Paul & Anor v The Royal Wolverhampton NHS Trust

This appeal raises a difficult point of law about the circumstances in which a defendant who owes a duty of care to a primary victim may be liable to a secondary victim for a psychiatric injury suffered as a result of witnessing the death or injury of the primary victim. This type of injury is referred to in the authorities, inaptly, as “nervous shock”. Several other cases have been stayed pending the outcome of this appeal.

The issue arises in proceedings brought in respect of the death of Mr Parminder Singh Paul by his wife Mrs Balbir Kaur Paul and their daughters Saffron and Mya. On 9 November 2012, Mr Paul was admitted to New Cross Hospital in Wolverhampton, part of the Defendant Trust, after complaining of chest and jaw pain. He was discharged on 12 November 2012 after various tests and investigations. More than 14 months later, on 26 January 2014, while out on a shopping trip with Saffron (then 12) and Mya (then 9), he collapsed and died from a heart attack. The circumstances are pleaded in the Particulars of Claim as follows:

“37.

On 26 January 2014 the Deceased was out shopping with his daughters, the Second and Third Claimants. He mentioned that he felt ill and both the Second and Third Claimant thought he was joking initially. The Second Claimant was walking slightly behind her father and the Third Claimant slightly in front, because they had had a minor argument and she was angry. The Third Claimant turned and initially saw her father leaning against the wall momentarily; she saw his eyes roll back. Both girls saw him fall backwards and hit his head on the floor. The Second and Third Claimant were alone with their father in the street. The Second Claimant tried to ring her mother and then an ambulance but in her distress, was unsuccessful. There was no one immediately around and she began shouting for help until eventually a lady came and used her telephone to call an ambulance. The Third Claimant managed to make contact with her mother but was too distressed to be understood. The Second Claimant took the telephone and told her mother what had happened. Both girls saw a man holding their father’s head and there was blood on his hands. The Second and Third Claimants were ushered into a nearby church by members of the public that had come to help, as the children were very distressed by what they were witnessing. The First Claimant arrived at the scene shortly thereafter. The Second and Third Claimants remember hearing their mother outside screaming their father’s name and going back outside to be with her. The Second and Third Claimant saw the ambulance crew who had arrived by this time put a foil blanket over their father. They were doing chest compressions. There was a crowd of people including the police. The Second and Third Claimants were taken to an aunt and uncle’s house.”

38.

The ambulance arrived at 15.57 and left the scene at 16.28, arriving at hospital at 16.43 but further resuscitation was felt to be futile and the deceased was declared dead at 16.51. As such, the second and third claimants witnessed the final event.”

Mr Paul’s heart attack was caused by ischaemic heart disease and occlusive coronary artery atherosclerosis. The Claimants’ case is that the failure to diagnose these conditions during Mr Paul’s stay in hospital in November 2012 was negligent. In particular, it is said that the Hospital should have performed coronary angiography on Mr Paul. This, it is said, would have revealed significant coronary artery disease, which could and would have been successfully treated by coronary revascularisation. The Claimants say that, had that occurred, Mr Paul would not have suffered a cardiac event in January 2014; and Saffron and Mya would not have suffered the psychiatric injuries which they say were caused by witnessing his collapse and death. The Claimants plead that Mr Paul’s collapse from a heart attack on 26 January 2014 was “the first manifestation of the Defendant’s breach of duty”.

The Defendant accepts that it owed a duty of care to Mrs Paul. It does not accept that it owed a duty of care to Saffron or Mya. It therefore applied to strike out their statements of case pursuant to CPR r. 3.4(2)(a) as disclosing no reasonable grounds for bringing the claims, alternatively, for summary judgment pursuant to CPR r. 24.2(b) on the basis that the claimants have no reasonable prospect of succeeding. In a detailed judgement handed down on 4 November 2019, Master Cook held that, on the facts pleaded, Saffron’s and Mya’s claims were bound to fail: [2019] EWHC 2893 (QB). He therefore struck out their claims. Permission to appeal was granted by Stewart J on 14 January 2020.

When a person suffers a tortious injury, the injury often gives rise to losses on the part of others (for example, employers and family members). In general, the law does not allow recovery by such secondary victims. An exception came to be recognised in cases where secondary victims suffered “nervous shock” having witnessed the death or injury of a primary victim. The appellate courts have, however, been careful to keep recovery for this type of injury within strict limits. The first authoritative statement of these limits from the House of Lords was in McLoughlin v O’Brian [1983] 1 AC 410. By the time of that case, the law had not countenanced recovery in respect of an injury to anyone other than a spouse or child of the claimant (though there was an exception for “rescuers”); and the claimant had to see or hear the injury directly (though there was an extension allowing recovery where the claimant came upon the “immediate aftermath” of the incident). In McLoughlin, the claimant did not directly witness the accident in which her husband and three children were involved and did not come upon the scene immediately afterwards. She did, however, attend hospital a few hours later, where she saw her husband and one of her daughters covered in mud and oil and her son seriously injured. She was told that her youngest daughter had died.

The leading speech (and the most influential in terms of the subsequent development of the law) was that of Lord Wilberforce. At 419, he observed of the law in this area that

“the courts have proceeded in the traditional manner of the common law from case to case upon a basis of logical necessity”. On the facts, he reasoned: “If one continues to follow the process of logical progression, it is hard to see why the present plaintiff also should not succeed,” given that the scene at the hospital was only incrementally less immediate than that which she would have seen if she had come upon the aftermath on the highway. As to the process by which the case fell to be decided, Lord Wilberforce said this:

“To argue from one factual situation to another and to decide by analogy is a natural tendency of the human and the legal mind. But the lawyer still has to inquire whether, in so doing, he has crossed some critical line behind which he ought to stop.”

Lord Wilberforce noted at 420 that Lord Atkin’s neighbour principle, enunciated in Donoghue v Stevenson [1932] AC 562, at 580, was that recovery should be limited to

“persons who are so closely and directly affected by my act that I ought reasonably to have them in contemplation as being so affected”. This was another way of saying that “foreseeability must be accompanied and limited by the law’s judgement as to persons who ought, according to its standards of value or justice, to have been in contemplation”. He continued:

“Foreseeability, which involves a hypothetical person, looking with hindsight at an event which has occurred, as a formula adopted by English law, not merely for defining, but also for limiting, the persons to whom duty may be owed, and the consequences for which an actor may be held responsible.”

Furthermore, “the statement that there is a ‘duty of care’ denotes a conclusion into the forming of which considerations of policy have entered”. Foreseeability did not, therefore, of itself lead to a duty of care. These considerations led to the adoption of a limiting principle that recovery for “nervous shock” must be limited to those “within sight and sound of an event caused by negligence or, at least, to those close, or very close, proximity to such a situation”: 421. At 422, Lord Wilberforce added:

“As regards proximity to the accident, it is obvious that this must be close in both time and space. It is, after all, the fact and consequence of the defendant’s negligence that must be proved to have caused the ‘nervous shock’.”

Lord Wilberforce approved the finding of liability in Benson v Lee [1972] VR 879 which he said was soundly based upon

“direct perception of some of the events which go to make up the accident as an entire event, and this includes the immediate aftermath.”

The next important decision is Alcock v Chief Constable of South Yorkshire Police [1992] AC 310, which concerned claims by those who had witnessed the deaths of their relatives in the Hillsborough Stadium disaster in 1989. Lord Keith summarised the test for liability at 396:

“in addition to reasonable foreseeability liability for injury in the particular form of psychiatric illness must depend in addition upon a requisite relationship of proximity between the claimant and the party said to owe the duty.”

As regards the class of persons to whom it may be owed, Lord Keith explained at 397 that reasonable foreseeability would be the guide. It would be necessary to prove the closeness of the tie between the claimant and the primary victim. A bystander would not ordinarily be within the range of reasonable foreseeability, but “could not perhaps be entirely excluded from it if the circumstances of a catastrophe occurring very close to him were particularly horrific”. Once the claimant was within the sphere of reasonable foreseeability, the proximity factors mentioned by Lord Wilberforce in McLoughlin fell to be taken into account. The first of these was “proximity of the plaintiff to the accident in time and space”. For this purpose, the accident was to be taken to include its immediate aftermath. At 398, Lord Keith dealt with the second proximity factor: the means by which the shock is suffered. Here, it was necessary that the shock be suffered “through sight or hearing of the event or of its immediate aftermath”. The claims of the two claimants present in the stadium during the disaster failed because they had not established the closeness of the ties with the primary victims. The claims of those not present in the stadium, but who had witnessed the disaster on television, failed because the claimants were not sufficiently proximate to the accident.

Lord Ackner set out at 401 a number of principles, the fifth of which sought to define with greater clarity the mechanism by which the psychiatric damage must be caused:

“‘Shock,’ in the context of this course of action, involves the sudden appreciation by site or sound of a horrifying event, which violently agitates the mind. It has yet to include psychiatric illness caused by the accumulation over a period of time of more gradual assaults on the nervous system.”

For Lord Ackner, there were three elements which operated as a control to limit liability even in the case where reasonable foreseeability was established: (1) the class of persons whose claims should be recognised; (2) the proximity of such persons to the accident in time and space; (3) the means by which the shock was caused: 403-4.

Lord Oliver grappled with the difficulty of establishing a principled basis for limiting liability in this area of the law. At 411, he said:

“Although it is convenient to describe the plaintiff… as a ‘secondary’ victim, that description must not be permitted to obscure the absolute essentiality of establishing a duty owed by the defendant directly to him – a duty which depends not only upon the reasonable foreseeability of damage of the type which has in fact occurred to the particular plaintiff but also upon the proximity or directness of the relationship between the plaintiff and the defendant. The difficulty lies in identifying the features which, as between two persons who may suffer effectively identical psychiatric symptoms as a result of the impression left upon them by an accident, established in the case of one who was present at or near the scene of the accident a duty in the defendant which does not exist in the case of one who was not. The answer cannot, I think, lie in the greater foreseeability of the sort of damage which the plaintiff has suffered. The traumatic effect on, for instance, a mother on the death of her child is as readily foreseeable in a case where the circumstances are described to her by an eyewitness at the inquest as it is in a case where she learns of it at a hospital immediately after the event. Nor can it be the mere suddenness or unexpectedness of the event, for the news brought by a policeman hours after the event may be a sudden and unexpected to the recipient as the occurrence of the event is to the spectator present at the scene. The answer has, as it seems to be, to be found in the existence of a combination of circumstances from which the necessary degree of proximity between the plaintiff and the defendant can be deduced. And, in the end, it has to be accepted that the concept of ‘proximity’ is an artificial one which depends more upon the court perception of what is the reasonable area for the imposition of liability than upon any logical process of analogical deduction. The common features of all the reported cases of this type decided in this country prior to the decision of Hidden J in the instant case and in which the plaintiff succeeded in establishing liability are, first, that in each case there was a marital or parental relationship between the plaintiff and the primary victim; secondly, that the injury for which damages were claimed arose from the sudden and unexpected shock to the plaintiff’s nervous system; thirdly, that the plaintiff in each case was either personally present at the scene of the accident or was in the more or less immediate vicinity and witnessed the aftermath shortly afterwards; and, fourthly, that the injury suffered arose from witnessing the death of, extreme danger to, or injury and discomfort suffered by the primary victim. Lastly, in each case there was not only an element of physical proximity to the event but a close temporal connection between the event and the plaintiff’s perception of it combined with a close relationship of affection between the plaintiff and the primary victim.”

At 416, Lord Oliver continued:

“No case prior to the hearing before Hidden J from which these appeals arise has countenanced an award of damages for injuries suffered where there was not at the time of the event a degree of physical propinquity between the plaintiff and the event caused by the defendant’s breach of duty to the primary victim nor where the shock sustained by the plaintiff was not either contemporaneous with the event or separated from it by a relatively short interval of time. The necessary element of proximity between plaintiff and defendant is furnished, at least in part, by both physical and temporal propinquity and also by the sudden and direct visual impression on the plaintiff’s mind of actually witnessing the event or its immediate aftermath.” At 417, he said this:

“To extend the notion of proximity in cases of immediately created nervous shock to this more elongated and, to some extent, retrospective process may seem a logical analogical development. But, as I shall endeavour to show, the law in this area is not wholly logical and whilst having every sympathy with the plaintiffs, whose suffering is not in doubt and is not to be underrated, I cannot for my part see any pressing reason of policy for taking this further step along the road which must ultimately lead to virtually limitless liability… further pragmatic extensions of the accepted concept of what constitutes proximity must be approached with the greatest caution. McLoughlin v O’Brian [1983] 1 AC 410 was a case which itself represented an extension not, as I think, wholly free from difficulty and any further widening of the area of potential liability to cater for the expanded and expanding range of the media of communication water, in my view, to be undertaken rather by Parliament, with full opportunity for public debate and representation, than by the process of judicial extrapolation.”

The House of Lords returned to the issue of damages for psychiatric injury in Page v Smith [1996] AC 155, in which the majority held that it was essential to distinguish between primary and secondary victims. In claims by primary victims, the only question was whether the defendant was under a duty of care to avoid causing personal injury to the claimant. If so, the type of damage – physical or psychiatric – did not matter. For secondary victims, however, the law insisted on “control mechanisms”, which were necessary to avoid the prospect of a negligent defendant being liable to all the world. “Thus,” said Lord Lloyd at 189:

“in the case of secondary victims, foreseeability of injury by shock is not enough. The law also requires a degree of proximity… this means not only proximity to the event in time and space, but also proximity of relationship between the primary victim and the secondary victim. A further control mechanism is that the secondary victim will only recover damages for nervous shock if the defendant could have foreseen injury by shock to a person of normal fortitude or ‘ordinary phlegm’.”

White v Chief Constable of South Yorkshire Police [1999] AC 455, like Alcock, concerned claims arising out of the Hillsborough disaster, this time by police officers who had been present on the day and had witnessed the events. The House of Lords held (Lord Goff dissenting) that the claims failed: The Chief Constable owed no duty of care as employer to protect his employees from pure psychiatric harm when there was no breach of the duty to protect them from physical injury; and a rescuer in whom psychiatric injury was not reasonably foreseeable could not recover for that injury. Lord Steyn concluded as follows at 500:

“My Lords, the law on the recovery of compensation for pure psychiatric harm is a patchwork quilt of distinctions which are difficult to justify. There are two theoretical solutions. The first is to wipe out recovery in tort for pure psychiatric injury… But that would be contrary to precedent and, in any event, highly controversial. Only Parliament could take such a step. The second solution is to abolish all the special limiting rules applicable to psychiatric harm… [P]recedent rules out this course and, in any event there are cogent policy considerations against such a bold innovation. In my view the only sensible strategy for the court is to say thus far and no further. The only prudent course is to treat the pragmatic categories as reflected in authoritative decisions such as the Alcock case… and Page v Smith… as settled for the time being but by and large to leave any expansion or development in this corner of the law to Parliament. In reality there are no refined analytical tools which will enable the courts to draw lines by way of compromise solution in a way which is coherent and morally defensible. It must be left to Parliament to undertake the task of radical law reform.”

In similar vein, Lord Hoffmann at 502 described the “control mechanisms” developed in the previous case law as “more or less arbitrary conditions which a plaintiff had to satisfy and which were intended to keep liability within what was regarded as acceptable bounds”. But, at 504, he ruled out any radical change to these conditions:

“It is too late to go back on the control mechanisms as stated in the Alcock case… Until there is legislative change, the courts must live with them and any judicial developments must take them into account.”

At 511, he returned to this theme:

“It seems to me that in this area of the law, the search for principle was called off in Alcock… No one can pretend that the existing law, which your Lordships have to accept, is founded upon principle. I agree with Jane

Stapleton’s remark that ‘once the law has taken a wrong turning or otherwise falling into an unsatisfactory internal state in relation to a particular course of action, incrementalism cannot provide the answer:’ see The Frontiers of Liability.”

In Taylor v Somerset Health Authority [1993] PIQR 262, the claimant’s husband suffered a heart attack at work, caused by the defendant’s negligent failure, many months before, to diagnose and treat his serious heart disease. He died shortly after being taken to hospital. The claimant went to the hospital, where she was told of the claimant’s death and identified his body. Auld J explained at 263:

“The question for the court is whether Mrs Taylor’s involvement, as I have summarised it, in and within about an hour after her husband’s death brings her within the ‘immediate aftermath’ – the fourth – principle formulated by Lord Wilberforce in McLoughlin v O’Brian…”

At 266, having set out the relevant parts of the Lord Ackner’s speech in Alcock, Auld J said this:

“All their Lordships in Alcock, in considering the application of the immediate aftermath test to the various claims before them, did so, understandably in the circumstances of the case, by reference to the accident, the disaster in the stadium. But regardless of the particular circumstances of that case, all their lordships clearly regarded some external, traumatic, event in addition to its primary consequence of injury or death as a necessary starting point when considering what Lord Oliver described, at p. 410, as ‘the essential but elusive concept of “proximity” or “directness”’.”

The argument on behalf of Mrs Taylor was that the event to which the proximity test applied was the consequence of the defendant’s negligence, namely her husband’s death from a heart attack. Thus, it was said that “the nature of the event in which the injury or death occurs is not significant”. For the defendant, two points were taken. First, there was no event to which the proximity test could be applied; and “Mr Taylor’s death long after the negligence which had caused it was the culmination of a natural process of heart disease, and the death, however unexpected and shocking to Mrs Taylor when she learned of it, was not in itself an event of the kind to which the immediate aftermath extension could be attached”. Second, it was said, even if Mr Taylor’s death at work could be considered an event of the kind to which the immediate aftermath extension could be attached, Mrs Taylor’s discovery of it at the hospital from the doctor and subsequent identification of the body did not satisfy the third of what later became known as the “control mechanisms” (which relates to the means by which the shock is caused).

Auld J held that the claim failed for both of the reasons advanced by the defendant. At 267, he said:

“The immediate aftermath extension is one which has been introduced as an exception to the general principle established in accident cases that a plaintiff can only recover damages for psychiatric injury when the accident and the primary injury or death caused by that occurred within his sight or hearing. There are two notions implicit in this exception cautiously introduced and cautiously continued by the House of Lords they are of:

(i)

an external, traumatic, event caused by the defendant’s breach of duty which immediately causes some person injury or death; and

(ii)

perception by the plaintiff of the event as it happens, normally by his presence at the scene, or exposure to the scene and/or to the primary victim so shortly afterwards that the shock of the event as well as of its consequences brought home to him.

There was no such event here other than the final consequence of Mr Taylor’s progressively deteriorating heart condition which the health authority, by its negligence many months before, had failed to arrest. In my judgment, his death at work and the subsequent transference of his body to the hospital where Mrs Taylor was informed of what had happened and where she saw the body do not constitute such an event.”

Auld J then went on to say that, even if the fatal heart attack could be considered an event to which the “immediate aftermath” extension applied, the doctor’s communication to her of that fact would not come within the extension.

In Sion v Hampstead Health Authority [1994] 5 Med LR 170, the claimant was a father of a young man injured in a motorcycle accident. The claimant did not see the accident but, for 14 days, stayed at his son’s bedside, watching him deteriorate in health, fall into a coma and die. The claim was struck out. On appeal, it was argued for the claimant that “there was no need to prove sudden shock (or an ‘external traumatic event’, to use the language of Auld J in Taylor v Somerset Health Authority…), or any shock”. Staughton

Peter Gibson LJ recorded an argument for the defendant (based in part on the decision of Auld J in Taylor v Somerset Health Authority) that “the injuries to or the death of a primary victim in themselves or itself do not qualify as the horrifying event causing the shock needed for a valid claim”, so that “it was a precondition of a claim that the incident which resulted from a breach of duty should have characteristics of suddenness and violence additional to the injuries or death of the primary victim”. Peter Gibson LJ rejected the argument:

“It is of course correct that in most of the decided cases there has been a sudden and violent incident resulting from a breach of duty, but it is the sudden awareness, violently agitating the mind, of what is occurring or has occurred that is the crucial ingredient of shock… I see no reason in logic why a breach of duty causing an incident involving no violence or suddenness, such as where the wrong medicine is negligently given to a hospital patient, could not lead to a claim for damages for nervous shock, for example where the negligence has fatal results and a visiting close relative, wholly unprepared for what has occurred, finds the body and thereby sustains a sudden and unexpected shock to the nervous system.”

But he nonetheless accepted the defendant’s primary submission that there was on the facts no “sudden and unexpected shock” and for that reason the claim was bound to fail.

Waite LJ agreed with both judgments.

In North Glamorgan NHS Trust v Walters [2002] EWCA Civ 1792, [2003] PIQR P16, the claimant’s baby was admitted to hospital on 17 June 1996 with signs of jaundice. The hospital negligently failed to diagnose that he was suffering from acute hepatitis. He needed a liver transplant. If he had received one, he would probably have lived. Instead, he was given other treatment during the week and allowed home at weekends. On the weekend of 26 July 1996, the claimant brought him back to hospital. On 30 July 1996, the claimant, who was sleeping in the same room as the baby, was awoken at about 3am to see and hear him having a fit. She was then told, wrongly, that he had not suffered any serious damage as a result of the fit. He was later transferred to another hospital, where the claimant learned that he had in fact suffered catastrophic brain damage. The baby died in her arms on the following day. Thomas J found the hospital liable to the claimant.

The lead judgment was given by Ward LJ. Clarke LJ and Sir Anthony Evans agreed. As Ward LJ said at [20], the issue was whether the claimant’s psychiatric injury arose from the “sudden appreciation by sight or sound of a horrifying event or its immediate aftermath”. He went on to analyse the case law, starting with Lord Wilberforce’s speech in McLoughlin. Citing the excerpt set out at the end of [6] above, he said at [23] of his judgment that “[o]ne looks to the totality of the circumstances which bring the claimant into proximity in both time and space to the accident”. It followed that, when Lord Wilberforce said that “[t]he shock must come through sight or hearing of the event or of its immediate aftermath”, he was not intending to confine “the event” to a frozen moment of time.

The first question in the appeal was whether the judge had erred in the holding that the

In White v Lidl UK [2005] EWHC 871 (QB), the primary victim suffered an accident in the supermarket carpark when a crash barrier which had been poorly maintained came through her windscreen. Her mental state deteriorated and, some months later, she committed suicide by hanging herself. The secondary victim was her husband, who found her hanging body and suffered psychiatric injury. The question for Hallett J was what constituted the shocking event. If it was the original accident, the claim would fail (since the secondary victim never witnessed that); if it was the suicide, the claim could succeed: see [11].

Hallett J held that the claimant was bound to fail. She said this:

“37… Even with the benefit of hindsight his injury was not reasonably foreseeable. He cannot bring himself within the category of people who suffer shock as a result of seeing or hearing a tragic event or its immediate aftermath. This is because in my judgement the relevant event for the purposes of this case is the incident with the barrier.

“38.

I agree with [counsel for the respondent] that one cannot simply ignore the incident in which injury was actually and negligently caused to the primary victim. Mrs White’s cause of action based on the respondent’s negligence arose at that time. Had Mr White come across the accident and his wife’s car and suffered shock as a result, he would no doubt have had a claim. But he did not. It took a second event six months later for that to happen. This was not a combination of circumstances making up one event or even one series of events of a seamless nature. Nor was there an inexorable progression. The shocking event in this case was a completely distinct event, a second event separated in time and space from the accident.

…

40.

It follows from what I have said that I do not accept that the decision in Walters extends the law in the way that [counsel for the Claimant] might wish. In Walters it is clear from the judgement that the event with which the court was concerned was the fit suffered by the baby and its aftermath. The defendant’s negligence caused the fit, which caused the brain damage, which in turn led to the death. Thus the event or series of events began with the fit, “the negligent infliction of damage” and continued “through to the conclusion of its immediate aftermath”. That is why there was, as the court found an inexorable progression from fit to death, which occurred in the mother’s sight and hearing. It was in that context that the court found the claimant succeeded in bringing herself within the class of people with a legitimate claim in law.

41.

As I have indicated, I am not persuaded that Mr White can do the same.

He could not have sued for his own personal injury had his wife’s health simply deteriorated after the accident. A distraught parent cannot sue for the progressive assaults upon him or her caused by the despair of looking after a brain-damaged child. Sadly, not all those who suffer can be compensated in damages. It is difficult to see therefore why the law should be extended to cover Mr White’s reaction to his wife’s death simply because six months after the accident it was for him an undoubtedly shocking event.”

In Taylor v A. Novo (UK) Ltd [2014] QB 150, the claimant’s mother sustained an injury to her head and foot when a fellow employee caused a stack of racking boards to fall on her. She seemed to be making a good recovery but, three weeks later, she collapsed and died at home as a result of deep vein thrombosis leading to pulmonary embolism, caused by the accident. Her daughter did not witness the original accident, but did witness her collapse and death at home and suffered post-traumatic stress disorder as a result. Again, the question for the court was whether the claimant could rely on the death at home as the relevant “event” for the purposes of establishing proximity.

Lord Dyson MR gave the leading judgment, with which Moore-Bick and Kitchin LJJ agreed. At [11], he cited the remarks of Auld J in Taylor v Somerset, set out at [16] above. At [12], he set out Peter Gibson LJ’s observations in Sion, set out at [17] above. At [29] he noted that the defendant’s negligence had two consequences separated by three weeks in time. There had been a “single accident or event” – the falling of the racking boards on to the claimant’s mother. If the daughter had witnessed the accident and suffered psychiatric illness as a result of that, she could have recovered. She could not, however, recover for injury caused by the death some three weeks later for “two inter-related reasons”.

The first reason (see [30]) was that

“if the judge is right, Miss Taylor would have been able to recover damages for psychiatric illness even if her mother’s death occurred months, and possibly years, after the accident (subject, of course, to proving causation). This suggests that the concept of proximity to a secondary victim cannot reasonably be stretched this far. Let us now consider the situation that would have arisen if Mrs Taylor died at the time of the accident and Miss Taylor did not witness the death, but she suffered a shock when she came on the scene shortly after the ‘immediate aftermath’. In that event Miss Taylor would not have been able to recover damages for psychiatric illness because she (possibly only just) would have failed to satisfy the physical proximity control mechanism. The idea that Mr Taylor could recover in the first situation but not in the others would strike the ordinary reasonable person as unreasonable and indeed incomprehensible. In this area of the law, the perception of the ordinary reasonable person matters.” 30The second reason (see [31]) was that:

“In my view, the effect of the judge’s approach is potentially to extend the scope of liability to secondary victims considerably further than has been done hitherto. The courts have been astute for the policy reasons articulated by Lord Steyn to confine the right of action of secondary victims by means of strict control mechanisms. In my view, the same policy reasons militate against any further substantial extension. That should only be done by Parliament.”

Lord Dyson continued:

“32.

It follows that, in my view, the judge was wrong to hold that the death of Mrs Taylor was the relevant “event” for the purposes of deciding the proximity question. A paradigm example of the kind of case in which a claimant can recover damages as a secondary victim is one involving an accident which (i) more or less immediately causes injury or death to a primary victim and (ii) is witnessed by the claimant. In such a case, the relevant event is the accident. It is not a later consequence of the accident. Auld J put the point well inTaylor v Somerset Health Authority[1993] PIQR P262: see para 11 above. Ms Taylor would have been able to recover damages as a secondary victim if she had suffered shock and psychiatric illness as a result of seeing her mother's accident. She cannot recover damages for the shock and illness that she suffered as a result of seeing her mother’s death three weeks after the accident.

33.

I turn to the authorities relied on by Mr Bartley Jones. It follows from what I have said that in my view the reasoning of Auld J in the Taylor case was correct. As I have explained at para 13 above, the observations of Peter Gibson LJ inSion v Hampstead Health Authority[1994] 5 Med LR 170were obiter dicta and they are therefore not binding on this court.”

At [35], Lord Dyson distinguished Walters in this way:

“The court was able on the facts of that case to hold that the event was a ‘seamless tale with an obvious beginning and an equally obvious end… played out over a period of 36 hours’. It was ‘one drawn-out experience’. I do not see how this sheds any light on the question that arises in this case where the injuries and death suffered by Mrs Taylor were certainly not part of a single event or seamless tale. The judge held (correctly) that the sustaining of the injuries and the death were distinct events. The question whether the death, being a separate event, was a relevant event for the purposes of a claim by a secondary victim did not arise in the Walters case.”

Shorter v Surrey and Sussex Healthcare NHS Trust [2015] EWHC 614 (QB) was another case of clinical negligence. The hospital had failed to diagnose an aneurysm as a result of which she suffered a subarachnoid haemorrhage about a week later. Her sister was with her in hospital from the point when she was told about the haemorrhage to her death. The claim failed because there was no shocking event equivalent to the seizure in Walters. Materially for present purposes, however, Swift J said this about Taylor v Somerset and Walters:

“209.

Cases of clinical negligence present particularly difficult problems. The factual background of cases can be very different and often quite complex. The nature and timing of the ‘event’ to which the breach of duty gives rise will vary from case to case. In Taylor v Somerset Heath Authority, the claimant’s husband’s heart attack and death occurred as a consequence of negligent treatment which had occurred many months before. The claimant did not observe the occurrence of the heart attack or death. She came onto the scene an hour later and viewed her husband’s body at the mortuary. The trial judge found that there was no ‘qualifying event’, just the final consequence of her husband’s progressively deteriorating heart condition which the defendant, by its negligence many months before, had failed properly to treat. It was not the kind of external, traumatic event which, when perceived by a secondary victim, would give rise to a successful claim for damages. He further found that, even if the heart attack and death were to be treated as a qualifying ‘event’, the claimant did not see her husband’s body soon enough after his death to convey to her the shock of the heart attack as well as its consequence.

210.

In the case of Walters, it is not clear how long prior to the baby’s seizure the negligence had taken place. It is, I suppose, arguable that the negligence continued from the point when the wrong diagnosis was made right up to the time of the seizure. However, in that case, the Court of Appeal made clear

(paragraph 34 of Ward LJ’s judgment) that the ‘event’ was a convenient description for ‘the fact and consequence of the defendant’s negligence’ and that it had begun ‘with the negligent infliction of damage’, i.e. at the time of the baby’s convulsion. That was the time when the consequence of the negligence first became evident. There would of course have been ongoing consequences affecting the baby’s biological processes for some time previously but it was only at the time of the convulsion that those consequences became evident and impacted on the claimant. The Court of

Appeal found that the ‘event’ began at that time and continued for the 36 hours up to the baby’s death.”

At [213], Swift J distinguished Walters in this way:

“In the case of Walters, the trial judge and the Court of Appeal laid considerable emphasis on the start of the ‘event’, when the mother awoke to find her baby rigid and choking after a convulsion, with blood pouring out of his mouth. Ward LJ likened that to the ‘assault upon her senses’ the mother would have suffered if she had seen her child bleeding in a seat after a road traffic accident. That sort of ‘assault upon the senses’ is, it seems to me, of a very different order to the scene in the A & E Department at [the hospital] on 12 May. Indeed, even if Mrs Sharma had for a short time been in the state described by the Claimant, I do not consider that the sight would have come within the type of ‘event’ described in Walters and the other relevant authorities. Mrs Sharma’s condition was fluctuating; she did not have obvious injuries; she was not – or at least did not appear at that stage to be – in any obvious or immediate danger.”

In Wild v Southend University Hospital NHS Foundation Trust [2014] EWHC 4053 (QB), [2016] PIQR P3, the claimant suffered psychiatric damage after being told of the death of his wife’s baby in utero as a result of negligent treatment by clinicians working for the defendant hospital. It was argued for the claimant that Taylor v A. Novo could not preclude a claim “in a case where the first manifestation of the injuries sustained by the primary victim occurs in front of (or within sight or hearing of) the secondary victim (or where he comes across the primary victim in the immediate aftermath of this injury) but is separated in time from the act or omission constituting negligence”. In a clinical negligence case where the first manifestation of the negligent act or omission was a shocking event seen, heard or otherwise directly experienced by the secondary victim, a claim would lie: see [35]. It was argued at [36] that it could be seen from the reference to Walters that the Court of Appeal in Taylor v A. Novo had not intended to state any new principle.

The deputy judge, Michael Kent QC, expressed “difficulty” with that proposition because of Lord Dyson’s approval of Auld J’s observation in Taylor v Somerset and his observation that Peter Gibson LJ’s remarks in Walters were obiter: see [37]-[38]. At [39], he noted that the term “external event” was “explained by the context of these claims which is that they are all made by those who are not directly participating in the events which have been engulfed the primary victims and which are in that sense external to the claimant”. At [42], he said that it was “arguably going too far” to argue, as the defendant had, that Lord Wilberforce’s reference to the “fact and consequence of the negligence” meant that the negligence must itself be synchronous with the sustaining of shock by the secondary victim. At [43], he thought it “a little unlikely” that the Court of Appeal in Walters had overlooked the fact that the fit was a result of the earlier negligent treatment, so that the cause of action had already accrued prior to the start of the relevant “event”. Ultimately, however, it was not necessary to resolve any of these points, because the claimant learned of the death after it had happened and witnessed no shocking event. This was fatal to the claim: [44]-[53].

In Liverpool Women’s Hospital NHS Foundation Trust v Ronayne [2015] EWCA Civ 588, [2015] PIQR P20, the claimant sustained a psychiatric injury from the shock of seeing his seriously ill wife in hospital, suffering from septicaemia and peritonitis, as a result of a negligently performed hysterectomy in which her colon had been stitched. The claim succeeded at first instance. On appeal, Tomlinson LJ (with whom Beatson and Sullivan LJJ agreed) said this at [17]:

“I consider it telling that there is, so far as the experienced counsel who appeared before us were aware, only one reported case in which a claimant has succeeded at trial in a claim of this type in consequence of observing in a hospital setting the consequences of clinical negligence. That is in my view unsurprising. In hospital one must expect to see patients connected to machines and drips, and as [counsel for the defendant] put it, expect to see things that one may not like to see. A visitor to a hospital is necessarily to a certain degree conditioned as to what to expect, and in the ordinary way it is also likely that due warning will be given by medical staff of an impending encounter likely to prove more than ordinarily distressing.”

Walters, the facts of which were set out in detail at [18]-[25], was the sole exception.

At [29], Tomlinson LJ set out the relevant passage from the findings of the trial judge dealing with the identification of the “event”. Applying Walters, the trial judge held that this was not the negligent stitch, but the injury flowing from it “as and when it became actionable harm, and that was when it became apparent on the 17^th July 2008 that the Claimant’s wife had or may have suffered the development of a life-threatening infection”. At [35]-[40], Tomlinson LJ held that the judge had been wrong to treat what the claimant had witnessed over a period of about 36 hours as one event. At [41], he found that the judge had also been wrong to regard it as horrifying. There does not seem to have been any challenge to the judge’s identification of the “event” as the point where the negligence became apparent; and Tomlinson LJ made no comment about it.

Various other cases were cited. I can deal with them more briefly. In Morgan v Somerset Partnership NHS Foundation Trust, on 29 February 2016, HHJ Denyer QC, sitting in the

In Tanner v Sarkar, the claimant had witnessed the death of her 2 year-old brother as a result of a GP’s negligent failure to refer him for treatment. Her claim, in the Great Grimsby County Court, failed because she had failed to establish a recognised psychiatric injury as a result of witnessing the death. In a judgment given on 12 December 2016, HHJ Buckingham added that the claim would have failed anyway because:

“the only event capable of constituting the relevant event in law is the tortious omission by [the GP]… to refer [the claimant’s brother] for urgent treatment.

That omission caused immediate harm to [the claimant’s brother] by allowing his condition to advance and further deteriorate and treated. That consultation was not of itself a shocking and horrifying event it was not therefore capable of causing psychiatric injury in [the claimant].”

Finally, in Werb v Solent NHS Trust, a psychiatric hospital allowed a 15-year old boy to go home on leave. On the following day, he committed suicide by throwing himself in front of a train. The claim was for psychiatric injury suffered by his father when he discovered the suicide. On 15 March 2017, Master Roberts declined to strike the claim out, holding that there was no evidence that the deceased had suffered any injury until, by reason of the defendant’s negligence, he took his own life.

The Master started by recalling at [32] Lord Dyson’s observation in Taylor v A. Novo, [31], that this area of the law is “arbitrary and unsatisfactory” and that this was a reason for approaching with caution any attempt to extend the scope of liability to secondary victims. Any such extension should be left to Parliament.

At [34], the Master indicated that he did not accept that clinical negligence claims involve the application of principles different from those applicable in other claims. At [35], he identified the question as “whether the death of Mr Paul is capable of being the relevant ‘event’ for deciding the proximity question”. In the Master’s view, a trial of the facts was not required to answer this question. This was because the facts of the current case could not be sensibly distinguished from those of Taylor v Somerset Health Authority. The ratio of that case was that the claimant’s husband’s death from a heart attack could not amount to a relevant event for the purposes of the proximity test: [36]. This decision had been expressly approved by the Court of Appeal in Taylor v A. Novo: [37]. In Walters, “the events from the misdiagnosis… could be seen as one event connected in space and time”: [39]. It was wrong to focus on the death of Mr Paul is being the first point at which the consequence of the defendant’s negligence became apparent. This overlooked the need for a “proximate connection between the initial negligence and the shocking event”: [40].

The Master concluded as follows at [41]:

“In the circumstances the Second and Third Claimants’ secondary victim claims are, in my judgment, bound to fail. Mr Paul’s tragic death 14½ months after the negligent incident, in circumstances separated in space and time from the negligence I must assume occurred in the hospital, cannot possibly be said to be the ‘relevant event’ for deciding the proximity required to establish liability under the established control mechanisms. It follows that the Defendant’s application must succeed and the secondary victim claims of the Second and Third Defendants will be struck out. They will of course retain their loss of dependency claims under the Fatal Accidents Act 1976.”

For the Appellants, Laura Johnson began by emphasising that, on an application to strike out, it must be assumed that the claimants will establish their pleaded case in its entirety. On that case, Mr Paul’s collapse on 26 January 2014 was the point at which he first suffered injury as a result of the defendant’s breach of duty, or the point at which that injury first became apparent. Insofar as the defendant argued that the negligent omission must have caused, or failed to arrest, a biological process leading immediately to actionable damage to Mr Paul’s heart or circulatory system, this would depend on expert evidence and so was not suitable for resolution at this stage.

Ms Johnson submitted that the Master had erred in concluding that it was necessary to establish temporal proximity between the negligence and the shocking event. There is no such requirement. McLoughlin and Alcock were conventional accident claims, where there was no gap between the negligence and the damage. Nonetheless, in neither case was it said that there must be temporal proximity between the negligence and the shocking event. The proximity required is, rather, between the claimant and the shocking event. Thus, it would be perfectly possible to imagine liability for nervous shock in a case where a claimant witnessed the death or injury of a close family member caused by (a) a collapse of scaffolding caused by its negligent erection some months earlier; or (b) the collapse of a water tank in a theatre, caused by its negligent installation six months earlier. It would be a surprising contention if a person who witnessed such events could not recover as a secondary victim because he or she had not witnessed the negligent act or omission, only the shocking event caused by it.

Ms Johnson relied on Walters to submit that clinical negligence claims are no different in principle. In that case, there was no suggestion that the negligence (the failure to diagnose on 17 June 1996) was temporally proximate to the shocking event (which began with the baby’s seizure on 30 July 1996). Thus, Walters is authority for the proposition that the shocking event can be an event caused by the defendant’s negligence rather than the defendant’s negligent act or omission itself.

As to Taylor v Somerset Health Authority, Ms Johnson submitted that nothing was known of the circumstances of Mr Taylor’s heart attack, because Mrs Taylor had not witnessed it. Auld J’s judgment cannot be read as saying that a death from heart attack can never satisfy the Alcock criteria unless the negligent act or omission is coincident with the shocking event. Even if Auld J did mean that, Walters shows that he was wrong.

Taylor v A. Novo could not be taken as having endorsed that view. It was a “two event” case; and by distinguishing Walters, the Court of Appeal was accepting that there might be a different result in a case where clinical negligence gave rise to a single event separated in time from the negligent treatment.

Finally, Ms Johnson submitted, relying on W v Essex County Council [2001] 2 AC 592, 598A-F (Lord Slynn), that the Master was wrong to strike out these claims, given the developing and uncertain nature of the law in this area.

For the Respondent, Mr Bagot QC submitted that the law was settled and, applying that law to the pleaded facts, the Master was correct to conclude that the claims were bound to fail.

Adopting the language of Lord Oliver in Alcock, Mr Bagot submitted that the Appellants’ focus on identifying the “event” must not be allowed to obscure “the absolute essentiality of establishing a duty owed by the defendant directly to [the claimant] – a duty which depends not only upon the reasonable foreseeability of damage of the type which has in

The Master was correct to focus on two cases, which together provide a complete answer to the Appellants’ claims: Taylor v Somerset Health Authority and Taylor v A. Novo. Mr Bagot submitted that the Respondent had never argued, and the Master never said, that a gap in time between the breach of duty and the consequent shocking event precluded a secondary victim claim. The argument was simply that, as noted by Auld J in Taylor v

Mr Bagot submitted that, on a proper analysis of the case law, Mr Paul’s heart attack and death on 26 January 2014 were consequences of the Respondent’s negligence. They were not capable of being themselves an “event”. The Master was right to look for a “a proximate connection” between the negligence and the shocking event: this flowed from Lord Wilberforce’s observation in McLoughlin that it was necessary to show that the injury had been caused by the “fact and consequence of the defendant’s negligence”.

In this case, there was on the Appellants’ case a completed tort as soon as, in breach of duty, the Respondent negligently failed to arrange coronary angiography and instead decided to discharge Mr Paul on 12 November 2012. Some damage to Mr Paul was occasioned to him on that day by the failure to improve or stabilise his condition or, at the very least, to prevent it from worsening. That being so, the heart attack was the later consequence of the hospital’s negligent infliction of damage and could not itself qualify as an “event”. In this respect the analysis of HHJ Denyer QC in Morgan v Somerset and of HHJ Buckingham in Tanner v Sarkar was correct. The proposition that a hidden and symptomless but non-negligible physical change was actionable was now established beyond doubt by the Supreme Court’s decision in Dryden v Johnson Matthey plc [2019] AC 403.

Mr Bagot submitted that Taylor v A. Novo was fatal to the claim. It established that “the relevant event is the accident. It is not a later consequence of the accident.” If it were otherwise, this would be – as Hallett J put it in White v Lidl – “tantamount to saying that provided the claimant can establish a causal link between their injury and the defendant’s negligence then they can have a claim”. Walters had been correctly analysed in White v Lidl and Wild as a case where the “event” began with the seizure, which was treated as being the “negligent infliction of damage”. Here, by contrast, the negligent infliction of damage happened 14½ months before the heart attack.

The Master was correct to conclude that the facts of this case could not be sensibly distinguished from those of Taylor v Somerset. The claim in that case failed for two separate reasons. The first was that the heart attack, many months after the negligent treatment, was not capable of being an “event”. This was the reason that Lord Dyson endorsed in Taylor v A. Novo.

A survey of the authorities indicates a degree of frustration about the lack of coherent principle underlying the law governing claims for psychiatric damage suffered by secondary victims. Lord Steyn, in White v Chief Constable of South Yorkshire Police, described the law in this area as “a patchwork quilt of distinctions which are difficult to justify”. Lord Hoffmann in the same case observed that the “search for principle” had been “called off” in Alcock. What they seem to have meant was that there are only two coherent positions for the law to take: allow secondary victim claims for psychiatric damage generally (subject only to foreseeability) or disallow them generally (as Prof. Jane Stapleton had argued). But coherence is not the only desideratum. Another is the need to ensure that developments in the common law should be incremental, because radical developments, particularly when they have polycentric public policy implications, are in our constitutional system for Parliament. In this area of the law, the House of Lords has decided that a move to either of the two logically coherent positions would be too radical a step for the courts to take. The solution adopted in Alcock was to say “thus far and no further” – i.e. to freeze the law in something like its (then) current state. The method by which this was done was to insist on what later came to be known as the “control mechanisms”, even though doing so perpetuated distinctions which it was impossible to justify in principle.

It is important to understand, but also not to overstate, the consequence of this approach. The consequence is that, at the level of the “control mechanisms” themselves (what might be termed the “macro” level), certain kinds of analogical reasoning which might otherwise have been open to courts are not now open. The courts cannot reason: “If the law allows recovery for post-traumatic stress disorder caused by witnessing a sudden death, and science tells us that the occurrence of the same disorder is equally foreseeable in one who witnesses a drawn-out death, principle dictates liability in the latter case.” Such reasoning is closed off, not because the argument is a bad one, but because constitutional considerations dictate that this step is for the legislature to take. The courts, for their part, are limited by the existing control mechanisms. Once the basis for their adoption is recognised, any appeal to extend them by analogical reasoning is deprived of its force.

This does not, however, tell us anything about how to decide (at what might be called the “micro” level) whether, on a given set of facts, a claimant has satisfied the existing “control mechanisms”. Here, there is no constitutional reason why the courts should not apply their usual analogical tools. More specifically, there is no reason to favour a conservative posture in which liability is accepted only where it has already been found to exist on indistinguishable facts. There is nothing to inhibit the courts from aiming for maximal coherence in the principles which govern the circumstances in which the existing control mechanisms will be satisfied. In doing so, they are bound by the rules of precedent, but are otherwise unconstrained.

Mr Bagot emphasised the importance of the overarching requirement to show proximity between claimant and defendant. In my judgment, however, the appeal to proximity, in this sense, does not assist. As Lord Dyson made clear in Taylor v A. Novo at [26]-[28], the term “proximity” has been used in two different senses in the case law: first, “a legal term… used as a shorthand for Lord Atkin’s famous neighbour principle” which “describes the relationship between parties which is necessary in order to found a duty of care owed by one to the other”; second, in the specific context of secondary victim cases, “physical proximity in time and space to an event”, i.e. “one of the control mechanisms which, as a matter of policy, the law has introduced”. When Lord Oliver spoke in the passage from his speech in Alcock set out at [9] above of the “absolute essentiality” of showing proximity between claimant and defendant, he was using the word in the first sense. He made this clear later on in the same passage, when he observed that “the concept of ‘proximity’ is an artificial one which depends more upon the court’s perception of what is the reasonable area for the imposition of liability than upon any logical process of analogical deduction”. In the context of secondary victim claims, the control mechanisms are, as Lord Dyson put it in Taylor v A. Novo, “the judicial response to how this area should be defined”.

Contrary to Mr Bagot’s submission, therefore, it is not sufficient simply to pose the question “Were the parties in a sufficiently proximate relationship?” In the first place, the answer to that question is far from obvious. More importantly, however, it is to be answered not (as Mr Bagot suggested) by an appeal to basic intuition, but by asking whether the control mechanisms are satisfied. In this case, there is no dispute that, if Mr Paul’s collapse from a heart attack on 26 January 2014 was capable of being a relevant “event”, each of the “control mechanisms” is satisfied on the facts pleaded: there was a parental relationship between the Claimants and the primary victim; the injury for which damages were claimed arose from a sudden and unexpected shock to the Claimants’ nervous systems; the Claimants were personally present at the scene; the injury suffered arose from witnessing the death of the primary victim; and there was a close temporal connection between the event and the Claimants’ perception of it, combined with a close relationship of affection between the Claimants and the primary victim.

It follows from the foregoing analysis that the key question in the present case is whether Mr Paul’s collapse from a heart attack, 14½ months after the allegedly negligent treatment, is capable of constituting a relevant “event”. There are three possible reasons why not. The first is that the “event” has to be synchronous, or approximately synchronous, with the negligence which gives rise to it. Mr Bagot was at pains to emphasise that he had not advanced this argument before the Master and was not advancing it before me. In my judgment, he was right to eschew it; and the reason why is relevant to the arguments he did advance. Although McLoughlin and Alcock were both cases where the negligence was close in time to the “event”, there is nothing in any of

A second possible reason why it might be said that Mr Paul’s collapse from a heart attack could not be a relevant “event” is that liability depends on there being a negligent act, rather than an omission. Again, Mr Bagot did not advance this argument. Again, he was right not to do so. Even if it were possible reliably to distinguish between cases of negligent acts and negligent omissions, there is no reason of principle why an event caused by the former should be treated any differently from an event caused by the latter. In any event, Walters was a case of negligent omission; that case would have been wrongly decided if liability required a negligent act. White v Lidl was another case of negligent omission; and although the claim in that case failed, Hallett J made clear at [38] that it would have succeeded had the secondary victim witnessed the “accident” and suffered psychiatric injury as a consequence.

A third possible reason why Mr Paul’s collapse could not be an “event” is that the Claimants were absent from the “scene of the tort”. This was very much part of Mr Bagot’s submissions. In my judgment, however, it takes matters no further. In “accident” cases, like McLoughlin, Alcock and Taylor v A. Novo, where the breach of the duty and the damage caused are coincident in time and place, the “scene of the tort” is also the scene of the negligence. When the negligence and the damage are separated, and assuming that there is no requirement for the negligence and the damage to be synchronous, the “scene of the tort” can only mean mean “the scene where damage first occurred”. In the context of the tort of negligence, this is the point when the tort becomes actionable or complete.

It was Mr Bagot’s case that the tort here became actionable at the time of or immediately after Mr Paul’s admission to hospital in November 2012, because the negligent failure to diagnose Mr Paul’s condition would at that point have set in train, or failed to arrest, the biological processes that eventually led to his death. Mr Bagot submitted that Dryden v Johnson Matthey showed this to be so even if, as pleaded in the Particulars of Claim, Mr Paul’s collapse was the “first manifestation of the Defendant’s breach of duty”. The difficulty with this submission is that Dryden v Johnson Matthey shows only that nonmanifest biological changes can constitute actionable damage. The question whether, in this case, the Defendant’s negligence caused actionable damage prior to Mr Paul’s collapse is a question of fact, which would no doubt have to be determined on the basis of expert evidence. At this stage, Ms Johnson’s position is that there is no indication of any damage prior to the moment of Mr Paul’s heart attack. For the purposes of this strike out application, I have to proceed on the factual basis most favourable to the Claimants,

The argument does not end there, because Mr Bagot submitted that, even on this basis, the claim could not succeed; and Ms Johnson submitted that the claim could succeed even if it were to turn out that there had been damage prior to Mr Paul’s heart attack on 26 January 2014. It is necessary to consider these positions in turn.

The essence of Mr Bagot’s submission was that a heart attack caused by clinical negligence many months before was not the kind of “event” that was capable of giving rise to liability for psychiatric injury in secondary victims. One strand of the argument was that it was not foreseeable that psychiatric injury would result from witnessing the consequence, many months later, of a negligent omission in the clinical setting. I would reject that argument. As Lord Wilberforce said in the passage from his speech McLoughlin set out at [6] above, foreseeability in this context involves “a hypothetical person, looking with hindsight at an event which has occurred”. There is no dispute that it is foreseeable that a negligent failure to diagnose a heart condition could result in a heart attack. Heart attacks can happen in many different ways. Not all of them are sudden and shocking events, but some are. I would regard it as eminently foreseeable that a negligent failure to diagnose and treat a heart condition might result in a sudden and shocking event that, if witnessed by close family members, might occasion psychiatric damage. It is true, as Mr Bagot submitted, that in Lord Wilberforce’s formulation it is the

“fact and consequence” (emphasis added) of the defendant’s negligence that must be shown to have caused the psychiatric damage in the secondary victim. But the need to consider whether the control mechanisms are satisfied arises only once causation is established. Put another way, the control mechanisms limit liability for psychiatric injury which is proved to have been caused by the defendant’s negligence. That being so, it is difficult to see how the need to establish causation could be relevant to identifying the kind of “event” that can give rise to liability for damage.

It seemed to be a further strand of Mr Bagot’s argument that the need to show that shock was caused by the “fact and consequence” of the defendant’s negligence entailed a requirement that the secondary victim must, at the time of the “event”, perceive not only the injury to the primary victim but also the fact of its causation by the defendant’s negligence. I can see no justification for any such requirement and no support for it in the authorities. Close family members who witnessed the Hillsborough disaster directly, and were able to show the necessary ties of love and affection, were in principle entitled to claim for psychiatric injuries caused by witnessing the deaths of their loved ones. Many of these secondary victims would have had no idea at the time whether the disaster was the result of negligence and, if so, whose negligence. There is no suggestion in the cases that this made any difference and no reason in principle why it should. A claimant must establish the chain of causation from the negligent act to a shocking event causing injury to the primary victim and then onwards to the psychiatric injury in the secondary victim; there is no additional requirement that this chain of causation, or any part of it, should have been manifest to the secondary victim at the time of the event.

The centrepiece of Mr Bagot’s argument was the decision of Auld J in Taylor v Somerset Health Authority and the endorsement of its reasoning by the Court of Appeal in Taylor v A. Novo. The claim in that case failed for two reasons. It is the first that is important for present purposes. The heart attack there (occurring at work and not witnessed by the claimant) was not an event that could give rise to liability because it was not “an external, traumatic, event caused by the defendant’s breach of duty which immediately causes some person injury or death”: see the extract cited at [16] above. One way of reading this is that the “event” has to be external to the primary victim. It is possible to glean some support for that reading from Auld J’s earlier statement that Alcock required “some external, traumatic, event in addition to its primary consequence of injury or death” (emphasis added): see the extract at [14] above. But if that is the proposition Auld J was advancing, it would be impossible for a claimant to recover damages for psychiatric injury caused by negligent omission in a clinical setting, because the “event” – however sudden and shocking – would always be internal, rather than external, to the primary victim.

On that analysis, the seizure in Walters would also have been internal to the primary victim; and that case would have been wrongly decided. Lord Dyson in Taylor v A. Novo did not say it was wrongly decided, although he distinguished it. If it were necessary for the event to be external to the primary victim, that would have provided an easy answer to the claim in Shorter. Yet there is no trace of that line of reasoning in the judgment of Swift J (which was informed by the Court of Appeal’s decision in Taylor v A. Novo). The claim in that case failed not because injuries caused by negligent omission in a clinical setting are incapable of founding liability for psychiatric injury in a secondary victim, but because, unlike in Walters, the events in that case did not qualify as a single, sudden, horrifying event. The same is true of the injuries in Ronayne. Again, Tomlinson LJ distinguished Walters because there was a series of events over a period of time, not because an event internal to the primary victim could never qualify. His observation that it will be rare for such an event to qualify if it takes place in a hospital setting does not, of course, cover this case and in any event serves to underline that allowing recovery in cases of clinical negligence is unlikely to lead to an unmanageable flood of cases.

There is another way of understanding Auld J’s requirement in Taylor v Somerset Health Authority for “an external, traumatic, event caused by the defendant’s breach of duty which immediately causes some person injury or death”. This is to read “external” as

“external to the secondary victim”. I note that this is how Michael Kent QC read the requirement in Wild: see [36] above. To my mind, it is also a plausible reading in the context of Taylor v Somerset Health Authority. The argument being advanced by the claimant there was that “the nature of the event in which the injury or death occurs is not significant” as long as it produces a sudden shock in the secondary victim. If “external” means “external to the secondary victim”, Auld J’s requirement for an “external, traumatic event” can properly be read as a rejection of this argument.

In my judgment, the ratio of Taylor v A. Novo is that, in a case where the defendant’s negligence results in an “event” giving rise to injury in a primary victim, a secondary victim can claim for psychiatric injury only where it is caused by witnessing that event rather than any subsequent, discrete event which is the consequence of it, however sudden or shocking that subsequent event may be. It is true that, at [30] of his judgment (see [29] above), Lord Dyson reasons that it would be undesirable to allow recovery in a case where “death had occurred months, and possibly years, after the accident”. But this is a concern about delay between “the accident” (i.e. the event) and its later consequence. As I noted at [63] above, there is nothing to suggest that there would be any reason to

At [32] of his judgment (see [31] above), Lord Dyson was careful to say that “accident” cases were “[a] paradigm example” of those in which a claimant can recover damages as a secondary victim and that “[i]n such a case” the relevant event is the accident, rather than a later consequence of it. This careful formulation seems to me to allow for nonparadigm cases where there is no “accident”, but some other kind of event – such as in Walters. The passage at [35] in which Walters is distinguished appears to recognise that an event which is external to the secondary victim, but internal to the primary victim, could in principle qualify if it is sufficiently sudden and horrifying and leads immediately or “seamlessly” to death or injury in the primary victim. This is consistent with the express endorsement at [33] of Auld J’s reasoning in Taylor v Somerset Health Authority only if the requirement for an “external, traumatic event” is read as requiring an event external to the secondary victim. That reading is supported by Lord Dyson’s implicit disapproval, in the very next sentence, of Peter Gibson LJ’s observations in Sion. (Peter Gibson LJ had said that the “crucial ingredient” of shock was a “sudden awareness, violently agitating the mind” and it did not matter if the incident giving rise to it involved no violence or suddenness at all. So, for him, there did not even have to be an event external to the secondary victim.)

On this analysis, I would hold that the Master was wrong to conclude that these claims are bound to fail on the facts pleaded. Here, unlike in Taylor v A. Novo, there was on the facts pleaded only one event: Mr Paul’s collapse from a heart attack on 26 January 2014. On the facts pleaded, it was a sudden event, external to the secondary victims, and it led immediately or very rapidly to Mr Paul’s death. The event would have been horrifying to any close family member who witnessed it, and especially so to children of 12 and 9. The fact that the event occurred 14 ½ months after the negligent omission which caused it does not, in and of itself, preclude liability. Nor does the fact that it was not an “accident” in the ordinary sense of the word, but rather an event internal to the primary victim. In a case where such an event is the first occasion on which damage is caused, and therefore the first occasion on which it can be said that the cause of action is complete, Taylor v A. Novo does not preclude liability. As I have explained, for the purposes of this strike-out application, I must assume that the present is such a case.

The analysis set out above is sufficient to demonstrate that this appeal must be allowed and the Master’s order striking out these claims set aside. However, Ms Johnson made clear in argument that, in her submission, it was possible for the claims to succeed even if the negligent failure to diagnose had given rise to actionable damage prior to Mr Paul’s collapse on 26 January 2014. Since I have heard argument on that matter, it is appropriate that I should make some brief observations about it.

Although Ms Johnson was right to point out that Taylor v A. Novo was not a clinical negligence case, there is no indication in any of the authorities that the principles applicable to such cases are different. Lord Dyson’s reliance on and endorsement of Taylor v Somerset Health Authority, and his reasons for distinguishing Walters, make clear that they are not. That being so, Taylor v A. Novo would preclude liability in the present case if there were a relevant “event” prior to Mr Paul’s collapse on 26 January

In Taylor v A. Novo, however, there was something that could properly be described as an “event” prior to that witnessed by the secondary victim – the collapse of the racking boards on to Mrs Taylor (the primary victim). That event coincided with or immediately preceded the moment when actionable damage was first suffered by her, which was also the moment when that damage became manifest. The same will be true in most “accident” cases. But in the present case, there was nothing that could naturally be described as an “event” before Mr Paul’s collapse on 26 January 2014, even on the assumption that some actionable damage was suffered before that date. The Court of Appeal in Taylor v A. Novo did not need to, and did not, decide whether a defendant could be liable to a secondary victim in such a case. It did not say, for example, that an “event” can qualify only if it coincides with or immediately precedes the first actionable damage to the primary victim.

If it is necessary to identify a stopping point after which the consequences of a negligent act or omission can no longer qualify as an “event” giving rise to liability for psychiatric damage in a secondary victim, the most obvious candidate is the point when damage to the primary victim first becomes manifest or, as Swift J put it in Shorter (see [33] above), “evident”. In Walters, this was the point when the baby suffered a seizure, even though, as Swift J noted, “[t]here would of course have been ongoing consequences affecting the baby’s biological processes for some time previously”. Had the death been a discrete event, rather than the end of a single, seamless event starting with the seizure, there could have been no liability to secondary victims for psychiatric damage caused by witnessing it. In a case where the shocking event is the point where the damage becomes evident, it is difficult to see why recovery should depend on the happenstance of whether, possibly unbeknown to the primary or secondary victims, actionable damage had previously been suffered.

I would therefore hold that the principle in Taylor v A. Novo is no bar to recovery in this case if it is shown that Mr Paul’s collapse from a heart attack on 26 January 2014 was the first occasion on which the damage caused by the hospital’s negligent failure to diagnose and treat his heart condition became manifest.

I do not accept Mr Bagot’s submission that this approach should be rejected because it will open the floodgates to claims from secondary victims in an unacceptably large number of cases. In the first place, arguments of this sort are best addressed at the “macro” level, i.e. when setting the control mechanisms, rather than applying them to the facts of individual cases. In any event, I do not accept the premise. Even though defendants are in principle liable to secondary parties for psychiatric damage caused by witnessing an event in the primary victim caused by clinical negligence, it will still be necessary to establish that the event in question was sudden, unexpected and shocking in the relevant sense. Cases like Shorter and Ronayne show that this will not be easy, particularly where the event occurs in a clinical setting. Moreover, even if there is a qualifying shocking event, it will remain necessary to show that it was that event, and not some later discrete consequence of it, that caused the psychiatric injury. These are stringent limits on recovery for psychiatric damage by secondary victims. Any additional limit would involve modification of the existing control mechanisms; and that is a matter for Parliament or, possibly, for the higher appellate courts.