Young v AIG Europe Ltd

Mr Neil Young, the Claimant (C), is now aged 77 years having been born on 4 April 1938. On 13 May 2013 he was involved in an accident caused by the negligent driving of the Defendant’s (D) insured. He suffered personal injury as a result of that accident. The question for me to determine is one of causation only.

On 15 January 2015 Master Eastman made the following order:

C suffered a myocardial infarction (heart attack) (“MI”) on/about 15 May 2013. D accepts that the accident caused the MI. C also sustained a spinal haematoma, the neurological symptoms of which first developed on/about 17 May 2013 and this rendered him paraplegic. The spinal haematoma and its consequences are also admitted as having been caused by the road traffic accident.

On 4 June 2013 C suffered a non-haemorrhagic stroke. This caused left-sided paralysis which apparently will have the effect of substantially increasing C’s claim (and particularly his care needs) if C succeeds on the disputed issue of causation.

There is expert cardiological evidence from Professor Hall for the Claimant (report 9 January 2015) and Dr Saltissi for the Defendant (report of 12 November 2014). There is a joint statement of the cardiologists dated March 2015.

The core of the issue arises from paragraph 8 of the amended defence which, in relation to the causation of the stroke says:

Embolism arising from (a carotid artery) atheromatous plaque unrelated to the spinal injury; (Footnote: ²)

In summary, D submits that the medical evidence merely establishes a material increase in risk and that this is not the applicable test of causation in the circumstances of this case; that C’s case at its highest is that D’s negligence merely added a new discrete risk factor (the MI) to the existing risk factors of age, male gender, ex-smoking habit, hypertension, hypercholesterolemia and overweight/obesity. Therefore, D submits that the court cannot find that the road traffic accident probably caused the stroke.

The joint statement of the neurologists is short. There is agreement as to the 4 potential mechanisms for the non haemorrhagic stroke, as now recorded in the amended defence. (Footnote: ⁴) The statement continues:

Further questioning of Professor Venables, subsequent to the joint statement, revealed the following: (Footnote: ⁵)

Before dealing with the cardiology evidence, a short glossary may help:

LV – left ventricular chamber of heart

Akinesia – static LV wall without contraction

Thrombus – blood clot

Mural Thrombus – blood clot which forms on the wall of the LV

Embolus – usually a piece of thrombus which has broken away and is carried towards the brain by the circulation of blood.

Of the 4 potential causes for C’s stroke, the cardiologists (and neurologists) agree that cause (d) is the most unlikely. They agree that if cause (b) was responsible for the stroke then it was caused by the accident. Professor Hall says that of the 4 causes (b) is the most likely, but he cannot say that it is more probable than not that that was the mechanism which caused the accident. He accepts that (a) and (c) are serious possibilities and, although (d) is very unlikely, it cannot be excluded.

What Professor Hall does say is that on the clear balance of probabilities the accident caused the stroke or, to put it another way, absent the accident C probably would not have suffered the stroke he did suffer. The evidence is that the stroke was likely to have been caused by a thrombotic mechanism but he cannot say which of the 4 probably did cause it, though (b) is the most likely candidate of the 4. In other words:

The stroke (it is agreed) caused the MI and the MI is the most likely of the 4 causes of the stroke, though not such that on the balance of probabilities it did cause the stroke.

If it was not the MI (cause (b)) then there were multiple factors driving the thrombotic process. These were the initial trauma, the spinal cord surgery and immobility in hospital. (Footnote: ⁶)

Professor Hall accepted that C was at some increased background risk by reason of the facts set out in paragraph 7 of this judgment. Nevertheless, for the reasons given, his view was that the accident very probably did cause the thrombus which caused the stroke by 1 of the 4 possible pathways. In addition there was the close temporal relationship.

Dr Saltissi’s overall view was that (d) was a “non starter”, (b) was unlikely and that (a) and (c) were possibilities which were possibly linked to the accident. He said he genuinely believed that there was an equal likelihood that this was a coincidental event or an event caused by the accident. I pressed him and asked him that if he had to “put his money anywhere” what would his order of likelihood be? He said (of course extremely qualified) that it would be (c), (a), (b), (d).

There was substantial cardiological evidence on the possible pathways. It is right that I summarise this evidence. I shall begin with pathway/ cause (b).

The essence of Professor Hall’s opinion for making (b) the most likely cause of the 4 is that the risk of stroke in the post MI period is greatly enhanced. For this he relies upon the article by Witt (Footnote: ⁷). On the other hand Dr Saltissi says that the temporal link between the MI and the stroke exists, but he considers that there are reasons why the link, whilst possibly causal, is more likely to be co-incidental.

Dr Saltissi says that the Witt paper is a blunt instrument and there are aetiological features in this case which undermine the reasoning that one can deduce from the epidemiological increased risk of stroke following MI that the balance of probabilities is that the MI caused the stroke. (Footnote: ⁸) Indeed he suggests that cause (c) is more likely than cause (b). He says that Witt not only has limitations which make it non applicable in this case; it also contains inconsistencies.

Dr Saltissi gives 4 essential reasons for his opinion that the balance of probabilities favours a cause other than (b), notwithstanding the Witt article. These need some analysis.

It is agreed by the cardiologists that the degree of LV damage shown on C’s echocardiograms is no more than moderate. There is no evidence of extensive damage. Dr Saltissi says that although the quality of scans does vary, an echocardiogram would almost always show a significant area of akinesia and often thrombus. This is important according to him, because where a cardiogenic embolus is caused by MI, a mural thrombus has previously formed. A principal reason for the thrombus formation is that the MI damages the heart walls such that it no longer contracts. The lack of contraction means that the blood in contact with that part of the wall is static and so the thrombus forms. The absence of mural thrombus/extensive LV akinesia demonstrated on the echocardiograms is, he contends, a strong indicator against cause (b). He says that to establish the presence of a cardiogenic embolism, at least on the balance of probabilities, it is necessary to identify either mural thrombus or extensive LV akinesia (or both) on the echocardiograms.

Professor Hall disagrees with Dr Saltissi on this. He says that mural thrombus or extensive akinesia would not necessarily be evident on an echocardiogram. He is not surprised by the fact that no thrombus is seen in the LV and says that deterioration of areas of damage to the myocardium will depend on the quality of the imaging. In addition he says that the first and third echocardiograms were described as being technically difficult, and so it would be hard to identify thrombus or accurately delineate areas of abnormal contraction. Professor Hall also says:

One would not expect to see a thrombus on the first echocardiogram because it was on the day of the MI and clots take time to form. (Footnote: ¹⁰) As regards the second and third echocardiograms, they were too late because, although thrombi may stay in situ, they also can fully embolise or dissolve naturally. Therefore the fact that there was no thrombus a number of months later does not take the matter any further.

When the echocardiogram specifically refers to the study being “technically difficult”, the fact that that is remarked upon is important to demonstrate that it may not show everything. It may not indicate the true detail of the impairment and, if a thrombus had been there, it may have been missed on the echocardiogram.

The echocardiogram report stated “The posterior wall contracts with a bouncy motion. Inferior wall appears dyskinetic”. This means that the wall moves out when it should move in and makes it a candidate for thrombus formation because the blood flow is not being normally expelled from that area.

Dr Saltissi responded that he considered that Professor Hall was over diminishing the importance of the echocardiogram. The technicians were able to determine the motion of the heart wall and were able to comment upon it. He accepted that the echocardiogram was not a perfect tool but emphasised that none of the three echocardiograms showed a thrombus. Whilst he accepted that the first echocardiogram was early, he said it is the larger clots which occur earlier and are picked up on the echocardiogram and embolise. As regards to the second and third echocardiograms, he said that some thrombi do dissolve and some fully embolise but some also organise and would be visible.

There is clearly a problem with the echocardiogram evidence. It is difficult to say that it is supportive of C’s case. I accept Dr Saltissi’s evidence that it is some indication against there having been a left ventricular clot. However it is by no means conclusive.

The cardiologists also disagree as to whether extensive akinesia is necessary so as to make cause (b) likely.

Professor Hall says extensive akinesia is not necessary. As regards to the aetiology he says that the main pre-requisite is an area of myocardial damage. This is by definition present when there is a proven MI. A non-ST elevation MI is nearly always sub endocardial (Footnote: ¹¹). In a STEMI damage goes through the heart wall, therefore there is greater myocardial damage and more likelihood of clot. Nevertheless non ST elevation MI provides a site where the clot can form prior to (potentially) becoming an embolism if it breaks off. He suggests that in the context of MI a reduction in contraction may favour the occurrence of thrombus on the wall of the LV but there is another very important factor. This is that when the inside lining of the heart becomes inflamed and damaged, it acts as an area where thrombus will form, since the clotting mechanism is activated by that type of inflammation as a result of the MI.

Dr Saltissi on the other hand says the mere presence of some myocardial damage is necessary but not sufficient. The damage has to be sufficiently focal, extensive and severe to have impaired LV mural contraction. Non – ST elevation MI affects only the sub endocardium and not the full thickness of the wall. Therefore it less commonly causes a wall motion abnormality sufficient to cause an attached mural thrombosis.

Professor Hall says that Witt (table 2) shows that there is no difference in the incidence of anterior myocardial infarction, ST elevation MI or presence of Q waves in patients who suffer a stroke and those who do not. (Footnote: ¹²) Further, Witt says (Footnote: ¹³):

How does Dr Saltissi respond to this? His first comment is that he considers the Witt paper to be internally inconsistent. He cites Witt (Footnote: ¹⁴) (relying on the sentence I have underlined):

Of course the first part of that sentence points out the controversy which is a factor of some importance in this case. Again there was argument about this. I take it into account but do not have to determine the specific point.

Dr Saltissi says that he accepts that a clot can form and embolise from anywhere within the LV. Nevertheless he says it is considerably more common at the cardiac apex, on the arterial wall following infarction, and after a full thickness (Q wave/ST elevation) MI rather than a partial thickness/non ST elevation inferior MI as in C’s case. Professor Hall says that when a clot is actually seen in the heart it is more often at the cardiac apex but the clot can occur anywhere and with any kind of MI.

C had been on warfarin prior to his admission to hospital. On 14 May 2013 (the day prior to the MI) his INR was just below the therapeutic range. It is agreed that mural thrombus is susceptible to prevention by anticoagulation. Dr Saltissi accepts Professor Hall’s point that in the time interval between the INR on 14 May 2013 and the MI the level of anticoagulation will have reduced, though unlikely to have returned to its normal level.

If the stroke was caused by an embolus originating in the heart, then Dr Saltissi says that it is only possible to say that the thrombus causing the stroke occurred at some point prior to the onset of the neurological symptoms on the morning of 4 June 2013. (Footnote: ¹⁵)

What happened between the MI and the stroke in terms of anticoagulation? C did not receive further oral anticoagulation after the MI until the time of the stroke. C was treated with heparin (Footnote: ¹⁶) between 15 – 17 May 2013 and between 21 May 2013 and the date of the stroke on 4 June 2013. It therefore appears that there were only 3 days out of the 22 between admission and the stroke when C was not anticoagulated. In fact, anticoagulation (heparin is an anticoagulant) was a problem with C because of the risk of bleeding. Therefore, although he did receive heparin, it was only in therapeutic dosage for the first day/so. Thereafter it was in prophylactic dose of Low Molecular weight heparin. This would reduce the risk of clotting but not abolish it.

According to Professor Hall the anticoagulation result is something of a red herring since, wherever it occurred, C did suffer a stroke by reason of a blood clot. Therefore the anticoagulation did not prevent the clot which caused the stroke. This tells us nothing about where or why the clot formed. Dr Saltissi opined that an in situ clot in the brain was less likely to respond to anticoagulation than a clot in the heart, but I do not consider that the anticoagulation is any pointer against cause (b).

In short cause (b) is entirely possible as the pathway for the stroke to have occurred. On the cardiology evidence it was not, perhaps because of the points made by Dr Saltissi, agreed in part by Professor Hall, a cause which was more probable that the other 3 causes combined; nor did Professor Hall claim this. He said it was the most likely of the 4 candidates. (Footnote: ¹⁷)

If the cause of the stroke was not (b), then Professor Hall contends that each of the other 3 candidates was probably caused by the accident. He said that C would have been in a “metabolic mayhem” by reason of the major trauma of the accident, immobility in hospital, the MI, the severe spinal cord injury and attendant surgery. All these have the capacity to cause a stroke (Footnote: ¹⁸). Shortly after the MI, blood tests were taken. These were taken in relation to the heparin prescription. They show a normal platelet count and mildly raised APTT. (Footnote: ¹⁹) However, according to Professor Hall it is the activity of platelets that is relevant to clotting. Most clots occur when there is a normal platelet count. In clinical terms more subtle tests which might indicate the “metabolic mayhem” were not done. This is because it is universally known that adverse metabolic changes do occur after major trauma etc. That is why, to the extent possible, C was given anticoagulant therapy. A research study could test for the metabolic changes but hospitals do not do these subtle tests because they are known to occur and treatment is prescribed. Professor Hall said that is why C was given Low Molecular Weight Heparin. In the circumstances of a bleed around the spinal cord, if the hospital could have properly avoided anticoagulant treatment, they would have done.

Dr Saltissi did not accept the term “metabolic mayhem”. He said that the main reason for the anticoagulation would be because of the immobility to stop a venous thrombosis. His opinion was that it was very rare for a patient to suffer arterial thrombosis. He accepted that the hospital had done the normal clotting tests and also that after major trauma (including surgery etc) it is possible to get increased clotting. The clotting tests which were done would not necessarily show any metabolic disorder, as Professor Hall had stated. Dr Saltissi’s point was that there was no actual evidence that there was any serious metabolic upset which may have led to a clot in a place other than the heart. He accepted that there was a possibility that if the pathway was cause (a) or (c) then that was caused by the accident, but this was not a probability.

Professor Hall relies on Witt. He accepts that epidemiologically the risk will be at its highest the earlier after the MI that a patient is. The SMR (Footnote: ²⁰) of 44 is for the whole period 0 – 30 days. In the second month the SMR drops to 4.5. What we can say is that at 31 days (and even more so at 20 days) for the Witt cohort it would have been lower than 44 and higher than 4.5. There were other matters:

C had background risk factors. However, if one looks at Witt (Table 2) a not insubstantial number of the cohort had some such background risk factor(s).

On Professor Hall’s evidence, as a result of the accident, C was put in a worse risk category than the Witt cohort because of the “metabolic mayhem”, including that caused by the MI. Therefore either the MI was directly causative of a cardiological embolus or it was part of the metabolic mayhem which led to a clot forming there or elsewhere.

Both cardiologists agree that Witt is a very reputable paper. There is no reason not to use it as a starting point. Dr Saltissi said it should be read with some caution for the following reasons:

It reviewed a US population in a small area. The authors specifically say:

The patients in the study had an MI between 17 and 36 years ago. Management of MI has transformed in that period. C was treated in a way in which one sub group was treated i.e. with low molecular weight heparin. We do not know if the SMR of that sub group was the same or similar to that of the cohort as a whole. Also the cohort as a whole would be treated differently in 2013 in England such that the overall SMR would probably be somewhat different from 44.

Witt was a community based paper not a hospital study.

Witt includes all strokes, though 95% were ischaemic. There is no break down of the cause of the ischaemic strokes.

It is suggested that strokes following STEMI are associated with a worse short term prognosis than N-STEMI.

There is a statistical quirk which neither cardiologist could explain but which suggested that the Killip class (which measures the seriousness of the MI), according to the P value, shows that there is a highly significant difference between those who sustained a stroke and those who did not sustain a stroke.

All those factors have to be borne in mind and all have some importance. Nevertheless there is no better source of information than the Witt paper.

Dr Saltissi referred to a review paper by Weir. (Footnote: ²¹) That paper states:

Professor Venables was asked about the Witt paper. I will take his evidence on this matter shortly. Mr Willems QC probed in relation to the statistical evidence. Professor Venables had to be cautious about certain responses because (a) he is not a statistician or epidemiologist and (b) this was not a matter of dispute between him and Professor Chadwick; therefore the detail of Mr Willems’ questions had not, perfectly reasonably, been anticipated by Professor Venables.

Professor Venables accepted that roughly speaking C’s background risk of a stroke was about 3% per annum i.e. 0.25% per month. The standardised risk for the population as a whole is about 2 per 1000. Even in relation to a 75 year old, C’s background risk was significantly increased. As regards the Witt paper:

The highest risk of a stroke after MI was in the first 30 days. It may be that there is a short period after the MI before the peak incidence occurs and then the incidence will fall.

Witt does not separate out the background risk of the cohort since someone prone to MI is also more prone to a stroke.

Part of the Witt cohort would have had a stroke coincidentally and it is not easy to say what percentage would have had a coincidental stroke i.e. a stroke absent the MI.

In Table 2 of Witt under the heading “cardiovascular risk factors” a substantial percentage of the Witt cohort had cardiovascular risk factor(s) and, since the risk factors often appear together, it is likely that a not insubstantial number had more than one risk factor. Indeed the heading “Comorbid Conditions” shows this.

The Witt paper includes patients who had STEMIs (39%). They would be at greater risk of stroke than somebody who did not have a STEMI.

These are some of the main points which Mr Willems QC explored with Professor Venables and which, in broad terms, Professor Venables accepted. I have not gone into the minutiae of the figures which were put to Professor Venables and on which he made some guesses, because it is not fair to record some of the details because of his lack of epidemiological expertise/notice of the questions. What Mr Willems was mainly seeking to establish is that the SMR of 44 had to drop substantially as it was not known to what level it dropped in relation to C personally.

There is no doubt that one cannot just transfer the SMR of 44 to C as an individual. It is also true that a number of points can be made which appear to reduce that SMR of 44 in relation to him. There are however factors, which cannot be statistically measured, which increase C’s risk since he not only suffered an MI as a result of the accident but also the original major trauma and the spinal operation to decompress the haematoma. Thus there were three, not one, major stressors. In relation to this Professor Venables said that the body’s response to injury is to heal and the body’s response in healing is inflammation. Even something as trivial as a cut on the finger causes inflammation which is part of the body’s stress response to injury in order to advance the healing process. That was much magnified by the events following the sort of injury which C had.

At the end of his evidence Professor Venables said that he still thought the epidemiological evidence was very strong and drove his view that (b) was the route probably followed in C’s case. As regards (a) and (c) Professor Venables said he thought that certainly they could be in the causative chain. If the argument is accepted that you can activate a plaque in the coronary arteries and cause a heart attack, then a plaque in the ascending aorta, the common or internal carotid artery or the origin of the middle cerebral artery is exactly the same physiologically. If somebody could prove that the pathway was not (b) then Professor Venables said that one has exactly the same mechanism for cause (a) and (c) and this gives the same result. So if C had not had the trauma of the accident, (a) and (c) probably would not have happened. This hinges on the pathophysiological similarity of (a), (b) and (c). Wherever an atheromatous plaque is, once it has been activated by the pathophysiological chain of events that go on after trauma, then it does not matter whether it is in the leg, heart or brain. In short, Professor Venables believed that the accident probably caused the stroke through pathway (b) but, if not by pathway (b), by (a)/(c). In response to Dr Saltissi’s suggestion that the background risk was equally likely to have been the cause rather than the accident, Professor Venables said that he did not see how a surge in stroke incidence against a background in Witt of myocardial infarction can just be a coincidence. He said that one does not see those sorts of coincidences where risks are raised 4 fold, 10 fold or 44 fold. Indeed he felt that whichever route (i.e. pathway (a), (b) or (c)) was involved, then he had no doubt that in some way this was triggered through the accident.

Professor Chadwick gave brief evidence. He said that having heard Professor Venables (Footnote: ²⁴) there was nothing which changed his view or led him to wish to clarify his view. He said from the point of the injury leading to the MI, then subsequently the paraplegia, and the stroke were part of a chain of events. He also thought route (b) was more likely, though conceding Dr Saltissi’s point that there was an absence of direct evidence to support (b). He strongly agreed with Professor Venables that whether it was route (a), (b) or (c) there was no real doubt but that there was a causal link from the accident rather than background risk. He said he thought that was the evidence that Witt provided because the SMR, no matter what precise figure you give for a point in time, is a big number and it is impossible to get away from that fact. In answer to the question how confident he was that C would not have had the stroke if he had not had the accident, Professor Chadwick said he was clinically confident and clinical confidence was rather higher and demanded a rather higher level than the balance of probabilities.

This is an unusual case in that the Claimant’s cardiologist and both neurologists are firmly of the opinion that the accident caused the stroke. The sole dissident is Dr Saltissi. Of course, it does not follow from that that the Claimant must succeed. It is necessary to examine the quality of the evidence. This is why I have set it out in such detail. Also it is the court’s function to decide causation, not that of the doctors.

It is conceded by the Defendants that if the court is satisfied on the balance of probabilities that the accident caused the stroke, it is not incumbent upon C to prove which one of the four possible pathways caused it. The Defendant submits, however, that the evidence can merely demonstrate that the accident increased the risk of C’s stroke. Mr Willems QC invited me to say, if I felt I could, having heard his submissions on this point, whether I was persuaded that C had proven causation on the balance of probability. If not, there would have been further argument as to whether material increase in risk enabled C to succeed. After hearing Mr Willem’s submissions, I accepted his invitation and told him that I was convinced that, absent the accident, the stroke would not have occurred on the balance of probabilities.

Having reviewed the evidence in detail my conclusions can be crystallised as follows:

C had a background risk which was higher than a healthy 75 year old. It was about 30% risk over 10 years or, roughly speaking, 0.25% each month.

If one effectively discounts pathway (d) (Footnote: ²⁵) then there were 3 possible pathways to the stroke. Any one of these pathways could have happened absent the accident and any one of them could have happened because of the accident.

Whilst the precise figures in the Witt paper will never apply to an individual, I accept the evidence of the majority that, to put it in Professor Chadwick’s words, the SMR, no matter what precise figures you give for a point in time, different points in time, after the myocardial infarction, the SMR is a big number, and it is impossible to get away from the fact that it is a big number.

Further, Witt dealt with incidence of stroke after MI. C was subject to other stressors, most particularly the major trauma of the accident and of the spinal surgery. Pathophysiologically these are likely to have triggered, if not pathway (b), then pathway (a) or (c).

The evidence does not depend solely on the epidemiology in the Witt paper. It depends also on the pathophysiological explanation and experience of two eminent neurologists. To the extent that a cardiologist can properly have an input into the issue on pathways (a) and (c) (as opposed to (b)) then that evidence is supported by Professor Hall. Dr Saltissi disagrees on pathway (b) and indeed on pathways (a) and (c). He accepts the possibility of the accident having caused a stroke through pathway (a) or (b) or (c) but believes that coincidence i.e. background risk is equally as likely. Of course he may be right but very probably, in my judgment, he is wrong.

In Alphacell v Woodward [1972] AC 824, 827 Lord Salmon said:

For those reasons the answer to the preliminary issue ordered to be tried by the Master is that the road traffic accident that occurred on 13 May 2013 on the clear balance of probabilities caused the stroke suffered by the Claimant on 4 June 2013.