" verdict, the resolution of the wider point is clearly of less significance than it would otherwise be. Take, for example, the Divisional Court's example of Miss Thomas' attack having been relatively mild and the ambulance arriving quickly "but its journey to the hospital [having been] extended because the crew stopped for 10 minutes at a public house, with the result that when she arrived at hospital her life could not be saved". I would regard such a case as falling comfortably into the "neglect" category: Miss Thomas on that scenario would already have been a dependent of an ambulance crew who then grossly failed in her care. But undoubtedly there will be cases which fall outside the category of "neglect" and yet appear to call for an inquest on the basis already indicated, namely cases involving a wholly unexpected death from natural causes which would not have occurred but for some culpable human failure, a category of cases already perhaps recognised by the editors of Halsbury's Laws and Statutes. It is the combination of their unexpectedness and the culpable human failing that allowed them to happen which to my mind makes such deaths unnatural. Deaths by natural causes though undoubtedly they are, they should plainly never have happened and in that sense are unnatural.
47. An inquest will, of course, be held only if the coroner has reasonable cause to suspect such a combination of circumstances. That does not mean that he will have to make detailed investigations into every hospital death. Mr Burnett's fears in this regard are to my mind misplaced. Nor would I expect such a view of the law to involve any substantial increase in the number of inquests now requiring to be held.
48. I need hardly add that this approach must not be allowed to circumvent the clear bar constituted by Rule 42 of the Coroners Rules 1984: the verdict must not appear to determine any question of criminal liability on the part of a named person or any question of civil liability.
49. It follows from all this that I for my part would have upheld the judgment below even had I not concluded that an inquest was in any event required here because of the possibility of a "neglect" verdict (not itself a violation of Rule 42 - see Jamieson).

The costs below

50. The final issue arising on this appeal relates to the costs below which were ordered to be paid by the Coroner. The order was made apparently without argument to the contrary and certainly without reference to the Divisional Court's judgment in R v HM Coroner for Lincoln ex parte Hay [2000] Lloyds RepMed. 264, 278. Basing himself on Hay, Mr Burnett submits that the Coroner ought not to have been required to pay Mr Touche's costs.
51. In Hay Brooke LJ reviewed a number of coroners' cases, noted Rose LJ's judgment in R v Newcastle-under-Lyme Justices ex parte Massey [1994] 1 WLR 1684 as to magistrates, and continued:

"In my judgment, that situation [with regard to magistrates] is quite different from the situation here when a coroner is carrying out his important statutory duty to conduct an inquest. In this context the relevant principle appears to be that if a coroner not only files an affidavit but also appears and contests the making of an adverse order in an inter partes adversarial mode, then he or she is at risk as to costs. If, on the other hand, the coroner, as is fitting for somebody holding judicial office, swears an affidavit to assist the court and then appears in court, more in the role of an amicus than as a contesting party, then the court is likely to follow the normal rule set out in Jervis and make no order as to costs provided that it does not express strong disapproval of his or her conduct. ... It goes without saying that the Court is greatly assisted by coroners who depose to what took place before them and then appear in Court to assist the Court in an amicus role."

52. Jervis, I may perhaps note, states (at page 348 of the 11th edition):

"If the coroner does appear at the hearing, and loses, then the court has a discretion whether to order the coroner to pay the successful applicant's costs, even though he acted reasonably. But such an order has only rarely been made; usually no order is made unless the coroner's behaviour called for strong disapproval. One additional factor against making a costs order is where the applicant is legally aided and therefore it would only be the public paying the public."

53. No order for costs was made in favour of Mrs Hay. She, be it noted, was legally aided. Mr Touche is not. But nor, let it be made clear, has the coroner's behaviour attracted the least criticism. On the contrary he has conducted himself impeccably and, if a costs order is to be justified against him, that can only be because he failed in the event to defeat the challenge.
54. I have, I confess, some difficulty with the approach in Hay. In the first place I can find no basis in earlier authority for the suggested distinction between the coroner's appearance on the one hand as "a contesting party" ("contest[ing] the making of an adverse order in an inter partes adversarial mode"), and on the other as "an amicus". Secondly, it seems to me difficult in practice to apply this distinction. How does one tell which role the coroner is playing? Both postulate that he will be resisting the challenge and arguing the relevant law. It can hardly be by reference to the force of his (or his counsel's) submissions. Amici curiae, indeed, play different roles according to the requirements for their assistance: sometimes they argue a case which otherwise would go by default, canvassing any arguments available, however unpromising; sometimes they address wider considerations or cover a particular interest not otherwise represented. What role, on the Hay approach, is this coroner playing? True, the point at issue is one of considerable importance. Yet, the point having been resolved against him at first instance, he then appealed. An amicus does not (cannot) appeal. Having appealed, as Mr Burnett recognises, the coroner is to be treated like any other appellant: if he wins, he recovers his costs; if he loses, he pays the respondent's costs. Why then should the position be different below? Indeed, had the coroner won below, he would certainly have asked for and, no doubt, been awarded his costs. Of course, as Brooke LJ observed, the Court is greatly assisted by the coroner not merely swearing an affidavit but also appearing to argue the case - particularly in a case like the present which raises a true point of law of general application. But it would seem hard on the applicant that the more important the point, the less likely he will be to recover his costs - even though the point will obviously be of greater importance to coroners as a whole than to him as an individual.
55. Naturally I recognise that if coroners who appear on the challenge and lose are regularly to be condemned in costs they may be more reluctant to be represented at the hearing, so that the court would be deprived of their assistance. That would be a pity. But it would always be open to the court to ask for an amicus and at least then the applicant's position as to costs would be fair: he would simply have to bear his own costs irrespective of the outcome. On Mr Burnett's argument, the applicant gets the worst of all possible worlds.
56. Mr Burnett makes the additional point that Mr Touche's costs at first instance were probably no greater (or very little greater) than had the coroner chosen not to be represented: the application would still have had to be made and the hearing was in any event concluded within a day. Why then should the coroner have to pay the applicant's costs simply because he chose to appear? The answer seems to me to lie in the anomaly whereby a judicial officer (assuming only that he has done nothing calling for specific disapproval) can generally, by choosing not to appear, exempt himself from any costs liability even though his decision is found unlawful. In my judgment that anomaly ought not readily to be extended.
57. There is this further consideration. S.13 of the Coroners Act 1988 expressly provides that where the High Court, on an application brought with the Attorney General's fiat, is satisfied inter alia that the coroner "refuses ... to hold an inquest which ought to be held" (s.13(1)(a)), it may inter alia "order the coroner concerned to pay such costs of and incidental to the application as to the court may appear just" (s.13(2)(b)). Indeed, this very challenge was brought before the court by way of a s.13 application as well as by judicial review. Given this express statutory discretion to award costs against the coroner whenever justice demands, a discretion unqualified by any need to find misconduct on his part (or even, I may observe, any reference to his appearance before the court), I see no sufficient reason to subject its exercise to limitations as rigorous as those suggested by Hay.
58. Coroners are, it is well known, funded as to their legal costs by the relevant local authority - here, we are told, by four London Borough Councils. There is no question of the coroner personally having to pay the applicant's costs. Were it otherwise, indeed, he would hardly be appealing.
59. In the result, given that Parliament has chosen not to heed repeated pleas by the court that there be power in this sort of case to order costs out of public funds, I would make the same order as to costs as the Divisional Court made below per incuriam the decision in Hay, and would accordingly dismiss the appeal on this issue too.

LORD JUSTICE ROBERT WALKER:

60. I agree and I add a few words of my own on what My Lord, Simon Brown LJ has called the wider argument. I agree with Simon Brown LJ that his judgment in Thomas, although concurring in the result, is in its reasoning significantly different from that of the other members of the court. I also respectfully agree that the majority view is not entirely satisfactory.
61. The expression "unnatural death" in s.8(1)(a) of the Coroners Act 1988 does not have a single clearly defined meaning. (As Lord Sumner said in a different context in Weld-Blundell v Stephens [1920] AC 956, 983, "Everything that happens, happens in the order of nature and is therefore 'natural'".) Often "unnatural" means little more than abnormal and unexpected, and that rather muted shade of meaning would appear to be consistent with the legislative purpose of the Coroners Act 1988.
62. In particular, I doubt whether the naturalness or unnaturalness of a death should be determined exclusively in terms of causation, especially if that is seen as requiring a search for a single "dominant cause of death". That is the expression which Simon Brown LJ used in Homberg (at p.370) in summarising the majority view in Thomas. The better way forward is to look for a combination of circumstances rather than a single dominant cause.

LORD JUSTICE KEENE:

63. I agree that, for the reasons given by Simon Brown LJ, there is reasonable cause to suspect that Mrs Touche's death was contributed to by "neglect" in the sense used in Jamieson [1995] QB 1 and was for that reason alone "unnatural" in terms of section 8(1)(a) of the Coroners Act 1988. That is sufficient to determine this appeal on the substantive issue. I also agree that the appeal on costs should be dismissed for the reasons given in Simon Brown LJ's judgment.
64. On the "wider point", as it has been described, I find myself arriving at the same conclusion as Simon Brown and Robert Walker LJJ, but by a somewhat different analysis of the judgments in the leading case of Thomas. I do not discern any necessary conflict between the judgments of Dillon LJ and Farquharson LJ on the one hand and Simon Brown LJ on the other in that case. It is well-established that there may be more than one cause of death in a given situation, and that is illustrated by the possibility of a verdict which incorporates a finding that neglect contributed to the death in question. Jamieson reflects that situation, as do a large number of other authorities. Simon Brown LJ's judgment in Thomas was dealing with the obvious possibility that a death may have more than one cause.
65. But it is not the case that the judgments of the other members of the court in Thomas seek to deny that possibility. They did not deal with the possibility of there being more than one cause of death, but it cannot be that they took the view that, however complex the factual situation, it had to be forced onto the Procrustean bed of a single cause. Their judgments, properly read, amount to a finding that, on the facts of that particular case, there was only one cause. But there was no pronouncement that there has to be in all cases only a single cause of death. In short, I concur with the view expressed by the Divisional Court in the present case to the effect that they did not detect any conflict between the judgments in Thomas. They were right to say that the possibility of a coroner sometimes having to find more than one cause of death was not canvassed by Dillon LJ.
66. I therefore do not find any difficulty in terms of precedent arising in this case. The approach spelt out by Simon Brown LJ on the wider issue in his judgment in the present case is not in conflict with the decision in Thomas, and it provides a practical, workable approach in this difficult area of law. For that reason also I would dismiss the substantive appeal.
ORDER: Appeal dismissed for the reasons set out in the Judgments. The respondent to have his costs on a detailed assessment. (Order does not form part of approved Judgment)