England and Wales Court of Appeal (Civil Division) Decisions

The Royal Victoria Infirmary & Associated Hospitals NHS Trust v B (Child) [2002] EWCA Civ 348 (14th March, 2002)

IN THE SUPREME COURT OF JUDICATURE
IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE-
QUEEN’S BENCH DIVISION-
NEWCASTLE UPON TYNE DISTRICT REGISTRY
(HIS HONOUR JUDGE LANGAN QC sitting as a Deputy High Court Judge)

		Royal Courts of Justice Strand, London, WC2A 2LL
		14 March 2002

TUCKEY LJ:

1. This is an appeal by the Defendant hospital authority from the judgment of HHJ Langan QC sitting as a High Court Judge in Newcastle. The then seven-year-old Claimant (B) claimed damages for severe cerebral palsy caused by the Appellant’s admitted negligence. The only issue the Judge had to decide was the multiplier to be applied to the agreed cost of future care (£150,000 per annum) for the rest of her life. This involved deciding B’s life expectancy and then discounting that number of years to produce the multiplier. The Judge decided that B “should be treated as a child who will live to the age of twenty five years”. This meant that the cost of future care would be incurred over a period of 17·42 years. So for this period and taking a discount rate of 3% the Judge applied Ogden Table 38 (“multipliers for pecuniary loss for term certain”) which produced a multiplier of 13·61.
2. The Appellant says that the Judge was wrong on both counts. Firstly it says he should not have decided how long B was expected to live, but “the average number of remaining years of life of a cohort of individuals with (her) age and condition”. This statistical approach was derived from the evidence of Professor Strauss, Emeritus Professor of Statistics at the University of California. Using this approach Professor Strauss said B’s life expectancy was an additional 14·9 years. Secondly, the Appellant says that the Judge should not have used Ogden Table 38 but, based on Professor Strauss’s evidence, a whole life multiplier which took account of the fact that B might die earlier or live longer than her life expectancy. This produced a multiplier of 10·5. So the Appellant says that the Judge awarded £466,500 too much.
3. The facts as agreed or found by the Judge can be shortly stated. B was born on 5 October 1993. When she was twelve weeks old she developed a bronchiolitic illness which resulted in her admission to the Appellant’s hospital in Newcastle. There she suffered a severe hypoxic episode, the consequences of which have been devastating. She is quadriplegic, cannot sit up or roll over, has no useful movement of her hands and suffers from muscle spasm which is partially controlled by drugs administered by a pump inserted in her spinal column. She is cortically blind, cannot speak and gives little indication that she understands what is going on around her. Feeding is very difficult and there is a significant risk that she will require a gastrostomy so that she can be fed by tube. She is doubly incontinent and totally dependent upon others for her care.
4. As well as Professor Strauss the Judge heard evidence about B’s expectation of life from two consultant paediatricians. He preferred the evidence of Dr. Kovar called on behalf of the Defendants. Dr. Kovar’s opinion, based on seeing B at hospital and at home, published information by Professor Strauss among others and his clinical experience and judgment was that on the balance of probabilities B would not live beyond the lower end of an age range of twenty five to thirty five. Whether or not B was tube fed was the principal factor affecting survival although there were other uncertainties as well. Hence the range. The Judge described Dr. Kovar’s approach as “holistic” and one which inspired confidence. He accepted Dr. Kovar’s assessment of twenty five to thirty five years as correct and continued :

“It is fair to say ........ that Dr. Kovar acknowledged that his wide range of twenty five to thirty five came from Professor Strauss’s published work, that his own particular input was to place B at the bottom of the range, and that – I quote Dr. Kovar’s oral evidence – “If a court came to the conclusion that the right range was Strauss’s twenty two to twenty five I would be at the lower end of that. I would have to move down.” For reasons which I gave when dealing with Professor Strauss’s evidence, I am not persuaded that a range should be selected purely by reference to his statistical work. In my judgment, sufficient weight is given to that work and to Dr. Kovar’s understandable caution by placing B at the bottom and not just in the lower part of the Kovar range.”

5. I have already referred to Professor Strauss’s definition of life expectancy. He said it was impossible and unnecessary to predict the survival time of an individual. He had compiled a database for everyone who has received services for developmental disabilities from the State of California between 1980 and 1996. There were 40,000 cases of cerebral palsy of which 923 were children who had attained B’s age and were tube fed. He had interrogated his data base and constructed a life table which showed that B’s life expectancy was an additional 14·9 years if she was tube fed and 17·8 years if she was fed by others. If one took the median, as opposed to the average age, the figures were 10·5 and 14·4 additional years.
6. The Judge criticised Professor Strauss’s evidence because it ignored the individual claimant, downplayed the role of the clinician to an unacceptable degree and was too mechanistic. The characteristics by which Professor Strauss identified his cohort were unrefined and too narrow in scope. He concluded that :

“Those criticisms ........ are in my judgment sufficient to demonstrate that it would be wrong, at least in this case, to depart from the conventional manner of determining life expectancy. This is done by reference to general life tables in the ordinary run of cases and on the basis of medical evidence in the special type of case like B’s.

I accept that statistics are a useful tool in the hand of the clinician but where reliable medical evidence is before the court they should not displace the expertise of the clinician. They provide, rather, a useful background to, and cross-check on, the work of the doctor.”

7. Ogden Table 38 only discounts for accelerated receipt. Professor Strauss contended that a further discount should be made for mortality within the life expectancy determined by the court. The Judge rejected this contention for reasons of convention (the suggestion was quite novel), authority (Thomas –v- Brighton HA [1999] AC345) and principle (it amounted to a double discount which meant that if B survived exactly to her life expectancy she would receive less than she had been deemed to require for her care).
8. Mr. De Navarro QC for the Appellant has mounted a full-scale attack on the Judge’s conclusions. I have already referred to two of the reasons why he says that Judge was wrong. But his first reason is the Judge should not have preferred Dr. Kovar’s evidence to that of Professor Strauss. He submits that the Judge overlooked or took insufficient account of the fact that Dr. Kovar’s range (as opposed to his position within the range) was not based on clinical judgment but on Professor Strauss’s work and that he accepted at trial that if that work showed the range to be lower (22 to 25 years) then his clinical judgment was that B was at the lower end of that range. Dr. Kovar had accepted that interrogation of Professor Strauss’s database was the appropriate starting point and “the best we have got”. Furthermore Dr. Kovar’s range of 25 to 35 was based on a mathematical error.
9. In support of these submissions we were shown passages from the evidence of the Claimant’s paediatrician, Dr. Lloyd, whose evidence at trial was that the range was twenty-five to twenty eight. The Judge however recognised that this was a coincidence and did not place reliance on this evidence in reaching his conclusion. For this reason I do not think it helps us either to refer to Dr. Lloyd’s evidence. The simple question we have to answer is whether it was open to the Judge to reach the conclusion he did based on the evidence of Dr. Kovar. This is not the same question as whether he should have preferred the evidence of one of the Appellant’s experts to that of the other. That is one of the curious features of this case. The Appellant needed Dr. Kovar’s evidence to refute Dr. Lloyd’s evidence that the outlook for B was optimistic so her expectation of life was at the top of any range. The Appellants however did not need Dr Kovar to contradict Professor Strauss, but if they called him, on the face of the reports exchanged before trial, that is what he would do. The real question is whether he maintained the views expressed in his reports at trial.
10. We have been taken through the transcript of Dr. Kovar’s evidence. There are admittedly passages where he comes close to admitting that he would defer to Professor Strauss’s data base in determining the range and that his clinical assessment was only relevant as to where B fell within that range, but he did not say this in terms. Given that he was the Appellant’s expert, if that really was his view, one would have expected him to say so clearly from the outset of his examination-in-chief and he did not.
11. This is not surprising if one looks at Dr. Kovar’s latest report. He refers to Professor Strauss’s earlier published work “Prognosis for survival in function in children with severe developmental disabilities [1997]” which showed that B had a probable (i.e. median) life expectancy to at least fifteen years and to “Life expectancy of adults with cerebral palsy” which showed that a fifteen year old would probably survive to twenty five if tube fed and thirty to thirty five if not tube fed. Dr. Kovar went on to say, “on this reasoning there is a 25% chance of her surviving to twenty five to thirty five years”. This is obviously a mathematical error, but I do not think it seriously undermines the approach taken by Dr. Kovar in this report. He continued :

“I understand Professor Strauss is of the view that this approach is flawed. (He has the benefit of the original data). He calculates B’s life expectancy (defined as average number of remaining years of life of the cohort of individuals with that age and condition) as 14·9 additional years, that is to the age of 22·2 years. He argues that this value is more appropriate for the court to consider than “median”. I have argued as above in order to offer a probability view.”

So by the time he wrote this report Dr. Kovar was aware of Professor Strauss’s life table for B. He cannot be criticised, as Mr. De Navarro sought to do, for offering a probability view since that was the view offered by Professor Strauss in his published work.

12. Dr. Kovar’s report continued with references to other published work by Hutton and then says :

“Experience does recognise that survival into adulthood is not uncommon. There is a view in the UK that reliance on the combined Strauss data over-estimates life expectancy. B’s disabilities moreover are more marked than most in the high LAS group in Hutton.

In order to arrive at a balanced opinion on which to advise the court my opinion in this case is based on

- my assessment of the severity of B’s disabilities and the positive and negative factors discussed above.

- my understanding of the published data pointing to survival into adult life of children like B whilst recognising the question marks around this data as discussed above, seeing the dedicated care she receives at home,

- my experience over years of similar children now young adults in their twenties, but equally of several apparently well CP children – adolescents who were disabled as B is now who later died in childhood or in their teens from intercurrent infection, or suddenly and unexplained, with incidental illness, or in association with a seizure (epilepsy), and

- taking account of various colleagues views on prognosis and life expectancy in similar cases of children both similarly and worse affected than B,

- applying my own clinical intuition and judgment.”

These considerations led him to conclude, “B is unlikely to live beyond the lower end of this twenty five to thirty five year range”.

13. Professor Strauss’s more general studies referred to in Dr. Kovar’s report produced more optimistic forecasts of expectation of life than the customised life table he produced for B. One is bound to ask why. The explanation is said to be found in the second paper which says :

“A limitation of the life expectancy and survival analysis reported here. ...... and in other studies is that they are based on only a crude classification of functional level. More appropriate estimates of life expectancy and survival probabilities for a subject with a given profile of age, gender, functional skills, etc. may be obtained using the methods described here.”

This less crude classification is what Professor Strauss says he performed in this case based on data which he had asked for from the paediatricians but which the Judge described as unrefined and too narrow in scope. But by producing these less crude statistics Professor Strauss was looking at matters which were very much for the clinician e.g. functional skills.

14. It follows I think that Dr. Kovar’s evidence can be explained on the basis that he took the range from Professor Strauss’ more general work and then used his own clinical judgment and the other matters to which he refers in his report to refine his conclusion. This is why I am not surprised that Dr. Kovar maintained the views which he expressed in his report. It follows that I think it was open to the Judge to base his conclusion about B’s expectancy of life upon Dr. Kovar’s evidence.
15. I should perhaps add that the passage which I have quoted (paragraph 6) in which the Judge referred to his criticisms of Professor Strauss’s evidence suggests that the Judge thought that acceptance of Professor Strauss’s evidence would be a departure from the conventional manner of determining life expectancy. If he had meant by this that the court should not have regard to relevant statistical evidence he would have been wrong. But I do not read his judgment in this way. What he is saying is that it would be wrong to decide expectation of life purely by reference to Professor Strauss’s statistics. That would be a departure from convention and certainly was the tenor of some of the material emanating from Professor Strauss who had written a separate report identifying what he described as “the many problems and misconceptions” contained in the paediatrician’s reports.
16. The next question is whether the Judge misunderstood the task he had to perform. Mr. De Navarro submits that he fell into error by attempting to predict B’s actual survival time rather than estimating her life expectancy. The former is both unpredictable and unnecessary: unpredictable because no-one can predict actual survival times; unnecessary because life expectancy can be determined from a life table of a comparable population (the cohort).
17. The Judge’s task was to award a lump sum for future loss so as to put B, as nearly as possible, in the same position financially as if she had not been injured. The sum awarded might prove too little or too much but the Judge was still required to make his best estimate of what that sum should be. The court must do the best it can (see Wells –v- Wells [1999] 1AC 345 at 363, 364, 389 and 395).
18. In the ordinary case where the court is considering a claim for future loss for life it will be able to use actuarial tables as its starting point for determining the multiplier. In such a case the court will be using statistics relating to the general population to predict the life expectancy of the claimant in question. But it was common ground that such statistics could not be used in this case because B’s expectation of life had been substantially reduced by her disability. So in this case the Judge had to decide what B’s expectation of life was. In other words he had to make his best estimate of how long she would live. Without that estimate he could not determine what lump sum to award so as to give her the means to provide the care she needed for the remainder of her life. Whether an actuary or statistician would call this estimate her “expectation of life” is not to the point. It is inescapably what the Judge had to decide. In making that decision in this case it is common ground that the Judge could and should take into account relevant, statistical and/or actuarial evidence. Professor Strauss’s evidence was admitted without objection for this purpose and the Judge made it clear from the passages from his judgment which I have quoted that he took this evidence into account. What the Appellants’ submission comes down to is not an argument about the question the Judge asked himself but an assertion that the Judge was bound to accept Professor Strauss’s evidence. He was not. It is for the Judge and not experts to decide issues of this kind, although of course in a case such as this, he will be much guided by expert evidence.
19. On the view I have taken of Mr. de Navarro’s first point the Judge was entitled to accept Dr. Kovar’s evidence, based as it was on his own fairly limited experience and the reported experience of other members of his profession and others including Professor Strauss. Statistics alone cannot determine any case of this kind; some medical input is required as well. Other factors may also be relevant such as the standard of care which the parents are able to provide.
20. In his skeleton argument Mr. de Navarro boldly contended that “resolution of the appeal should render it unnecessary to call Professor Strauss to give oral evidence in other cases in the future ........” This contention is presumably based on the assumption that the appeal is allowed. But if it is not, I do not think we should be taken to have accepted this. In this case no statistician was called by the Claimant. This obviously created difficulties for the Judge, as he said. But based on this one case I do not think it would be right to say that a customised life table prepared by Professor Strauss based on his Californian data base should be given the status of the tables produced by the government actuary. This is not, I emphasise, to say that Professor Strauss’s evidence or the evidence of any other statistician or actuary is inadmissible. In an appropriate case such evidence may well provide a useful starting point for the judge, but if it is to serve this purpose Professor Strauss or any other such expert should be required to give evidence if his report is not agreed. Such evidence, together with medical evidence, should provide a satisfactory inter-disciplinary approach to the resolution of issues of the kind which arose in this case.
21. So I turn to the final question on this appeal which relates to the multiplier taken by the Judge. Should the discount have simply been for accelerated receipt (Ogden Table 38) as the Judge decided, or should it also have included a discount for mortality to reflect the fact that B would not live exactly to her life expectancy?
22. The simple answer to this point, one would have thought, was that the quantification of B’s expectation of life already took into account the chance that she might die earlier or live longer and that she would almost certainly not die on the predicted date. If one also had to take mortality into account in determining the discount this would, as the Judge said, be a double discount.
23. Mr. De Navarro however, sought to demonstrate that this was not the case with the aid of a table. This showed, he said, that even on the Ogden Tables there was a small difference in the multipliers, but in this case where there was a larger chance of B dying earlier than living longer, the difference would be greater. Professor Strauss’s calculations suggested that a further reduction of about 1·6 should be made.
24. I remain wholly unpersuaded by this exercise although I confess I did not fully understand it or the reason why the figures differ. However, I feel safe in my ignorance because it seems to me that the Judge’s approach followed that of the House of Lords in Wells –v- Wells and the other cases decided with it. Thus, in Thomas –v- Brighton HA the agreed medical evidence was that the claimant had a life expectancy to age sixty. Both the trial judge and the Court of Appeal had taken the Table 38 multiplier but reduced it to discount for further contingencies. The House restored the full Table 38 multiplier. Lord Lloyd said at 378 D to E

“There is no room for any discount in the case of a whole life multiplier with an agreed expectation of life. In the case of life expectancy the contingency can work in either direction. The plaintiff may exceed his normal expectation of life or he may fall short of it.

There is no purpose in the courts making as accurate a prediction as they can of the plaintiff’s future needs if the resulting sum is arbitrarily reduced for no better reason than that the prediction might be wrong. A prediction remains a prediction.”

There is no difference in principle between an agreed expectation of life and one determined by the judge.

25. Lord Lloyd made similar comments about an earlier decision by McCullough J in Janardan –v- East Berkshire HA [1990] 2 Med. L.R. pointing out that the effect of the increased discount was that the claimant would run out of money before he reached the age to which the court had decided he would live. This would also be the effect of reducing the multiplier in our case. This together with the fact that we are bound by the decision in Wells –v- Wells persuades me that the Judge applied the right discount in this case.
26. For these reasons I think this appeal should be dismissed.

Sir Anthony EVANS:

27. I agree entirely with Tuckey LJ’s judgment and that this appeal should be dismissed. As to the underlying principle, there is no doubt. The court, as in any action where damages are claimed for the long-term consequences of personal injury, must do its best ‘to arrive at a lump sum which represents as nearly as possible full compensation for the injury which the [claimant] has suffered’ (per Lord Lloyd in Wells v Wells [1999] 1 AC 345 at 363f). In fairness to both parties, the award must be no greater and no less than that amount. In making its assessment, the court must make necessary assumptions on the basis of the evidence that is available to it (per Lord Hope of Craighead (at 390A). ‘The calculation should make the best use of such tools to assist that process as are available’ (per Lord Hope at 390E).
28. In the present case, experienced leading counsel agreed all the figures relevant to the calculation of the amount of damage, except one. The agreed cost of special nursing and other care for the claimant, for the whole of the rest of her life, is £150,000 per year. By how many years should that figure be multiplied? Given that £150,000 was the multiplicand, what should be the multiplier?
29. The appropriate method of establishing the multiplier since the judgment of the House of Lords in Wells v Wells is to assess the appropriate life expectancy figure and then discount it, so as to take account of immediate payment of a lump sum figure. The correct starting point is to derive a life expectancy figure from published actuarial tables, but this is subject to special factors applying in particular cases (per Lord Lloyd in Wells v Wells at 379F).
30. Here, the claimant was just over seven and a half years old at the date of the trial in May 2001. Tragically, the disabilities from which she suffers are such that actuarial tables which are relevant for the population at large can have no application in her case. The court has to make the best estimate that it can of her individual life expectancy, based primarily upon the opinions of medical expert witnesses. There is also scope for statistical evidence, where it is available, based upon records of other cases of comparable disabilities. Such evidence was available here from a distinguished statistician, Professor David Strauss PhD (Sussex) FASA, who is now based in California and has specialised in this field. Unsurprisingly, his published work was known, or made known, to the distinguished paediatricians who gave evidence, Dr Ben Lloyd for the claimant and Dr Ilya Kovar for the defendants.
31. Dr Lloyd and Dr Kovar made similar estimates of the claimant’s life expectancy. Dr Lloyd expected that she would live to between 25 and 28 years of age, but he had demonstrated some inconsistency and the judge placed little or no reliance upon his evidence. Dr Lloyd considered that there were positive factors which outweighed negative aspects of B’s condition, and that this justified taking a figure at the upper end of the range. Dr Kovar was more pessimistic. He put the range at 25 to 35 years, and he thought that her case should be placed at the lower end of it. The judge expressly preferred Dr Kovar’s evidence on this point.
32. The two doctors, therefore, made similar estimates, but they reached them by different routes. ‘We have come in the end to very similar numbers, by slightly different routes, with different emphasis’ (Dr Kovar transcript page 26E).
33. Dr Strauss gave evidence of a different kind. He identified a group or ‘cohort’ of 923 cases from his accumulated data which corresponded with B’s age and her condition as it was reported to him. He then gave specific figures for her life expectancy; 14.9 years if she should be compared with children who need tube feeding, 17.8 years if she should not. Adding these figures to her current age, his opinion as a statistician was that the average age to which a child of her age and physical condition could be expected to live was within a range of 22.4 to 25.3 (say, 22 to 25) years.
34. Much of the evidence at the trial and of the submissions which we heard on this appeal was taken up with a close examination of Dr Strauss’ statistical concepts and methods. Even with the assistance given by his evidence, I cannot claim more than an imperfect understanding of them. What emerged, as I understand it, is that he was content to use average rather than median figures, and that he could offer no view on whether tube feeding should or should not be taken into account in B’s case. Therefore, he could only produce a range derived from the two figures which he gave. He insisted, however, that he was ‘definitely not trying to predict survival time’ (transcript page 22B). This led to some uncertainty as to what his life expectancy figures were, if not just that, but as I understand his evidence he was saying no more than that the figures were an average derived from his records rather than a reliable prediction in any individual case.
35. The issue raised by this appeal has been described as an issue of principle: was the judge entitled to prefer the view of (one of) the doctors to the evidence given by Dr Strauss, based on relevant data (described as ‘the best we have got’) and his statistical expertise?
36. In my judgment, this is not a question of principle, and whether it is or not, the answer is clear: the evidence of a statistician is both relevant and admissible, and the judge must take account of all the evidence, including this, when deciding what assumption he should make as to the future life span of the claimant. On analysis, it became clear that the issue raised by the defendants/appellants is rather more mundane. It arises from a forensic difficulty in which they found themselves at the trial. Their medical witness, Dr Kovar, knowing of Dr Strauss’ evidence, maintained his opinion that the best estimate of B’s life span was at the lower end of the bracket, from 25 to 30 (or 35) years. Dr Strauss’ bracket was lower, 22 to 25 years, and Dr Kovar conceded (in re-examination) that “logically”, if that was the correct range, he “should move down in relation to that as well” (transcript page 33G). The defendants relied upon Dr Kovar’s evidence in order to challenge the more optimistic medical prognosis expressed by Dr Lloyd, and they did this successfully. Their complaint now is that the judge was wrong to accept Dr Kovar’s evidence rather than that of their other witness, Dr Strauss, as to what the appropriate range or bracket should be.
37. In my view the judge dealt with this admirably in paragraph 66 of his judgment, culminating in -

“.... I am not persuaded that a range should be selected purely by reference to [Dr Strauss’] statistical work. In my judgment, sufficient weight is given to that work and to Dr Kovar’s understandable caution by placing B at the bottom and not just in the lower part of the Kovar range.”

38. The judge also said, and I agree:

“I accept that statistics are a useful tool in the hand of the clinician but where reliable medical evidence is before the court, they should not displace the expertise of the clinician. They provide, rather, a useful background to, and cross-check on, the work of the doctor.”

39. I would add only this. I doubt whether such a clear distinction exists between “statistical” and so called “clinical” evidence, as Mr de Navarro asserted in his attractively presented submissions. When a doctor gives expert evidence, the court relies on the witness for a professional opinion which takes account of contemporary knowledge and expertise, not limited to his or her own personal experience but including reports of the experiences of other doctors and other published material. The published material clearly includes statistical information such as was produced by Dr Strauss, though not, of course, the individual assessment made by him in the present case. Whether or not such an assessment has been made, the court must still rely primarily, in my judgment, on expert medical witnesses before reaching a conclusion in the particular case. It would be wrong to allow a statistician, or an actuary, to do more than inform the opinions of the medical witnesses and the decision of the court, on what is essentially a medical, or clinical, issue.
40. On the second issue, Dr Strauss explained why, in his opinion, when the projected total cost of future care is discounted for accelerated payment of a lump sum by way of damages, the discount should also include the risks of mortality, as shown by actuarial tables for the period of life expectancy. Nothing could be clearer, in my opinion, than that those risks are totally different from the reasons why, tragically, the period for a young child is as short as it is in the present case. There appears also to be a mathematical basis for this submission, derived from the proposal that the individual has a greater chance of pre-decreasing the average period than of extending it. However that may be, a multiplier which produced a lesser figure than is necessary to provide for the injured person until he or she reaches the average age cannot, in my judgment, be consistent with the underlying principle as it is applied by the courts.

THORPE LJ:

41. I have had the advantage of reading the judgment of my lord, Tuckey LJ, in draft. I gratefully adopt his statement of the facts and his conclusion and reasoning on the appellant’s second submission. I differ from him only on the appellant’s first submission for reasons which I will endeavour to explain.
42. At the trial the judge’s first task was to establish the claimant’s life expectancy. On that finding depended his second task, namely to fix the appropriate multiplier. At the outset of his judgment he stated that the claimant’s counsel urged the adoption of the orthodox method for establishing the claimant’s life expectancy. The judge thus described the method:

“The first orthodoxy or convention is that where the life expectancy of a claimant is in issue, it is most appropriately determined on the basis of medical evidence. In this case, it is said, interrogation of a data-base which contains the history of patients whose condition is similar to B’s will provide an answer to the question – ‘what is her life expectancy?’ which is more likely to be accurate than anything coming from the doctors who have seen her.”

43. I do not accept the orthodoxy is thus safely stated. Further what follows does not seem to me to be a fair summary of Mr de Navarro’s submission.
44. The judge returned to what he regarded as conventional in paragraph 47 of the judgment when he said:

“.... It would be wrong, at least in this case, to depart from the conventional manner of determining life expectancy. This is done by reference to general life tables in the ordinary run of cases and on the basis of medical evidence in the special type of case like B’s.”

45. With that citation I agree to the extent that, obviously, in the majority of personal injury cases, where the claimant is not left with significant permanent disability, life expectancy is established by the use of life tables produced by the Government Actuaries Department, conveniently found in the tables for the calculation of damages compiled by the members of the Professional Negligence Bar Association. Equally where the claimant has a partial permanent disability it will be conventional to make use of the same tables with appropriate adjustment to reflect the disability factor. No doubt the judge in making that adjustment will depend upon expert medical advice specific to the claimant.
46. However in cases such as the present where the defendant’s negligence has resulted in massive permanent disability the life tables have nothing to offer. But I do not accept, as the judge appeared to accept, that in such cases life expectancy depends upon the expert medical evidence either alone or as the primary contribution. Obviously the assessment of life expectancy in any case can produce nothing better than an informed estimate. The task is simple and uncomplicated wherever the life tables apply since they are founded upon a statistical assessment of a very large human population. Without that sure guide the task becomes much more difficult. Furthermore error may lead on the one hand to substantial under compensation and on the other hand to an excessive abstraction of public or insurer’s funds. So a judge needs all the help that he can get. It should not be regarded as a field in which voices of two distinct disciplines compete for dominance. What is surely required is interdisciplinary co-operation. The statistician alone cannot provide the judge’s needs. His conclusions are only drawn from a relatively small human population group whose disabilities roughly match those of the claimant. Equally a medical expert cannot meet the judge’s needs. Although he can offer profound understanding of the claimant’s case, his knowledge of comparable cases is limited to what chance has brought to his professional care during the course of his career.
47. This was frankly acknowledged by the claimant’s expert, Dr Lloyd, when he said (at 2/730):

“Precise attempts at precise calculations in life expectancy is not part of my daily work.”

48. In answer to the next question as to whether it was a straightforward area he replied:

“Well, I have been doing it for medico-legal purposes for a few years and it is not at all straightforward, and this case brought home to me even more than that – it brought home to me even more. It’s a very complex matter. I think everyone finds it complex, difficult and imprecise.”

49. The collaboration between the statistician and the medical expert is simple to define. The statistician provides the starting point when he reports outcome for the sample of past comparable cases that he has established. As this case illustrates, Professor Straus is a world expert who has the largest collection of comparables. His database contains particulars of about 200,000 people, 40,000 of whom have suffered cerebral palsy. Again as this case illustrates, he relied upon medical experts, Dr Lloyd for the claimant and Dr Kovar for the defendant, to complete a relatively detailed questionnaire to enable him to determine where in the spectrum of cerebral palsy disability B lies. With that information he was then able to draw from his database of 40,000 a group of 923 whose age and disability broadly match B’s.
50. From that group of comparables Professor Strauss was able to construct a bespoke life expectancy table to match the facts of B’s case. That provided the judge with a starting point rather than a conclusion. It was then the task of the medical experts to bring to bear both their profound understanding of B’s case and also their general clinical experience to guide the judge as to how the starting point drawn from a group of comparables should be adjusted in reliance upon their expertise.
51. It appears that this collaboration was broadly understood between the three experts. It was common ground that Professor Strauss’ database was the best available and his interpretation provided the starting point. It was equally common ground that it was for the doctors to advise the judge as to specific refinements. Dr Lloyd accepted Professor Strauss’ opinion and contended for a significant increase in B’s life expectancy to reflect specific factors which he identified to be on balance strongly positive. Dr Kovar, by relying upon Professor Strauss’ published work, adopted a significantly higher starting point but curiously arrived at a conclusion similar to Dr Lloyd’s by assessing the overall balance of specific considerations as being strongly negative.
52. On the assessment of the specifics the judge preferred the evidence of Dr Kovar to that of Dr Lloyd. Thus for me the central question in the appeal is whether the judge was entitled to accept Dr Kovar’s conclusion when his starting point was drawn from Professor Strauss’ published work rather than from Professor Strauss’ advice to the judge in his written reports and in his oral evidence. Before coming to that question I will consider briefly the judge’s assessment of Professor Strauss’ contribution. In paragraph 35 the judge said:

“I have to confess to having found the evidence of Dr Strauss, whether written or oral, extremely difficult to follow.”

53. I have considerable sympathy with that confession, since I too having no training in the statistical field, have at times struggled when reading Professor Strauss’ reports or the transcript of his evidence. However in the end it is necessary to recognise that Professor Strauss is a statistician of great eminence. The claimant did not call any contrary evidence, indeed the claimant’s only expert accepted Professor Strauss’ opinion. The judge criticised him for introducing the median figure, when asked how long B was likely to live. In my view the criticism was misplaced. To a statistician the median figure has obvious attractions. But it was never any part of the defendant’s case that the judge should adopt the median figure, or even give it any weight, in the performance of his task.
54. Nor do I accept the judge’s criticism that Dr Strauss had focussed ‘on the cohort to which an individual appears to him to belong rather than on the individual himself’. That criticism seems to misunderstand the function of the statistician and the division of function between the statistician and the medical expert. Professor Strauss’ only contribution was to focus on a group of comparables. His only focus on the individual was derived from the paediatrician’s answers to his questionnaire which enabled him to identify the group of comparables.
55. Nor was Professor Strauss to be criticised for saying that the determination of the actual survival time of a person is both impossible and unnecessary. The judge went on to say that that was his essential task, however impossible it might be to fulfil with precision. But that is only to state the different function of the expert and the judge. The judge’s final criticism was that Professor Strauss appeared to him to down play the role of the clinician ‘to a wholly unacceptable degree’. But there were inevitable tensions in the presentation of the defendant’s case, to which the judge nowhere refers. Mr de Navarro needed Dr Kovar to establish the strongly negative balance of the factors specific to B. But he could not introduce that evidence without Dr Kovar’s elevated assessment of the starting point. Whilst Dr Lloyd had gradually reduced his assessment of the starting point in reaction to Professor Strauss’ opinion, Dr Kovar had made no such adjustment. Thus in his final submissions Mr de Navarro’s criticism of Dr Kovar, and it seems to have been very muted criticism, was in his interpretation of Professor Strauss’ published material rather than in his own field of paediatric medicine.
56. This fundamental conflict internal to the defendant’s case was not mentioned by the judge in relation to the evidence of Dr Kovar either. That may be because Mr de Navarro wanted to have his cake and to eat it. If he too openly and too robustly criticised Dr Kovar’s adoption of Professor Strauss’ published material and then his failure to adjust to Professor Strauss’ opinion in the case, he risked to undermine the judge’s confidence in preferring Dr Kovar to Dr Lloyd in the paediatric field. After all, perhaps the judge’s principal stated reason for rejecting Dr Lloyd was an impression of inconsistency derived from Dr Lloyd’s gradual shifts in reaction to Professor Strauss’ opinion.
57. The judge’s adoption of Dr Kovar’s figure for life expectancy is essentially explained in paragraph 58 of his judgment:

“Dr Lloyd seemed to me to be attempting to fit B into or around the Strauss figures which had for him become rather too much of a straight jacket. Dr Kovar, on the other hand, brought to the case what I have described as a holistic approach. He said that he put into the equation – and I accept his evidence as to this – his examination of B, the published information, his own clinical judgment, his experience and the views of colleagues. To me that is an approach which inspires confidence.”

58. The problem with this explanation is that it simply ignores the criticism that Dr Kovar should not have relied upon his original interpretation of the published material, refusing to make any adjustment in reaction to Professor Strauss’ developing opinion. The closest Dr Kovar came to adjustment was in re-examination when there was the following exchange (2/868):

“Suppose the court came to the view, apparently supported by Dr Lloyd, that certainly subject to the question of moving it one way or the other, that the proper range is Professor Strauss’ 22-25?

A – I would be closer to 22 than 25, because I believe B is in the worst category of Professor Strauss’ database, and I then throw in the add-on which is things such as the Baclofen Pump, the pepper and salt, so I would move down. I have used that range from Professor Strauss’ data. I am told that that was quite inappropriate that this should be the range, on the strength of his data, and logically I move down in relation to that as well.”

59. Mr de Navarro criticises the judge for accepting Dr Kovar’s opinion on two other grounds which I do not propose to consider partly because they were less persuasive but principally because in my opinion the primary submission succeeds. In my opinion the judge fell into error in failing to distinguish the role of the statistician and the role of the paediatric expert. The statistician had a role, a crucial role that did not conflict with the role of paediatrician’s. Had the judge separated the stages necessary to build up to his conclusion he would have been bound to conclude that Professor Strauss’ evidence as to the starting point was unchallenged save by an expert in another field who purported to base his differing opinion on the expert’s own publications. On the expert evidence, once the judge had rejected Dr Lloyd’s paediatric expertise, there was, in my opinion, only one tenable finding as to life expectancy and that is the finding for which Mr de Navarro contended.
60. Accordingly I would have allowed this appeal in part.
Order: Appeal dismissed; costs to be assessed; leave to appeal refused. (Order does not form part of the approved judgment)