Ipsofactoj.com: International Cases [2003] Part 11 Case 3 [CAEW]




- vs -





29 OCTOBER 2002


Lord Justice Latham

  1. The appellant, who is 50 years old, developed a non-Hodgkin’s lymphoma which manifested itself as a lump under his left arm in the autumn of 1994. He saw the respondent, his general practitioner, on the 22nd November 1994, who wrongly diagnosed it as a lipoma, that is a benign collection of fatty tissue, and negligently failed, as the judge found, to refer him to a specialist for confirmation or otherwise of his diagnosis. In 1995, the appellant moved with his family to Lincoln. On the 22nd August 1995 he saw his then general practitioner and complained, again, of the lump. This general practitioner also considered that the lump was probably a lipoma, but referred him on a non urgent basis to the Lincoln County Hospital for an investigation. In his referral letter he described the lump as having gradually enlarged over the past year. The surgeon who examined him on the 2nd November 1995 suspected a lymphoma and arranged for an urgent biopsy. This confirmed the surgeon’s suspicions.

  2. There was some delay thereafter as a result of difficulty in identifying the nature of the lymphoma; but a CT scan did not reveal any spreading of the disease elsewhere in the body. However, on 13th January 1996 he was admitted to hospital with acute and intense chest pain which was a result of the lymphoma having spread, in particular, into the left pectoral region. Chemotherapy was administered on six occasions and was then supplemented by a course of radiotherapy. Although the tumour responded, it did so incompletely. It was therefore decided that the appellant should be subjected to high dose chemotherapy, involving the harvesting of stem cells to preserve them from destruction, the administration of chemotherapy and the replacement of the stem cells. This treatment took place at Leicester in August 1996. He was discharged in early September 1996.

  3. In early 1998 the appellant suffered a relapse when he developed a tumour in the right axilla, which statistically gave rise to a very poor prognosis. The result was that the chemotherapy that he was then given was intended merely as palliative. He was told that he could not be cured. In this context cure means a period of remission of at least ten years since the disease was last evident. In April 1998 there was thought to have been another relapse, although this was never demonstrated histologically. Nonetheless a further course of palliative chemotherapy was prescribed. Happily, there has been no further recurrence of the disease since then.

  4. As might be expected, the effects on him and his life have been devastating. He suffered severe side effects from the original treatment, in particular the high dose chemotherapy treatment in August 1996. He had to give up work. He felt very ill and has continued to feel weak and lacking in energy ever since. Since his relapse in early 1998, he quite reasonably believes from what he has been told that he is living on borrowed time. The effects of that on him and his family can be readily imagined.

  5. The judge concluded that the failure of the respondent to refer the appellant for a specialist opinion in November 1994 delayed treatment by about nine months, that is from April 1995 to January 1996. He concluded that there would have been no material change in the tumour between November 1994 and April 1995, and that the important escalation occurred in the autumn of 1995. He also accepted that the delay in treatment had significantly reduced the appellant’s chances of survival, in the sense of a cure as I have described it, to 25%. Nevertheless, he dismissed the appellant’s claim for damages on the grounds that the appellant had not established on the balance of probabilities that the negligence had had a material effect on the outcome of the disease. He did so on the basis that the evidence before him established that for a person with the type of lymphoma from which the appellant suffered, the prospects of a cure were in any event less than 50%, in other words that it was more probable than not that he would have been in his present position even if treatment had commenced in April 1995. He based his reasoning on the decision of the House of Lords in Hotson v East Berkshire Health Authority [1987] AC 750. His conclusion, essentially, was that the medical evidence before him established as a matter of past fact that the lymphoma was of such a nature that as a matter of probability the appellant would not have been cured, in the same way that the probability was that the avascular necrosis from which the appellant suffered in Hotson would probably have occurred even in the absence of the intervening negligence by the doctor.

  6. Mr. Maskrey QC, on behalf of the appellant, firstly submits that the judge was wrong in his analysis of the medical evidence, and that properly analysed, it established that on the balance of probabilities, had the appellant been treated in April 1995, he would have only required one course of chemotherapy which would have resulted in a cure. He submits, secondly, that even if the judge’s assessment of the evidence was right, he failed to appreciate that the exercise on which he was engaged in relation to the statistical chances of survival for the appellant was not one of causation but of quantification. He submits that there can be no doubt that the respondent’s negligence caused loss and damage to the appellant. It was accepted by the judge that prompt treatment would have prevented the spread of the tumour. The spread of the tumour was itself an injury to the appellant as was the pain caused by its invading the left pectoral area. The cause of action was therefore complete. The sole question was the assessment of damages flowing from the completed tort. The award of damages involved therefore an assessment of the appropriate sum to compensate for the effects of the enlargement of the tumour. This required an assessment of the appropriate sum to compensate for the pain and damage to the pectoral tissue, an assessment of the extent delay resulted in more intensive, and therefore damaging, treatment, an assessment of the increased risk of relapse and adverse effect on prognosis and finally an assessment of the effect on the appellants expectation of life. All these matters, he submits, are common or garden problems confronted by courts when dealing with the assessment of damages and raise no issues of principle.

  7. His third and most fundamental submission is that as an alternative to his second submission, this court should revisit the question of whether or not loss of a chance in the sense of a diminution in a risk of an adverse outcome as a result of medical advice or medical intervention, should itself be recognised as damage giving rise to a claim in negligence. He submits that whatever may have been thought to have been the position after Hotson the decision of the House of Lords in Fairchild v Glenhaven Funeral Services Ltd [2002] 3WLR 89 permits this court to consider as a matter of policy whether or not such a claim can properly succeed.

  8. Mr. Edis on behalf of the respondent submits that there is no justification for any criticism of the way the judge approached the evidence. As far as the second submission of Mr. Maskrey is concerned, he submits that the appellant should not be permitted to pursue his claim on this basis which was no part of the way in which the matter was put before the judge. The case was argued, he submits, at all times on the basis that the medical evidence before the court was directed essentially to the likelihood of the appellant being cured, had the respondent referred him to a specialist as he should have done. As to the third submission, he submits that it is trite law which cannot be reopened before this court, at least in clinical negligence cases such as the present, that the loss of a chance cannot in itself amount to damage. The court has to assess on the balance of probabilities, what would have been the case but for the alleged negligence. If a determination of that issue results in the chance being evaluated at more than 50%, it becomes, for the purposes of the law, a certainty; if less than 50% the chance is treated as if it would not have occurred.

  9. In order to understand these submissions and the judge’s judgment, it is necessary to consider the course of the proceedings and the evidence before the court in some greater detail.

  10. The particulars of claim were dated the 3rd April 2000. The particulars of pain and injury were pleaded in the following terms:

    There was rapid spread and development of the claimant’s non-Hodgkin’s lymphoma disease at the end of 1995. Had the claimant been referred in about November 1994, the malignancy would have been diagnosed and treatment commenced with the disease having progressed no further than stage 1. The treatment would have been by means of either radiotherapy or CHOP Chemotherapy. There would have been a very high likelihood of cure. The claimant would have been unlikely to have developed any secondary recurrence, and would not have required the further treatment detailed above.

    The prospects of obtaining a cure by the time the treatment was in fact commenced were reduced below 50%. There was a high risk of secondary recurrence, as in fact occurred with the need for further treatment.

    The claimant’s precise prognosis remains uncertain.

  11. This pleading was clearly based on a report from Professor Goldstone, a consultant haematologist, dated the 1st July 1999. In this he described the appellant as suffering by November 1995 from anaplastic Ki-1 High Grade B Cell non-Hodgkin’s lymphoma. It had reached what he described as Stage 1E, that is that it had infiltrated other tissues. His view was that had the appellant been referred to a specialist in November 1994 he would have been “curable either by radiotherapy or by CHOP Chemotherapy with Stage 1 disease and a 10 year chance of survival of 84%.” He described the delay as massively reducing the prospects of obtaining a cure. His conclusion was:

    I think there is a significant possibility that the cumulative delay in the plaintiff’s case has compromised his chance of a cure from over 80% to under 50%.

  12. However it would appear that Professor Goldstone had not at that time taken into account a paper published by Falini et al accepted by the American Society of Hematology (sic) in December 1998 and published by the Society some time in 1999. This suggested that for prognosis purposes it was necessary to divide this particular form of lymphoma into two categories, described as ALK positive and ALK negative. The study on which the paper was based involved a total of 78 patients, of whom 25 were ALK negative. This suggested that those presenting in the ALK negative subgroup had a significantly reduced prospect of survival. The appellant falls into this sub-group. To take account of this material, and further material relating to the development of the appellant’s disease, Professor Goldstone and Dr Bunch, instructed by the respondent, prepared a report containing their joint answers to questions from the solicitors for the appellant and the respondent. Their overall view was:

    We are agreed that the cumulative delay in diagnosis and commencing treatment was associated with the claimant’s clinical stage and performance status and this equates with a poorer prognosis.

  13. In attempting to give percentage chances, it is clear that both doctors had some difficulty because of two uncertainties in particular. The first was the extent to which the Falini data should be relied upon because of the smallness of the sample; and the second was the fact that the appellant had survived by the time of the report, which was prepared in 2001, for three years without relapse which indicated a greater capacity for survival than the statistics themselves might suggest. They agreed, however, that taking account of the Falini paper to the extent that they thought appropriate, the appellant would have had a 55% chance of complete remission as a result of the first course of therapy, but with a 40% chance of relapse. They considered that it was “quite possible” that his individual prognosis had been reduced to less than 50% of what it would have been intrinsically at the outset. His actual prognosis they considered to be dependent now more on the fact that he had survived for three years rather than the statistical likelihood of a cure which was no more than 10-15%.

  14. At trial, only Professor Goldstone gave oral evidence. Having described the nature of the disease and the treatments which the appellant underwent, his evidence substantially consisted of an extrapolation of the figures given in the joint report into what was described as a working example. He took a cohort of 100 ALK negative patients, and making the allowances that the joint report had made for the Falini Paper, he followed them through to a conclusion, applying the probabilities of relapse to those who achieved remission at any stage, and the possibility of remission for those who had relapsed. The assumption was that the cohort consisted of patients with the same stage of disease as that from which the appellant suffered at the time when treatment should have been commenced. Of the 100, 55 would achieve complete remission as a result of chemotherapy. To these must be added four who could have achieved complete remission by some other route. Of the 59 who had achieved by whatever means complete remission, 40% would relapse, leaving 35 who would survive. Of those who relapsed and went on to high dose chemotherapy, six would survive. Of the remainder, only one would survive a further relapse. Thus, out of the cohort of 100, only 42 would be cured.

  15. As to the appellant’s prognosis, he considered that the fact that he had survived for three years since the last relapse suggested that his chances of ten years survival had increased from a statistical 10% to 20-30%.

  16. Counsel for the appellant and the respondent made written submissions to the judge. On the issue with which we are concerned, the appellant’s case was put in the following terms:


    On the evidence in the present case the following “facts” have clearly been established on the evidence, on the balance of probabilities:


    That Mr. Gregg’s disease would not have “up staged”.


    That Mr. Gregg would have had more limited treatment than he in fact underwent;


    That Mr. Gregg is likely to have joined the 55% of cohort of patients who obtained complete remission from such limited treatment, instead of being in the cohort of those who do not obtain such a response and thus require the further treatment that he in fact had i.e. high dose chemotherapy and stem cell transplantation.


    Thus applying the Hotson test, the claimants have established on the balance of probabilities (i.e. better then 50% prospect) that the outcome would have been different. This is a simple matter of past fact. The evidence in relation to these matters is agreed.

  17. The respondent also identified the issue as one of causation. The argument is best summed up in paragraph 4 of the supplementary submissions:

    It is not accepted that the evidence justifies the claim that the claimant has shown that with timeous treatment his prospect of avoiding relapse would have been over 50%. The worked example which was accepted by the claimant’s expert demonstrates that tracing through the (necessarily complex) sequence of events which might have followed earlier referral shows that on the balance of probabilities he would not now even be in a better position than he is. That is the crucial test to be applied. That is equivalent to and not as the claimant says confounded by Hotson. There the court had to decide whether a finding of a 75% chance of a similar result was in fact a finding on causation (of past fact) or one of an estimate of chances. It in fact amounted to a claim that the complainants avascular necrosis was not caused by the delay in treatment. In the present case the finding (which must follow from the evidence) that the claimant would probably have relapsed and that his treatment and progress would have followed the same course is a finding that his tragic history was not caused by any delay but instead by his underlying, very dangerous disease.

  18. The judge’s relevant findings of fact are set out in paragraph 38 of his judgment as follows:


    In the 9 month period between the spring of 1995 and January 1996 when treatment began, the claimant’s condition “upstaged” significantly so that he was less likely to achieve complete remission and had a poorer prognosis as a result. Specifically his chances of avoiding radical high dose chemotherapy, his chance of avoiding a relapse and his chances of ultimate survival were all reduced.


    Taking the claimant as an example of the whole population of anaplastic large cell lymphoma patients but with no adverse prognostic features he would have had a remission chance significantly higher than 55% and the chance of disease free survival of over 50%. For such a patient the addition of the adverse prognostic factors that came to affect him because of the delay, meant his initial chance of remission would have fallen to around 50-55% and his chances of overall survival had moved from over 50% to less than 50%.


    Those examples do not take into account the ALK negative factor. That is not something that can be ignored; but nor does the “immaturity” of the evidence allow for its extrapolation directly into the claimant’s case. The best evidence that Professor Goldstone could give in the whole population but taking into account as he in his clinical experience thought right the ALK negative factor was the worked example he gave; namely of 100 such patients 55 would achieve complete remission as a result of the initial CHOP treatment; 42 would ultimately survive of whom 35 would not have had to undergo high dose chemotherapy with cell stem treatment.


    It is not possible to say that without the adverse prognostic factors caused by the delay the claimant who is ALK negative, would more probably than not have become a disease free survivor, or that he would have avoided relapse and relapse after high dose chemotherapy. He may have done but it is not possible to say. It is possible to say on the basis of Professor Goldstone’s model that he would more probably than not have achieved complete remission with initial CHOP therapy and without high dose chemotherapy with stem cell harvesting.


    His chances of survival now for five years are less than 50%. In fact, twice in his evidence Professor Goldstone said 20-30%, so that in my judgment justifies a finding that he has a 25% chance now of disease free survival for 5 years.

  19. Having reviewed a number of authorities, and in particular Hotson, the judge’s conclusions were as follows:


    The time for evaluating what would have happened in the past is the trial. That is the time, it seems to me, when the fundamental rule is that what happened, and what would have happened, has to be established on the balance of probability. What is so established is taken as certain, what is not so established is ignored. There seems to me to be no sanction on the authorities for the view that some intermediate event between the breach of duty and the trial can change the basis on which proof of causation can be established. That rule is subject to some exceptions, for example where a worthwhile cause of action that has been lost has to be evaluated by reference to the chances of its success: Kitchen v Royal Airforce Association [1958] 1WLR 563; when an opportunity to win a prize is lost: Chaplin v Hicks [1911] 2AC; or; as explained by Stuart Smith LJ in Allied Maples Group v Simmons & Simmons [1995] 1WLR 1602 at 1611A where the claimant’s loss depends on the hypothetical action of a third party. In all those cases either by pragmatic development of categories (such as the chance of gaining employment or promotion) or by the route of saying that the chance itself is to be regarded as something of value, damages may be awarded for loss of a chance, so that a 30% chance that has been lost is compensated by an award of 30% of the benefit that would have accrued had the chance become a reality. In my judgment the authorities do not support the view that a loss of chance for a more favourable outcome from medical treatment is in a category of cases where compensation will be awarded on that basis. It seems to me that the reasoning in Hotson should drive me to the conclusion expressed above.


    The starting point is to prove what the claimant’s position was at the time of the breach of duty. That is a question of past fact. In this case Mr. Gregg had a lymphoma that was at a particular stage of development. But that is not the only past fact. What would have been the course of the disease but for the defendant’s negligence, is also, up to the date of trial, a past fact. I do not think that it is possible to apply the description “hypothetical” to it so as to put it in the quantification rather than the causation category of the inquiry. I do not think that there is a forensic difference between asking what would have happened had the doctor attended (in Bolitho v Hackney Health Authority [1998] AC 232 HL) and what would have been the outcome had the doctor given appropriate treatment. In this case the claimant has failed to prove that it is more probable than not that the outcome for him, looked at as at the date of trial, would have been materially different had he been treated nine months earlier. Since the burden of proof lies on the claimant that means it should be taken as more probable than not, that treated earlier, the claimant would nonetheless have at some stage to undergo high dose chemotherapy with stem cell harvesting, would have relapsed, and would thereafter have a very poor prognosis for survival, albeit one that improves as time goes by. That, it seems to me puts him in the same category as the claimant in Hotson who at the time of the breach of duty had an injury the consequence of which would in any event (on the balance of probabilities, and therefore treated as certain) have meant that the necrosis from which he came to suffer would have been suffered. To give effect to a chance less than 50% that the outcome might have been different would be to give effect also to the 25% chance for the claimant in Hotson that appropriate medical intervention would have avoided necrosis.


    In my judgment, therefore, the claimant has not proved that he has suffered an injury as a result of the delay, because he has not proved it more probable than not that he has.

  20. As to the appellant’s first submission, that the judge wrongly evaluated the evidence of Professor Goldstone, we indicated in the course of argument that we did not consider that the appellant could succeed on this ground. It is true that Professor Goldstone was clearly cautious during his evidence in his approach to the significance of the Falini paper, based as it was on such a small sample of patients. But his worked example, upon which the judge based his conclusions, was clearly founded upon the same assessment of the effect of the Falini paper as had informed the answers he gave together with Dr Bunch in their joint medical report. At no stage in his evidence did he resile from the answers given in that report, other than in expressing the caution to which I have referred. It does not seem to me to be possible to argue, in these circumstances, that the judge was not entitled to come to the conclusions that he did based upon that worked example. I would accordingly reject this submission.

  21. The second and third submissions raise more formidable and controversial issues. The second submission, put more fully, is as follows. There is no doubt that in the present case, the delay in diagnosis caused the tumour to enlarge, invade neighbouring tissue and cause severe pain. It was the enlargement of the tumour which reduced the chances of successfully treating it. This aspect of the plaintiff’s claim was never addressed by the judge. Nonetheless the judge’s findings amply support the submission that the appellant had indeed suffered injury which entitled the appellant to general damages for the pain and suffering which were the physical consequence of the spread of the tumour, an assessment of the extent to which delay resulted in more intensive therefore damaging treatment, an assessment of the increased risk of relapse and the adverse effect on prognosis, involving an assessment of the consequences to the appellant’s expectation of life. The appellant beguilingly submits that these are not issues of causation to be resolved on the balance of probabilities, but of quantification, so that the court is entitled to use the statistical information and provide a proportionate award in the same way as it does in any other case involving uncertainty as to prognosis, such as those involving the risk of osteoarthrosis or the risk of epilepsy.

  22. The respondent submits that the essential question for the court to determine was whether or not the appellant’s tumour was of such a nature that, on the balance of probabilities, the delay had no effect on its outcome. That was a question which had to be determined on the balance of probabilities which is the exercise which the judge carried out in a way which cannot be criticised on the evidence before him. Although the only information available to the judge was the statistical information, the judge was entitled to use it in the way that he did. Mr. Edis rightly points out that sympathy for the appellant should not result in the court compensating him for what would have been the inevitable consequences of his disease. He submits that the appellant’s claim is in essence a claim for loss of a chance, and the courts should not be prepared to entertain such a claim, in clinical negligence actions such as the present, when there is a means of evaluating what would have been the course of the disease as a fact, on the balance of probabilities. He submits that it is not open, certainly to this court, to find that such a claim is available to a plaintiff as a matter of law.

  23. The appellant’s submission is an undisguised attempt to side step the decision of the House of Lords in Hotson. In that case, the House of Lords held that the application of the ordinary burden and standard of proof meant that the damage about which complaint was made, namely the development of avascular necrosis, was caused solely by the accident, and not by the delay in treatment. Although there was damage in terms of pain for which the appellant obtained a small monetary award of damages, that had no relevance, according to Lord Bridge, to the damage about which complaint was made. He said at page 780:

    I would observe at the outset that the damage is referable to the plaintiff’s pain during the five days by which the treatment was delayed in consequence of failure to diagnose the injury correctly, although sufficient to establish the authority’s liability for the tort of negligence, have no relevance to their liability in respect of the avascular necrosis. There was no causal connection between the plaintiff’s physical pain and the development of the necrosis. If the injury had been painless, the plaintiff would have had to establish the necessary causal link between the necrosis and the authority’s breach of duty in order to succeed. It makes no difference that the five days pain gave him the cause of action in respect of an unrelated element of damage.

  24. There may be some debate about whether or not the factual position supports this analysis. And there may well have been underlying pathology resulting from the failure to diagnose which had an effect on the outcome. Be that as it may, the decision of principle made by their Lordships was clearly based upon this assessment of the factual situation. Mr. Edis submits that the judge was correct to conclude that this case required him to evaluate the nature of the tumour before the respondent’s negligence and come to a conclusion as a matter of fact whether it was such that the appellant would inevitably have had the poor prognosis which he now has.

  25. Which of these two approaches is right is the critical question in this case. I have come to the conclusion that Mr. Edis’ argument, which has the apparent attraction of simplicity, overlooks the very real differences between the facts in Hotson as assumed and described by Lord Bridge and the facts in the present case. I accept that on his analysis the fact that the tumour has enlarged and caused pain is not in itself a sufficient distinction. What is different it seems to me, is the all or nothing nature of avascular necrosis in Hotson, by that I mean the fact that the claim proceeded on the basis that avascular necrosis either would or would not have been the result despite timeous intervention. The lymphoma in the present case is different in kind. It was undoubtedly susceptible to treatment. That was agreed. The question was the extent to which the chances of treating it successfully had been reduced. The evidence which the judge used in order to come to his conclusion was that chemotherapy would have given the appellant, but for the negligence, a better than even chance of complete remission with a less than 40% risk of relapse thereafter. Although statistically it may be shown that less than 50% of the cohort with such a tumour would be cured, a result which failed to reflect the fact that on one analysis of the statistics, it could be said that he would probably have been cured, does not seem just. Nor does it seem just when the only real certainty is that the negligence has reduced the appellant’s chances of a cure.

  26. It seems to me, therefore, that Hotson is not an analogous case. It is an appropriate solution in a case where the conclusion of fact can clearly resolve the issue. In the Hotson case, the conclusion of fact which could be reached on the balance of probability as to the state of the hip immediately after the accident resolved the issue. Avascular necrosis was inevitable. In the present case, the tumour was treatable, unlike avascular necrosis which was not. The tumour grew and the chances of successful treatment were reduced. For these reasons I do not consider that Hotson stands in the way of Mr. Maskrey’s submission. The question is whether or not the reduced chance of successful treatment can properly sound in damages.

  27. There is no doubt that in the quantification of loss, which is essentially a jury question, judges are entitled to use statistical chances, whether strictly so called, or roughly assessed, as tools to provide a fair solution in any given case. In Mallett v McMonagle [1970] AC 166 at page 176 Lord Diplock said:

    The role of the court in making an assessment of damages which depends upon its view as to what will be and what would have been is to be contrasted with its ordinary function in civil actions for determining what was. In determining what did happen in the past the court decides on the balance of probabilities. Anything that is more probable than not it treats as certain. But in assessing damages which depend upon its view as to what will happen in the future or what would have happened in the future if something had not happened in the past, the court must make an estimate as to what are the chances that the particular thing will or would have happened and reflect those chances, whether they are more or less than even, in the amount of damages which it awards.

  28. If, as Mr. Maskrey submits, the question is one simply of quantification of loss, there can therefore be no difficulty in accepting the consequences of his second submission. But Mr. Edis submits that in essence that would amount to awarding damages for the loss of a chance in clinical negligence cases, which the courts have not been prepared to do before. This therefore engages Mr. Maskrey’s third submission. Questions of causation and quantification of necessity overlap. If there was clear authority for the proposition that the courts will not give damages for loss of a chance in medical negligence cases then that might affect the question whether or not the court should be prepared to do so by treating the question as one of quantification.

  29. Although Mr. Edis’ submission was based essentially on Hotson, he also relied on the decision of this court in Tahir v Haringey Health Authority [1998] Lloyds LR: Med 104. In that case the plaintiff had claimed that an operation was negligently delayed for 24 hours as a result of which he suffered from paraparesis leaving him with some permanent spasticity of gait and problems of bladder and bowel control. The Deputy Judge held that had this been established, damages would have been £30,000. However, she concluded that the only period of delay which amounted to negligence was two to three hours, and that the additional disability resulting from that period of delay could be properly compensated for by an award of damages of £4,000, simply being a proportionate sum reflecting the extent of the delay. The Court of Appeal held that her findings did not justify the conclusion that that period of delay caused any part of the plaintiff’s loss. In the course of giving the lead judgment, Otton LJ held that the issue was one of causation. As to that he applied the following propositions of law:


    The burden of proof in causation was upon the plaintiff.


    Causation is a question of past fact, to be decided on a balance of probabilities: see Mallett v McMonagle [1970] AC 166.


    If he proves that the negligence was the sole cause or a substantial cause, or that it materially contributed to the damage, he will succeed in full. See Bonnington Castings v Wardlaw [1956] AC 613 and McGhee v National Coal Board [1973] 1 WLR1.


    If he fails to cross this threshold then he fails to recover any damages: see Barnett v Chelsea and Kensington Hospital Management Committee [1969] 1QB 428.


    A plaintiff cannot recover damages for the loss of a chance of a complete or better recovery: see Hotson v East Berkshire District Health Authority [1987] AC 750.

  30. It seems to me, however, that the fifth proposition should be approached with some care. It is true that the House of Lords rejected the conclusion of both the judge at first instance, Simon Brown J, as he then was, and the Court of Appeal, that on the facts the 25% chance of saving the hip but for the delay provided the basis for an award of damages. But both Lord Bridge and Lord Mackay expressly reserved the issue of the principle. Lord Bridge said at page 782:

    The plaintiff’s claim was for damages for physical injury and consequential loss alleged to have been caused by the authority’s breach of their care. In some cases, perhaps particularly medical negligence cases, causation may be so shrouded in mystery that the court can only measure statistical chances. But that is not so here. On the evidence there was a clear conflict as to what had caused the avascular necrosis. The authority’s evidence was that the sole cause was the original traumatic injury to the hip. The plaintiff’s evidence, at its highest, was that the delay in treatment was a material contributory cause. But this was a conflict, like any other about some relevant past event, which the judge could not avoid resolving on the balance of probabilities. Unless the plaintiff proved on the balance of probabilities that the delayed treatment was at least a material contributory cause of the avascular necrosis he failed on the issue of causation and no question of quantification could arise ....

    The upshot is that the appeal must be allowed on the narrow ground that the plaintiff failed to establish a cause of action in respect of the avascular necrosis and its consequences. Your Lordships were invited to approach the appeal more broadly and to decide whether, in a claim for damages for personal injury, it can ever be appropriate, where the cause of the injury is unascertainable and all the plaintiff can show is a statistical chance which is less than even that, but for the defendants breach of duty, he would not have suffered the injury, to award him a proportionate fraction of the full damages appropriate to compensate for the injury as the measure of damages for the lost chance.

    There is a superficially attractive analogy between the principle applied in such cases as Chaplin v Hicks [1911] 2KB 786 (award of damages for breach of contract assessed by reference to the lost chance of securing valuable employment had the contract been performed) and Kitchen v Royal Airforce Association [1958] 1WLR 563 (damages for solicitor’s negligence assessed by reference to the lost chance of prosecuting a successful civil action) and the principle of awarding damages for the lost chance of avoiding personal injury or, in medical negligence cases, for the lost chance of a better medical result which might have been achieved by prompt diagnosis and correct treatment. I think there are formidable difficulties in the way of accepting the analogy but I do not see this appeal as a suitable occasion for reaching a settled conclusion as to whether the analogy can ever be applied.

  31. Lord Mackay said at page 786:

    .... I consider that it would be unwise in the present case to lay it down as a rule that a plaintiff could never succeed by proving a loss of a chance in a medical negligence case ....

  32. In declining to lay down any broad proposition of law in what he described as “this difficult area of the law” at page 789, Lord Mackay was clearly influenced by the decision of the Supreme Court of Washington sitting en banc in Herskovits v Group Health Co-operative of Puget Sound (1983) 664 P 2d 474. That was a similar case to the one with which we are concerned. The claimant suffered from a tumour. When he was first seen, the statistical chance of survival for more than five years was 39%. There was a delay which resulted in the chance of surviving more than five years being 25%. The defendant moved for summary judgment on the basis that the case could not succeed. The majority judgment was given by Dore J. His conclusion was:

    Rarely is it possible to demonstrate to an absolute certainty what would have happened in circumstances that the wrong doer did not allow to come to pass. The law does not in the existing circumstances require the plaintiff to show to a certainty that the patient would have lived had she been hospitalised and operated on promptly.

  33. He therefore concluded that the issue should be left to the jury to decide to what extent the claimant should be compensated. Brachtenbach J dissented on the basis that the statistics were a dangerous tool to use in such circumstances. All that they showed was that it could not be proved that the delay in diagnosis had caused any loss to the claimant. Pearson J agreed with the majority but was concerned about the direct application of statistical evidence in circumstances such as these. Nonetheless he concluded “that the best resolution of the issue before us is to recognise the loss of a less than an even chance as an actionable injury.”

  34. If the difficulty in determining the correct solution in circumstances such as this needs any underlining, courts in Canada have taken a different approach. In Lawson v Laferriere (1991) 78 DLR (4th) 609, a case involving the delayed diagnosis of breast cancer, the court dismissed the claim on the basis that the probability was that the delay would not have affected the outcome which sadly was the death of the patient, in other words applying the Hotson analysis.

  35. Although we have not been referred to any case directly in point in Australia, the High Court of Australia in Chappel v Hart [1999] Lloyd’s Law Reports: Med 223 had to consider the question of whether or not damages should be awarded to a plaintiff who had not been warned of the small risk of an adverse result of surgery which in fact occurred. Kirby J, who agreed with the majority that the plaintiff was entitled to damages, gave a thoughtful judgment in the course of which he said, at page 245:


    Valuing a lost chance

    A further way in which, in some circumstances, the difficulties of causation for a plaintiff are alleviated is by treating the plaintiff’s loss as a “loss of a chance”. In cases in which this approach is permissible or it may allow evaluation of the plaintiff’s loss in terms of comparing the chance of suffering harm (given the breach which has occurred) against those that would have existed (if the breach is hypothesised away). In CES v Superclinics (Aust) Pty Ltd (1995) 38 NSWLR 47at 56 to 57 I indicated my attraction to this approach as a more rational and just way of calculating damages caused by established medical negligence. It is clearly laid down by the authority of this court that, in some circumstances, a plaintiff may recover the value of loss of a chance caused by a wrongdoer’s act or omission. The approach also has some judicial support in the context of medical negligence in England [McGhee v National Coal Board [1973] 1 WLR1] Canada [see e.g. Laforest J dissenting in Lawson v Laferriere (1991) 78 DLR 94th 609 at 610] and the United States. A number of commentators favour this approach because of the failure of orthodox reasoning to do justice to some patients’ losses and because it invites more empirical capitulation of loss, with the use of statistics which might offer outcomes which are more accurate and fair to all concerned. On the other hand, the weight of judicial opinion in England [Hotson] and Canada [Lawson v Laferriere] and some academic writing appears to be critical of the application of loss of a chance throughout cases of medical negligence. In part this is because where medical negligence is alleged, “destiny ....[has] taken its course”, arguably making an analysis by a reference to chance inappropriate or unnecessary in the view of the critics of this approach. Alternatively, the loss of a chance calculation has been criticised on the ground that it would discard common sense, undermine the plaintiff’s onus of proving the case and submit the law to the “paralysis” as statistical abstractions.

  36. It is important to note that in coming to the ultimate conclusion that he did, namely that the plaintiff was entitled to recover damages, Kirby J referred to the speech of Lord Hoffmann in Environment Agency (formerly National Rivers Authority) v Empress Car Co (Abertillery) Ltd [1999] 2 AC 22 at page 29:

    The first point to emphasise is that common sense answers to questions of causation will differ according to the purpose for which the question is asked. Questions of causation often arise for the purpose of attributing responsibility to someone, for example, so as to blame him for something which has happened .... one cannot give a common sense answer to a question of causation for the purpose of attributing responsibility under some rule without knowing the purpose and scope of the rule.

  37. At page 246; Kirby J said:

    The “common sense” which guides courts in this area of discourse supports Mrs. Hart’s recovery. So does the setting of standards which uphold the importance of the legal duty that was breached here.

  38. This court considered the same question as that considered by the High Court of Australia in Chester v Afshar [2002] EWCA Civ 724. From the judgment of the court, which was given by Sir Denis Henry, it is plain that it agreed with the Australian High Court’s reasoning, and in particular the judgment of Kirby J. It considered that in that context, the remarks of Laws LJ in Rahman v Arearose [2001] QB 351 at page 367 were also helpful:

    So in all these cases the real question, is what is the damage for which the defendant under consideration should be held responsible the nature of his duty (here the common law duty of care) is relevant: causation, certainly would be relevant – but it will fall to be reviewed and in truth can only be understood, in the light of the answer to the question, from what kind of harm was it the defendant’s duty to guard the claimant?

  39. From this recitation of authority, I am satisfied that Hotson is not authority for the proposition that the loss of a chance is incapable of forming the basis for an award of damages in medical negligence cases. It is clearly an extremely difficult area of the law which has been grappled with with varying results, as I have indicated, in other jurisdictions. It has also been the subject of a number of academic writings, including an interesting and wide ranging article by Jane Stapleton in The Law Quarterly Review Vol. 104 July 1988 at page 389. This discusses some of the consequences if the law were to recognise that the loss of a chance simpliciter was damage for the purposes of an action. If carried to an extreme, this would open the door to claims for damages in the personal injury field by those who have suffered no injuries save for the statistical possibility of future harm, for example because they have been exposed to asbestos dust in the vicinity of an asbestos factory or asbestos workings, without there being any evidence of adverse effects at the time of the claim. It seems to me that there would be therefore good policy reasons for declining to extend the scope of the tort of negligence to speculative actions such as those, based on the loss of a chance simpliciter.

  40. However, different considerations apply in my view where injury has been occasioned which includes as a necessary consequence the loss of a chance of some benefit. In such a case the just answer to the question of whether or not that lost chance should sound in damages is to consider the nature of the duty of care in question, just as the High Court of Australia did in Chappell v Hart, and this court did in Chester v Afshar. For it is in the context of that duty that the limits to the recoverable damages can most sensibly be defined. In the present case, the respondent’s duty was to exercise such care as would reduce the risk of an undiagnosed cancer spreading and becoming less amenable to treatment. The spread of the cancer and the reduced chances of successful treatment were inextricably intertwined, unlike the plaintiff’s pain and the avascular necrosis in Hotson. It follows that the appellant can only obtain a proper remedy for the breach of the duty in question if the law is prepared to recognise the lost chance as part of the damages to be awarded for the injury which he has undoubtedly sustained.

  41. It will be apparent from the above that I do not consider that this is a case in which it is necessary to determine whether or not the loss of a chance can, of itself, be a permissible head of damage. In the present case the evidence clearly established that the cancer had spread, on the findings of the judge, by reason of the negligence of the respondent. That was all that was necessary to found his claim in negligence. Once that had been established, the question for the court was the extent to which the consequences, which included the reduced prospects of successful treatment could themselves be established as an issue of quantification of damage to be determined on the basis of the trite principle stated by Lord Diplock in Mallett v McGonagle.

  42. It is clear that the practical exercise of assessing the value of the lost chance in any given case will not necessarily be an easy task. As the Supreme Court of Washington recognised in Herskovits v Group Health Co-operative GHC of Puget Sound, statistics may be a dangerous tool in carrying out what is essentially a broad brush exercise namely the determination by the judge of a figure which is in his view fair and just to both the claimant and defendant in any given case. But the starting point must, it seems to me, be that where the injury that has been sustained carries with it a significant reduction in the chance of a benefit which would otherwise have been available, the judge is required to evaluate the value of that lost chance. I say significant, because it seems to me that otherwise it will be too marginal or speculative to amount to an injury capable of being recognised as damage.

  43. I would therefore allow the appeal.

    Lord Justice Mance


  44. HHJ Inglis found,

    • firstly, that, but for the negligent failure to diagnose the appellant’s tumour, the appellant would, statistically, have had a 42% chance of surviving for 5 years from the date of trial in mid-2001; and,

    • secondly, that the failure to diagnose the appellant’s tumour, causing some nine months delay in the start of treatment, reduced his chance of surviving for 5 years from mid-2001 to 25%, i.e. by 17%.

    I agree with Latham LJ that the appellant’s attempt to challenge the evidential basis for this conclusion fails for the reasons he gives.

  45. The claim is now put differently from the way it was put before the judge. Before the judge, the appellant relied upon “the reduced prospect of a good outcome” as the relevant actionable injury. His case was that the outcome would, but for the negligence, probably have been different, in that

    1. the tumour would not have grown (or “upstaged”),

    2. the appellant would have been likely to fall within the 55% cohort of patients who obtain complete remission from limited treatment and

    3. the treatment he required would have been less.

    The appellant’s failure to obtain complete remission under limited treatment was relied upon as “a simple matter of past fact”, satisfying the test of causation in Hotson v East Berkshire Health Authority [1987] AC 750. By the end of the trial at least, the appellant’s case had become that everything thereafter was a matter of assessing prospects, and the amount awarded could be discounted on a percentage basis, to take account of the risk of relapse.

  46. The judge addressed this case, by treating both the appellant’s position at the time of the breach of duty and the question “what would have been the course of the disease but for the Defendant’s negligence .... up to the date of trial” as questions of past fact (paragraph 43). He held that the appellant had “failed to prove that it is more probable than not that the outcome for him, looked at at the time of trial, would have been materially different had he been treated 9 months earlier”. On that basis, he viewed the appellant as being within the same category as the claimant in Hotson “who at the time of the breach of duty had an injury the consequences of which would in any event (on the balance of probabilities and therefore treated as certain) have meant that the necrosis from which he came to suffer would have been suffered”. He dismissed the claim accordingly.

  47. Mr. Maskrey QC now advances the appellant’s case by submissions at three different levels. The primary basis on which he now seeks to distinguish Hotson is by reliance on the development of an enlarged tumour as actionable “injury”. He submits that the appellant’s reduced prospects of survival were consequential upon that injury; and that damages are recoverable here, just as they are for a risk of osteo-arthritis after a negligently inflicted bone injury or a risk of epilepsy after a negligently inflicted head injury. His second submission is that Hotson has not finally closed the door on the recovery of damages for loss of the chance of a better outcome in the medical negligence field; and that (particularly following the House of Lords decisions in Fairchild v Glenhaven Funeral Services Ltd [2002] 3 WLR 89) this court should revisit that area. Thirdly, his notice of appeal and skeleton submit that the respondent’s negligence and the resultant delay made a “material contribution” to the appellant’s suffering and the course of his illness, and to his present poor prognosis, within the meaning of the principles accepted in Bonnington Castings Ltd v Wardlaw [1956] AC 613 and McGhee v National Coal Board [1973] AC 1. However, Mr. Maskrey also makes clear that, in whatever way the appellant’s claim is put, it is restricted to (i) the extra pain and treatment suffered by reason of the development of the enlarged tumour due to the respondents’ negligence plus (ii) the (statistically assessed) diminution of the appellant’s prospects of future survival. These submissions are based on the judge’s findings of fact and I consider that the appellant can be heard on them on this appeal, without injustice to the respondent.

  48. The problems they raise include:

    1. the legal test for the link requiring to be established between the respondent’s negligence and any relevant injury suffered by the appellant,

    2. whether the appellant suffered any and if so what actionable injury by way of, or supporting a consequential claim for, reduced prospects of survival and

    3. the basis of assessment of any damages for any such injury.

  49. As to (a), a claimant must generally establish on a balance of probabilities that his or her injury arose from the defendant’s negligence. Normally, this also means showing that “but for” the defendant’s negligence, the injury would not have been sustained. Fairchild, and Bonnington and McGhee as explained in Fairchild, stand as examples of situations in which the law, for exceptional reasons (particularly related to impossibility of proof), accepts as sufficient a link, not meeting those general tests. For what are no doubt associated reasons, they are cases where, despite the exceptionally relaxed factual link, the claimant was also able to recover the whole of his loss. At all events, no argument was apparently raised in any of these cases that there could or should be apportionment or pro rata recovery, whether according to the risk (perhaps incapable of assessment) that any particular defendant(s) actually caused the claimant’s injury or on any other basis. The development of the law in these cases to meet the needs of cases where it is scientifically impossible to prove that the claimant would probably not have suffered the injury but for the defendant’s negligence does however invite the question whether there may not be other (less extreme) circumstances in which a measurable possibility that the claimant’s outcome would have been better, but for the negligence, should be compensatable as such in damages.

  50. The law has developed a broad distinction in contexts such as the present between the causation of past events and hypothetical questions as to what would have happened (in the past as well as the future). The present case shows that the distinction may not always be easy or clear-cut. But, speaking generally, the actual sequence of events leading to a past event must be determined on a balance of probabilities. The position is more complex in relation to hypothetical questions as to what would have happened (in either the past or the future) “but for” a defendant’s breach of duty. There are certain well-recognised classes of case, in which the law takes account of the prospects (or “chances”) of a favourable or unfavourable outcome developing. This is so when a breach of contract deprives the claimant of prospects or chances to which he was entitled as a matter of contract or by reason of the existence of a duty of care: see e.g. Chaplin v Hicks [1911] 2 KB 786 and Spring v Norwich Union [1995] 2 AC 296. In other cases, the authorities have developed fixed rules to indicate when the relevant factual link falls to be determined on a balance of probabilities or by reference to an evaluation of prospects: see e.g. Allied Maples Group Ltd v Simmons & Simmons [1995] 1 WLR 1602. How far those rules apply or provide any analogy in the context of medical negligence (particularly failure in diagnosis) is an issue in this case. In the medical negligence context, it is however clear, as Mr. Maskrey submits, that, where the negligence is shown on a balance of probabilities to have led to a particular illness or condition, which itself creates a consequential risk of some possible future development or complication (such as osteo-arthritis after a bone injury), damages proportional to that risk will be awarded: see e.g. Hotson, 793H-794A per Lord Ackner; Davies v Taylor [1974] AC 207, 220B per Lord Simon.


  51. Against this background, I start with a more detailed examination of the third level of submissions relating to the cases on “material contribution”. These cases, as explained in Fairchild, show that it is in some circumstances just to allow a claimant to make a full recovery, even though he cannot establish, on a balance of probability, the extent to which his injury would have been avoided, or even that any part of his injury would have been avoided, “but for” the defendant’s negligence. So, in Bonnington an employee had contracted pneumoconiosis as a result of his exposure in the course of one employment to a noxious agent consisting of silica dust. The greater part of such exposure occurred without negligence, but a significant part occurred due to negligence. The effects of such exposure on the risk of his contracting pneumoconiosis were cumulative.

  52. McGhee marked a further step. The plaintiff’s inevitable exposure to coal dust had been negligently prolonged, by the defendant employer’s failure to provide facilities for washing before the employee went home. The plaintiff contracted dermatitis, arising from abrasion of the skin due to the presence of coal dust. The medical evidence left it uncertain whether repeated abrasions were a necessary precondition to the development of the disease or whether it could start from a single abrasion and spread, in which case a multiplicity of abrasions would merely increase the risk of its occurrence. The defendant was held liable on the basis that, in the circumstances, the material increase in the risk of injury caused by its negligence fell to be treated as a material contribution to the claimant’s injury.

  53. The existence of an “evidential gap” which cannot be filled on current scientific knowledge is a critical feature of cases in this category, although there may be scope for argument what degree of scientific uncertainty constitutes such a gap. Cases within the Fairchild principle contrast with cases such as Thompson v Smiths Shiprepairers Ltd [1984] QB 405, where the evidence enabled the judge to identify on a broad basis the extent to which the claimants’ deafness had already arisen during periods of employment before the defendants came under any duty to take effective precautions against deafness and the extent to which impairment of hearing arose subsequently, and to make an apportioned award accordingly. That the courts should, wherever possible, assign and apportion responsibility on a balance of probabilities is clear: see Holtby v Brigham & Cowan (Hull) Ltd [2000] 3 AER 421 (CA); Hatton v Sutherland [2002] EWCA Civ 76; [2002] 2 AER 1, especially at paragraphs 36-44 (CA). However, this is not the same as apportioning liability according to the risk that a particular defendant’s negligence has actually led to an injury.

  54. The boundaries of the line of authority exemplified by Bonnington Castings and McGhee were also examined in Wilsher v Essex Area Health Authority [1987] QB 730 (CA); [1988] AC 1074. The defendant’s hospital had negligently administered excess oxygen, thereby increasing the risk that the infant claimant would suffer RLF, an incurable condition of the retina impairing vision or causing blindness. The claimant suffered RLF, but there were other factors, applicable having regard to his medical history, which could without negligence have led to his RLF. Again, there was an evidential gap, but here the claim failed. The (pre-)existence of potentially contributory factors entirely extraneous to any negligence by the hospital was critical.

  55. In Fairchild the House of Lords examined and approved Bonnington Castings and McGhee. The claimants (C) in Fairchild had each suffered mesothelioma, after each had been negligently exposed to asbestos in the course of successive employments (by A, B, etc.). The risk of mesothelioma increases in proportion to the quantity of asbestos dust and fibres inhaled, but the disease can be caused by a single fibre, or a few fibres or many fibres. While “(5) any cause of C's mesothelioma other than the inhalation of asbestos dust at work [could] be effectively discounted, .... (6) C [could not] (because of the current limits of human science) prove, on the balance of probabilities, that his mesothelioma was the result of his inhaling asbestos dust during his employment by A or during his employment by B or during his employment by A and B taken together” (per Lord Bingham at paragraph 2).

  56. In such circumstances, the House of Lords held, C could recover in full from A and B. A and B could then seek contribution against each other, or any other employer liable in respect of the same damage. The exposure of C by each of A and B to an unjustified risk fell to be regarded as making a material contribution to the C’s disease. The majority of the House treated Bonnington, McGhee and Fairchild itself as instances of cases, in which justice calls for the application of an approach to causation other than the orthodox approach.

  57. Wilsher was regarded as correctly decided in Fairchild. Lords Bingham and Rodger (at respectively paragraphs 22 and 149 and 170, fourth point) approved in this regard the reasoning of Sir Nicholas Browne-Wilkinson V-C (as he was), dissenting in the Court of Appeal in Wilsher. He had identified a relevant distinction between disease resulting from exposure to a single noxious source (or at least “an agency that operated in substantially the same way”), to which the claimant has been exposed (even though the defendant’s negligence only relates to part of the exposure), and disease which may have resulted from “any one of a number of noxious agents” or of different events. Lord Hoffmann did not regard a distinction between a single and numerous noxious agents as principled (cf paragraph 72, and also Mustill LJ in this court in Wilsher at pp.771-772). He preferred to support the distinction between the results of McGhee and Wilsher on policy and pragmatic grounds related to the distinction in subject-matter between an employer’s liability claim and a national health authority’s duty to take medical care (cf paragraphs 69-70).

  58. The circumstances of the present case differ from those which in Fairchild called for a modified approach to the test of causation. First, Bonnington, McGhee and Fairchild were all cases concerned with an indisputable and already sustained injury or illness. Secondly, in those cases the claimant successfully held the defendants 100% liable for his injury or illness. In the present case, the claim is not, at least now, to hold the defendant 100% responsible for the total diminution in life expectancy, but is limited to those consequences (measured by reference to the reduction in prospects of survival) which are attributed to the enlarging of the tumour. Thirdly, the appellant’s primary case has not been that there was or is evidential impossibility along the lines that made it just in Bonnington, McGhee and Fairchild to award full damages, but rather that it is possible to prove relevant injury on the balance of probability, and that the risk of curtailed life expectancy can be measured in percentage terms as a consequence of that injury.

  59. Fourthly, whether one adopts the grounds on which the majority in Fairchild distinguished Wilsher or the more explicitly policy-based grounds of Lord Hoffmann, a doctor’s negligent failure to diagnose a pre-existing disease differs significantly from the fault of an employer who negligently brings about conditions in which an employee is or continues to be exposed to a noxious agent to which either the employee himself without negligence or some other employer by negligence has exposed the employee. The defendants in Bonnington, McGhee and Fairchild were, either individually or together with other negligent employers, the original source of the noxious agents which caused the plaintiffs’ injuries in those cases. Lord Bingham’s formulation of the issue is Fairchild is confined to such situations. Lord Rodger at paragraph 170 expressly left open other cases “where the other possible source of injury is a similar but lawful act or omission of someone else or a natural occurrence”.


  60. I come back therefore to the first two levels of submission made on the appellant’s behalf. The more radical submission is that the gist of the complaint can be presented as being for loss of a chance. No-one suggests that the statistical diminution in life expectancy relied upon by the appellant can sensibly be treated as minimal. Anyone would regard a 42% chance of survival as, at any rate, a fair, though not of course a good, chance of survival, whereas a 25% prospect, although the drop is “only” 17%, falls into the category of a poor chance of survival. On the other hand, quite apart from the respondent’s negligence, the appellant had cancer, which at least as a matter of statistical probability meant that he was anyway going to suffer a curtailed life, so that there cannot be a simple attribution, to the respondent’s negligence, of all the devastating effects on him of his cancer. The extra pain, suffering and distress resulting from the negligence should however be recoverable in damages.

  61. The respondent responds to this submission by relying upon Hotson as excluding any claim for loss of a chance. Although Mr. Maskrey submits that Fairchild enables the courts to take a more generous attitude to claims for loss of a chance, the subject was not debated in Fairchild - see per Lord Bingham at paragraph 34. Fairchild may, academically, invite the question whether there are less extreme circumstances where difficulties of proof may lead to a widening of the field for recovery of damages for loss of prospects, but the speeches do not themselves either identify or encourage this prospect.

  62. In Hotson the claimant when aged 13 fell from a rope and sustained an acute traumatic fracture of the left femoral epiphysis which the defendant hospital negligently failed to diagnose in time. The epiphysis (the cartilage separating the bony head from the neck of the femur in a growing body) died, leaving the claimant with a permanent disability (avascular necrosis). Simon Brown J had held that there was only a 25% chance that the claimant’s fall had left intact sufficient blood vessels to keep the epiphysis alive; and that, had sufficient blood vessels remained alive, the hospital’s negligent failure to diagnose the injury, allowing continuing bleeding in the area, would have blocked them, so killing the epiphysis. On this basis, the House of Lords (differing from both the judge and the Court of Appeal) held that the claim failed.

  63. The case was treated as raising a “simple” and “inescapable” question as to the causation of a past fact, to be decided upon the balance of probabilities: see pp.782C-D, 784G-H, 785C-E and 792C-D and G-H per Lords Bridge, Mackay and Ackner, with all of whom Lords Brandon and Goff agreed at pp.783D and 794B-C. The judge having held as a matter of probability that the blood vessels supplying the epiphysis were already blocked (so that the epiphysis was in effect already dead and the disability inevitable), the 25% prospect that he might (in effect) be wrong about this was irrelevant. Lord Bridge did not regard the case as an appropriate one in which to decide whether it could ever be appropriate, in a claim for personal injuries, where the cause of injury was “unascertainable” and all the plaintiff could show was a less than even statistical chance that, but for the defendant’s negligence, he would not have suffered the injury, to award a proportionate fraction of the full damages to compensate for the lost chance. He saw “formidable difficulties” in accepting the analogy of the principle applied in cases such as Chaplin v Hicks [1911] 2 KB 786 and Kitchen v Royal Air Force Association [1958] 1 WLR 563; and he went on specifically to say that:

    if the plaintiff had proved as a balance of probabilities that the authority’s negligent failure to diagnose and treat his injury had materially contributed to the development of avascular necrosis, I know of no principle of English law which would have entitled the authority to a discount form the full measure of damage to reflect the chance that, even given prompt treatment, avascular necrosis might well still have developed. The decisions in Bonnington .... and McGhee .... give no support to such a view.

  64. Lord Mackay at p.786A-B and F-G also drew attention to the decision in McGhee, when observing that “it would be unwise in the present appeal to lay it down as a rule that a plaintiff could never succeed by proving loss of a chance in a medical negligence case”. On the other hand, as he observed at p.786G, the House in McGhee did not treat the chance as an asset. Rather it treated it as establishing a sufficient, although unorthodox, causal link between the negligence shown and damage of a traditional nature, consisting of the disease contracted, and so awarded full damages for the latter. Lord Ackner at p.793H-794A distinguished the issue of (past) causation which he saw the case as raising from the assessment “where the cause of action has been established .... of that part of the plaintiff’s loss where the future is uncertain [involving] the evaluation of that uncertainty”. He gave as an example the “classic case” of the evaluation of the chance of epilepsy occurring after a brain injury. Lord Bridge pointed out at p.782C-D that the claimant’s suffering of extra and actionable pain during the five days by which treatment was negligently delayed was irrelevant to the liability for avascular necrosis, since there was no causal connection between the pain and the development of the necrosis.

  65. These passages in Hotson highlight the need to consider whether the particular injury was sufficiently linked with the duty broken. However, it cannot be necessary for a cause of action to exist prior to or independently of such injury, or there would be a distinction between contract and tort in a case where the necrosis could be shown as a matter of probability to result from negligence without any other injury. It cannot make all the difference whether such a claim is put in contract or tort, or is against a part of the National Health Service or against a private, contracting hospital.

  66. The crux of the House of Lords’ reasoning in Hotson is the treatment of the relevant issue of causation as one of past fact, to which the balance of probabilities test applied. The House also noted, without analysing further, the distinction between “causation” and “evaluation” or quantification. I do not regard these as labels identifying a distinction which is always either clear-cut or self-evident. Simon Brown J observed at first instance in Hotson [1985] 1 WLR 1036, 1048C-D that how the distinction is drawn in any particular case may well depend on policy considerations. The observation remains in my opinion true despite and perhaps because of the ultimate result in that case.

  67. Thus, in Davies v Taylor [1974] AC 207 the House was concerned with a Fatal Accidents Act claim in respect of an alleged dependency, by a deserting widow who asserted that there was a prospect of reconciliation. The House defined the test of dependency in terms of “a reasonable expectation of pecuniary benefit from the deceased”. Applying this test, the House held that, although the answer to the particular case depended essentially upon the claimant widow’s own position, intentions and evidence, it fell to be assessed, not on a balance of probabilities, but by assessing the prospect that she might have changed her mind and have returned to her husband, who wanted her back. (Nonetheless, it was held on the facts that there was no such prospect.)

  68. In relation to the distinction between matters to be determined on a balance of probability and matters to be determined by reference to the prospects or chances of their occurring, Lord Reid spoke in apparently general terms at p.212H-213C, where he said that, although the question whether a certain event did or did not happen is a matter of past fact, to be determined on the balance of probabilities, the hypothetical question whether the widow would or would not have returned to her husband was to be viewed as a question of future fact:

    But here we are not and could not be seeking a decision either that the wife would or that she would not have returned to her husband. You can prove that a past event happened, but you cannot prove that a future event will happen and I do not think that the law is so foolish as to suppose that you can. All that you can do is evaluate the chance.

    The husband had been killed in a road accident on 14th August 1968, the case had been tried on 16th October 1970 and it came before the House of Lords in 1972. Any reconciliation which would, but for the death, have occurred would presumably have occurred well in advance even of October 1970. Yet the hypothetical question whether it would have occurred was treated as, or as analogous to, one of future fact. A similar distinction to that drawn by Lord Reid had been drawn in the passage which Latham LJ has set out from Lord Diplock’s speech in Mallett v Monagle [1970] AC 166, 176E-G (again a Fatal Accidents Act case but a passage in general terms) and by Lord Browne-Wilkinson in Bolitho v City and Hackney H.A. [1998] AC 232, 239G.

  69. In the light of all these passages, it is clear that the judge in the present case fell too easily into a flawed analysis based on a simple distinction between past and future facts, derived solely by reference to the (fortuitous) date when trial happened to occur. Not only the fifth, but also the second proposition stated in Tahir v Haringey Health Authority [1998] Ll..L.R: Med. 104 must in my opinion be approached with caution, since they fail to distinguish between any of the different categories of causative problem.

  70. When the court will look at probabilities and when at prospects came under closely scrutiny in Allied Maples Group Ltd v Simmons & Simmons [1995] 1 WLR 1602 (CA). Again, the distinction was described as being between causation and quantification – see at pp.1609H and 1614C-D per Stuart-Smith LJ and at p.1619 per Hobhouse LJ. However, what the court did in Allied Maples was to recognise three different categories, each governed by its own fixed rule:

    1. Where the negligence consists in some positive act or omission, causation is a matter of historical fact, so that the court has, for example, to determine on the balance of probability whether the defendant’s careless driving caused the claimant’s loss consisting of his broken leg. Quantification may however depend upon future uncertain events, such as whether and to what extent the claimant will suffer osteo-arthritis, or will continue to earn at the same rate until retirement or whether, but for the accident, he might have been promoted.

    2. Where the negligence consists of an omission, for example to provide proper equipment or instructions or advice, causation depends not on historical fact, but on the answer to a hypothetical question, what would the claimant have done, if given the correct equipment, instructions or advice. The claimant must establish this on a balance of probability.

    3. Where on the other hand the claimant’s loss depends on the hypothetical action of a third party, either in addition to or independently of action by the claimant, the claimant need only show that there was a substantial chance of the third party acting in a way which would have avoided the loss, the evaluation of such chance being a question of quantification of damages. In support of the third rule, Stuart-Smith LJ cited inter alia Chaplin v Hicks and Kitchen.

  71. It may be that even the well-known rules in Allied Maples will require further elaboration. It has for example been suggested in McGregor on Damages (18th ed.) paragraph 381 that the first category in Allied Maples should embrace omissions, while the second category should embrace all circumstances in which the alleged loss depends on a claimant’s own conduct. It may also be asked how the categories and rules identified in Allied Maples fit with the reasoning in Davies v Taylor, where the claimant’s own hypothetical conduct was assessed in terms of prospects, not probabilities. The readiest explanation of the approach taken in Davies v Taylor is, however, that the particular statutory test under the Fatal Accidents Acts (“reasonable expectation”) required it. Thirdly, I would note that, although Lord Reid in Davies v Taylor spoke of a general impossibility of proof of hypothetical events, it seems difficult to explain the rationale of the distinction between the second and third categories and the rules developed to govern them in such absolute terms; rather it must, I would think, be the pragmatic consideration that a claimant may be expected to adduce persuasive evidence about his own conduct (even though hypothetical), whereas proof of a third party’s hypothetical conduct may often be more difficult to adduce.

  72. Where then does the present case stand? In my view, it is not directly covered by Hotson. This, I think, follows from the way in which the House of Lords analysed Simon Brown J’s findings as trial judge. Simon Brown J said at [1985] 1 WLR 1036, 1041:


    Even had the health authority correctly diagnosed and treated the plaintiff on 26 April there is a high probability, which I assess as a 75% risk, that the plaintiff’s injury would have followed the same course as it in fact has, that is he would have developed avascular necrosis of the whole femoral head with all the same adverse consequences as have already occurred and with all the same adverse future prospects.




    The reason why the delay sealed the plaintiff’s fate was because it allowed the pressure caused by haemarthrosis – the bleeding of ruptured blood vessels into the joint – to compress and block the intact but remaining blood vessels with the result that even had the fall left intact sufficient blood vessels to keep the epiphysis alive (which, as finding (1) makes clear, I think possible but improbable) such blood vessels would have become occluded and ineffective for this purpose.

    The House analysed these passages as a finding that a past state of facts had occurred, prior to any negligence, which made it (in Lord Ackner’s word) inevitable that the epiphysis would die, with all the adverse past and future consequences that the claimant had in fact suffered. If Simon Brown J’s reference to a 75% probability or risk related simply to his confidence in reaching a finding of past fact, then on conventional principles the court was bound to treat that past fact as having occurred, without any diminution of recovery on account of the limited evidential uncertainty about the correctness of the finding. Further, if the future consequences of that past fact were inevitable, with or without the subsequent negligence, there could be no question of assessing the prospects of their occurring on any percentage basis (other than 100%).

  73. How then should one view the judge’s findings in the present case? Like the evidence before him, his findings concentrate on the appellant’s prospects of a more favourable outcome, expressed statistically. In other words, they deal with the matter in similar terms to finding (1) in Simon Brown J’s judgment in Hotson, without the elucidation in finding (4), which enabled the House of Lords to transmute that finding into one of past fact. So, in the present case, Mr. Maskrey asserts that the only relevant past fact is that the respondent’s medical negligence led to the appellant’s tumour growing to an increased size and causing unnecessary pain and giving the appellant a cause of action for that at least. In all other respects, Mr. Maskrey submits, the court is concerned here with comparisons involving future and future hypothetical fact:

    1. between what has happened to date and what would have happened, but for the negligence and

    2. between what is in the future likely to happen and what would have happened in the future, but for the negligence.

    It is comparison (B), involving the critical area of life expectancy, that is central to this appeal; and Mr. Maskrey submits that all that can be said about it is that the appellant had statistically a better chance of surviving longer prior to the negligence that he does as a result of the negligence. This submission corresponds, as I see it, with the reasoning in Judge v Huntingdon Health Authority [1995] 6 Med LR 223 (a case with considerable factual similarity to the present), where the judge moved from findings of fact regarding the factual extent of the tumour at the date when it should have been diagnosed to an assessment of the prospects of a cure, bearing in mind such findings.

  74. Attractively put as Mr. Maskrey’s submission is, I do not consider that matters are so simple. Although the position was expressed before and by the judge in terms of statistical prospects, it cannot be a pure matter of future chance or future conduct (of the appellant or anyone else) whether someone falls within the category of the 58% category who are unfortunate or the 42% who are fortunate. While the evidence does not appear to have addressed this aspect specifically (presumably because of the way the case was presented), the most obvious influencing factors are, one would suppose, internal to the claimant at the time of the negligence, however unknown or unknowable they may be; and they consist of the precise characteristics, development and spread of the cancerous cells at the time of the negligence as well as the claimant’s precise physical characteristics and resistance. Other influencing factors may very well include subsequent events such as the particular medical treatment received, the patient’s subsequent life-style and his or her, or indeed others’, reaction to the stress inevitably incurred. All these factors may combine in bringing a case within either the 58% or the 42% band. Matters of (in this case at least, unidentified) past fact may therefore be very important, and there is no basis, at least on the evidence, for distinguishing their influence from any other influence. So, although the case may perhaps be supposed to be closer to Hotson than to any of the categories (all addressing purely hypothetical or future conduct) identified in Allied Maples, I do not think that it can be said to be covered by any previously identified category.

  75. Lord Bridge in Hotson was ready to accept that the approach taken in Hotson would not necessarily be appropriate in cases where it could be said that causation was scientifically impossible to ascertain: He said at p.782B:

    In some cases, perhaps particularly medical negligence cases, causation may be so shrouded in mystery that the court can only measure statistical chances. But that was not so here. On the evidence there was a clear conflict as to what had caused the avascular necrosis. The authority’s evidence was that the sole cause was the original traumatic injury to the hip. The plaintiff’s evidence, at its highest, was that the delay in treatment was a contributory cause. That was a conflict, like any other about some relevant past event, which the judge could not avoid resolving on a balance of probabilities.

    The (partial) report of Simon Brown J’s judgment at first instance in Hotson indicates that he heard detailed evidence about the likely chain of events involved in the fall and their consequences for the various blood vessels feeding the epiphysis (see in particular [1985] 1 WLR 1036, 1041C-E). His 75% confidence did not derive from statistics, but from detailed evidence particular to the fall and the plaintiff.

  76. The evidence and the judge’s findings of fact suggest that on the information before the court in the present case all that the experts could tell the court was about the general statistical pattern in the case of persons suffering from the type and size of tumour in issue in this case. It may be suggested that, since statistics are a questionable basis for assessing a balance of probability, the only way in which due responsibility can be sensibly assigned in the case of negligence such as the present is by reference to percentage prospects. The problems presented by purely statistical evidence are illustrated by the hypothesis of the traffic accident involving a cab in the town where there were only two cab companies, one with three blue cabs, the other with only one yellow cab: see per Brachtenbach J in Herskovits v Group Health Cooperative of Pugin Sound (1983) 664 P.2d 474, summarised by Lord Mackay in Hotson, at p.789B-D. This has elicited lively academic discussion: see C. Nesson, The Evidence or the Event? On Judicial Proof and the Acceptability of Verdicts (1985) Harvard. L.R. 1357 and D. Shaviro, Statistical Probability Evidence and the Appearance of Justice (1989) Harvard L.R. 530. The former article also considers the cases of Summers v Tice 199 P 2d (1948) and Sindell v Abbott Laboratories 26 Cal. 3d 588 (1980), mentioned by Lord Bingham in Fairchild at paragraphs 27 and 29 (although the solution in Fairchild of holding the defendants each 100% liable contrasts with the solution of statistical apportionment adopted in Sindell). Of course, in most cases there would be (or the claimant would be expected to adduce) further facts which would put further flesh on the exercise of assessing probability and so avoid this conundrum. This last point is emphasised in the powerful and comprehensive judgment of Gonthier J in Lawson v Lafferrière (1991) 78 DLR 609, 654d-g and 656d-f.

  77. In all these circumstances, I do not consider that the present case is on all fours with Hotson. The first reason is, in summary, that it is (at lowest) not clear that the appellant’s statistical prospects do depend on pure matters of past fact. Past fact, especially the precise, though unknown, characteristics of the cancer and the appellant, may be very important, but these are unlikely to be the only ingredients in the prospects. Secondly, there may anyway be a distinction between matters capable of being the subject of particular evidence, and matters which can only be addressed statistically. As I see it therefore, the present appeal raises a question not directly covered by authority. What is the appropriate approach in such a case? In particular, can and should the injury, or the “gist” of the complaint, be treated as being the diminution in prospects? Or as the enlarging of the tumour, in which case can the diminution in prospects be treated as a parasitical consequence, like a risk of epilepsy after a brain injury?

  78. Like Latham LJ and Simon Brown LJ, I have read with interest articles by (as she is now) Professor Jane Stapleton entitled The Gist of Negligence 104 LQR 213 and 389. In her submission the real unanswered issue, which should have been more directly addressed even in Hotson, is whether actionable injury can consist in the loss of a chance, rather than the necrosis of the epiphysis. However, it is clear that the reason that the House did not think it necessary to address directly the question when if ever damages may be recovered for loss of a chance was that it considered that issue to be pre-empted by the finding of fact which it identified, namely that the necrosis was inevitable prior to any negligence.

  79. That being the House’s approach in Hotson, the question arises whether a different principle applies, when and if no relevant finding of fact has been or, very possibly, could be made regarding a claimant’s position prior to a negligent failure to diagnose cancer, other than in terms of statistical prospects. While as the House saw it, this question did not arise for decision in Hotson, the speeches lend little encouragement to the introduction of any such different principle; cf Lord Bridge’s, Lord Mackay’s and Lord Ackner’s observations cited above.

  80. Firstly, I agree with Latham LJ that to allow a claim for worsened statistical prospects in the future, as a result of some past act of negligence, without proof of any past injury, would, clearly, be to go too far. Statistically, it may be possible to show that persons living in an area where some mining or industrial process has in the past been negligently conducted, are at greater statistical risk of succumbing to a disease such as pneumonicosis, asbestosis or mesothelioma. I do not think that English law can or should contemplate what would be essentially speculative actions for loss of life expectancy, based on no more than general statistical evidence.

  81. Even where a claimant has suffered physical injury (e.g. following a doctor’s failure to attend or operate upon him or an unsuccessful operation), and it can be shown that there has been some negligence which might have contributed to that injury, I do not think that a claimant can succeed simply on the basis that there is a chance that the negligence caused the injury. Lords Bridge and Ackner were at pains in Hotson to point out the absence of any such principle. The case of Bolitho v Hackney Health Authority [1998] AC 232, to which the judge referred in his judgment in this case, indicates that a claimant must, in a case of a doctor’s negligent failure to attend upon him, prove that the doctor, if he had attended, either would have taken steps which would have avoided the injury or, if he would not have done so, that any reasonable doctor (applying the Bolam test) should have taken such steps. These are tests based on probability. It would not be sensible to adopt different tests in such a case as the present. Further, if it was open to an English claimant to claim for loss of a chance in a case such as the present, then a claimant in the position of the claimant in Wilsher ought to be able to recover damages apportioned to reflect the percentage prospect that the negligence played a part, rather than nothing at all. There is no suggestion of this, even in argument, in any prior authority put before us, and it would represent a considerable increase in exposure.

  82. Ultimately, there is a policy choice whether to treat a case such as the present as falling within a category analogous to Hotson or to Bolitho, and so to require the claimant to prove that the medical negligence has, in probability, reduced his life expectancy; or whether to permit recovery for the prospects (probably derived from mixed, though unknown factors, including past facts and future circumstances) that it will be reduced. This is a question of policy to which, as Lord Mackay highlighted in Hotson and as Latham LJ highlights in his judgment, judges in different countries have reacted in every or almost every conceivable way.

  83. In Herskovits (above), where Dore J (in order to avoid a “blanket release” for doctors in cases where there was a less than 50% prospect) favoured recovery in full (an approach which Lord Mackay compared to that in McGhee), Brachtenbach J favoured the conventional balance of probability test and Pearson J considered that damages should be recoverable in the percentage of the lost chance. In Seyfert v Burnaby Hospital Society (1986) 27 DLR (4th) 96 (British Columbia Supreme Court), McEachern CJSC awarded 25% damages for loss of that prospect of avoiding a colostomy, following a negligent failure promptly to diagnose the extent of a stab wound.

  84. But in Lawson v Laferrière the Supreme Court of Canada (by a majority of six to one) refused to adopt a theory of liability for loss of chance in a medical negligence context, where death had actually occurred some time after a doctor’s negligent failure promptly to inform the patient that she had cancer. The doctor had treated the cancer as he thought fit, without informing his patient of it, and no criticism was made of the actual treatment. The argument, accepted in the court below, was that the claimant had been deprived of the opportunity of seeking further treatment by an alternative method which might have benefited her. Gonthier J’s judgment for the majority not only contains a comprehensive examination of principles and authority not just with reference to France, Belgium and Quebec; it also contains at pp.652-659 a critical analysis, in general terms, of the appropriateness or otherwise of an approach to causation in medical negligence cases in terms of loss of chance. Gonthier J identifies among the relevant considerations the exceptional nature of the cases in which damage can only be understood in purely probabilistic or statistical terms, the desirability of a unified basis of responsibility for medical negligence, covering acts and omissions, and his reluctance to treat “particular medical conditions as the equivalent of diffuse elements of pure chance, analogous to the non-specific factors of fate or fortune which influence the outcome of a lottery”. In the result, Gonthier J affirmed the role that statistical evidence has to play, with all other evidence, in the determination of causation as a matter of probability (leading to the recovery of the full value of the actual damage), and rejected the claimant’s case that the loss of a chance should be compensatable in damages measured according to the level of probability. It is right that Gonthier J also pointed out that frequently in the medical context (as in Lawson v Laferrière itself) the damage has already occurred, manifesting itself in sickness or death. But I do not understand him to have been suggesting this as the basis for a critical distinction in approach.

  85. While considering the present appeal, I have seen the force in the appellant’s submission (which to some extent echoes Dore J’s words in Herskovits) that both the nature and the purpose of the duty broken by the respondent doctor militate in favour of the recognition of responsibility, at least to the extent of any statistically measurable diminution of prospects: cf per Laws LJ in the passage in Rahman v Arearose Ltd [2001] QB 351, 367-368, cited with approval by Lord Bingham in Fairchild, at paragraph 12. On the other hand, the respondent cannot be equated with an employer who is the actual source of a noxious agent; and I think that to accept the appellant’s suggested approach in the present (very common) category of medical negligence case, involving failure to diagnose, would both open a considerable gate to claims based on percentages, and create a new category of case which would be difficult to distinguish in practice from other common cases of medical negligence (e.g. the Bolitho type case or cases of negligence in the course of unsuccessful medical treatment). If one looks at the matter as one of policy, I have come to the conclusion that the considerations in favour of an approach based on probabilities outweigh the argument that negligence that may in the future be causative of some injury should in cases like the present generally attract an award of damages according to the risk that such injury may emerge and it may have been caused by the negligence.


  86. The last question is whether the appellant can rely on the enlarged tumour as an “injury” enabling him to recover not merely damages for extra pain and treatment, but also consequential damages for diminished prospects of a full life. The enlarged tumour is a clear physical consequence of the respondent’s negligence. The 17% diminution in prospects of a normal life expectancy is claimed as a consequential or parasitical loss.

  87. Where there has been a negligently inflicted injury, it is, as I have said, common ground that damages will be recoverable for any increased risk of epilepsy or osteo-arthritis to which that injury may have given rise. In such cases negligence causing an injury has also introduced a distinct new risk, which the court has to evaluate. The prior risk of epilepsy or osteo-arthritis will commonly have been nil. However, this is not necessarily so. There is authority which considers the position if some background risk anyway exists. In Smith v Leech Brain and Co Ltd [1962] 2 QB 405 a pre-disposition to cancer was taken into account in diminution of damages otherwise recoverable for negligently inflicted skin cancer. The basis for this decision was that, but for the negligence, the deceased would not have succumbed to cancer at the time when he did, although he might have succumbed (because of his pre-disposition) at some later stage in his life. In such a case, there is no doubt both that the negligence caused the cancer and that the cancer curtailed the deceased’s life. That is sufficient to establish causation. So too was it in Chester v Afshar [2002] 3 AER 552, especially paragraphs 41-45, where this court held (following Chappell v Hart [1999] Ll.Med. Neg. 223; Aust. HC) that causation in respect of paralysis suffered during an operation was established by showing that, if the relevant surgeon had warned the claimant of the small risk of such misfortune, the claimant would have declined to have the operation at least for the time being.

  88. In such circumstances, there is no problem of causation similar to that identified in Hotson (where the claimant would anyway have suffered the same necrosis at the same time as he did). In assessing loss after causation of the relevant injury, including curtailment of life expectancy, has been shown, it is necessary to take into account the deceased’s probable life expectancy, state of health, career and earnings quite apart from any accident or injury. To take into account in this context a risk that a claimant’s life expectancy might, having regard to his pre-disposition, anyway have been affected by cancer or a later operation at a later stage of his career is no different from taking into account a risk that he might not have retained his job or level of earnings for the rest of his working life, or indeed from taking into account the possibility of promotion and increased earnings. When and how far this is an exercise to be undertaken in terms of risk assessment or probability does not require further consideration here, although it may perhaps arise in some other case (cf the references in Chester v Ashfar, paragraphs 41-45, to the potential issue as being whether it was “more likely than not” that the same risk would eventuate during any future operation, making the case an acceleration case, as well as the comments in McGregor on Damages (2d Suppl. to 16th Ed., para.385A) on Doyle v Wallace [1998] PIQR Q146).

  89. I do not therefore consider that the appellant’s revised way of advancing the claim can be accepted. The head of loss in respect of which the claimant is seeking to claim is diminution of his life expectancy. That represents a substantive head of claim, for which the claimant could recover if he could show, as a matter of probability, that the respondents’ negligence had caused it. If damages cannot be recovered for that as such, it is because the appellant cannot show (as I have concluded that he would have to) that he was not already going to suffer that head of loss, independently of the negligence. It is not possible to change the starting point to the enlarged tumour and to ask the court to assess the prospects that this made any difference to the claimant’s life expectancy. I would add that, if that were permissible, then

    1. it would seem to me that the claimant in Hotson should equally have been able to rely upon the completion of the process of occlusion of the blood vessels supplying the epiphysis as an injury, from which the (subsequent) necrosis of his epiphysis might (although only with 25% probability) have resulted; and

    2. whether a claimant could point to the development of some adverse physical condition or deterioration, in the course of events leading to an unsuccessful outcome to treatment or to an operation, could be expected to become the critical issue in many medical negligence cases.


  90. Not without hesitation, since I have found this a difficult case and my mind fluctuated on some of the points it raises, I would therefore dismiss the present appeal.

    Lord Justice Simon Brown

  91. I have the greatest sympathy both for the appellant and for his claim. As a result of his doctor’s negligence in failing to refer him for a specialist opinion his treatment was delayed by nine months and his chance of survival (i.e. his expectation of a cure in the sense explained by Latham LJ) was thereby reduced from 42% to 25%.

  92. In concluding that the appellant is on those basic facts entitled to recover damages, to include an award for the adverse effect of the respondent’s negligence on the appellant’s prognosis and expectation of life, Latham LJ has felt able to distinguish the facts of this case from those of Hotson v East Berkshire Area Health Authority [1987] AC 750 and to hold accordingly that this claim is not barred by the House of Lords decision in that case.

  93. I wish I could agree, but regretfully I cannot. My difficulty is in finding any distinction in principle between the physical effects of the delay, consequent upon the doctor’s negligence, in diagnosing the patient’s condition in each of the two cases. Here the delay caused the lymphoma to spread and become less treatable, giving rise to severe pain and the need for more intensive treatment. In Hotson it allowed the pressure caused by the bleeding of ruptured blood vessels into the joint to block whatever blood vessels had remained intact after the claimant’s fall (thereby denying him the 25% chance he had of avoiding a avascular necrosis) and caused five days of pain for which on any view he fell to be compensated.

  94. In Hotson at first instance - [1985] 1 WLR 1036, 1045-1046 - I said this:

    Time and time again courts evaluate past and future medical risks and award damages based on an assessment of the likelihood (a) of some adverse medical condition, like epilepsy or osteo-arthritis, developing consequent on the injury, or (b) that some pre-existing, perhaps degenerative, condition would in any event have manifested itself so as to cause the same or at any rate some lesser degree of disability as has been occasioned by injury. There is really no significant difference between that exercise and what the court is being invited by the plaintiff to do in the instant case. That is all the clearer when one recognises that on any view of the facts here the plaintiff has established his cause of action by proving damage to the extent of five days’ pain and suffering, a claim which cannot itself be ignored as de minimis. Even, therefore, if it were not permissible to compensate the plaintiff only for the loss of a chance, such head of loss could properly attach to the undisputed claim, immeasurably larger though the former happens to be. But this very point underlines how unsatisfactory it would be to suppose that the case should turn entirely on whether there is any directly provable injury, however slight. That will itself often be a matter of chance. If for instance a doctor negligently failed to diagnose cancer and there is proved to be a 25 per cent chance that had it been earlier diagnosed and treated it would have been cured but now cannot, there would be no clearly demonstrable injury upon which to hang a parasitical claim for the loss of a chance. In my judgment, therefore, it is sufficient to found a good claim in law that the plaintiff should establish that the defendant’s breach of duty has lost him a substantial chance of being better off. I should add that whether a chance significantly less that 25 per cent would be regarded as substantial for this purpose may be doubted: the courts will not wish to encourage professional negligence claims of a speculative nature.

  95. That approach, although substantially supported by the Court of Appeal, was, of course, rejected by the House of Lords and to their decision I shall shortly return. For present purposes I note only that I was there supposing that a claimant in a cancer case would have more, not less, difficulty than Mr. Hotson in advancing any claim other than upon the basis of loss of a chance. If, therefore, as I believe, this claimant is in no stronger position than was Mr. Hotson, the question arises as to whether he can nevertheless recover for his loss of a chance of surviving his cancerous growth having regard to the law as it has been developed, in particular following the recent decision of the House of Lords in Fairchild v Glenhaven Funeral Services Ltd [2002] 3 WLR 89, and bearing in mind that, ostensibly at least, the House of Lords in Hotson left open the possibility of a plaintiff in a medical negligence case succeeding upon proof only of the loss of a chance.

  96. I refer to the possibility of such a claim being only “ostensibly” left open by the House of Lords in Hotson - see in particular Lord Bridge at 782-783 and Lord Mackay at 786 - because it is difficult to understand how, consistently with the central basis of their Lordships’ decision, a claim in this context for loss of a chance could ever succeed. As was demonstrated in a penetrating article on the point written by Jane Stapleton (now a Professor) in The Law Quarterly Review (Vol 104 p389) the all-important question raised in Hotson was whether the damage forming the gist of the action in negligence could be reformulated in terms not of the claimant’s adverse medical condition but of the loss of a chance of averting it. Let me quote two paragraphs from the article at pages 392-393 (the first of which sets the scene by summarising the effect of the Court of Appeal’s approach):

    Once the plaintiff had identified his loss as loss of a chance and the Court of Appeal had held that it was compensatable as such, the plaintiff was entitled to damages for the value of that lost chance because he could establish on the balance of probabilities that the defendants’ fault had caused that damage. The balance of probability test on the issue of causation of past events is retained but its effects are transformed by the reformulation of the damage which forms the gist of the action.

    Unhappily the House of Lords did not resolve the difficult question raised by the plaintiff’s argument - namely whether reformulation of the gist in terms of loss of a chance should now be acceptable. Their Lordships focussed on the traditional (but in this context, irrelevant) question of whether the plaintiff had shown a causal connection on the balance of probabilities to the necrosis, which of course he had and could not. The defendants, therefore, succeeded in their appeal because the plaintiff had failed on the issue of causation as it related to damage formulated in terms of necrosis. Their Lordships regarded this finding as releasing them from having to determine whether loss of a chance could form the gist of a complaint in negligence. The loss of a chance issue was classified as one of valuation/quantification and it was held that unless the plaintiff first succeeded on causation no issue of quantification of the loss could arise. This reasoning unfortunately fails to address the essence of the plaintiff’s argument, which was whether a claim formulated in a different way (i.e. in terms of loss of a chance) was acceptable. In other words, the loss of a chance issue here is not about the quantification/valuation of the interest destroyed (all judges accepted the trials judge’s valuation of the lost chance as 25%) but whether loss of a chance could constitute the gist to which the causation enquiry could be directed. Consideration of the plaintiff’s argument was, therefore, essential before any issue of causation could be addressed. It cannot be over-emphasised that the formulation of the ‘damage’ forming the gist of the action defines the causation question. Logically one can only deal with causation after one knows what the damage forming the gist of the action is.

  97. As Ms Stapleton suggests a little later in her article, it is surely implicit from the speeches as a whole that their Lordships would have rejected the loss of a chance argument had they confronted it squarely.

  98. Is there anything in Fairchild to suggest that a contrary view would be taken today? The House of Lords there had to address the loosely related but ultimately very different question as to whether in the particular circumstances which arose there a claimant can recover damages even though he cannot prove that it was the defendant who caused his injury, being able to prove only that the defendant had materially increased his risk of injury. Consistently with the House’s earlier decision in McGhee v National Coal Board [1973] 1 WLR 1, their Lordships held that “a less stringent causal connection [than the usual ‘but for’ test] was … sufficient” - per Lord Nichols at paragraphs 44 and 45.

  99. As Ms Stapleton pointed out in the final section of her article at pages 406-407, a section discussing the relationship between Hotson and McGhee (and Wilsher v Essex Area Health Authority [1987] QB 730), the McGhee approach applies only where it is impossible to assess the extent of increased risk brought about by the defendant’s negligence (the Hotson point, of course, generally arising only where this increased risk is assessable) and, where it applies, it allows the claimant to recover in full - “a result even more radically pro-plaintiff than the Hotson argument”.

  100. As Fairchild makes plain, however, this more relaxed approach to the proof of causation - “treat[ing] a material increase in risk as sufficient to satisfy the causal requirements for liability” (per Lord Hoffman at paragraph 67) - will apply only in the most narrowly defined circumstances (see Lord Bingham at paragraph 2 and Lord Hoffman at paragraph 61). There is nothing in the speeches to suggest a more relaxed approach generally to proof of causation nor, to my mind, to invite any fresh attempt to advance the Hotson argument.

  101. That said, it seems less than wholly satisfactory to leave the House of Lords speeches in Hotson as the final word on the loss of a chance argument. If I consult a doctor about a specific condition and, through the doctor’s negligence in diagnosis or treatment, reduce from 49% to 5% my chance of averting an adverse outcome, not everyone would think it “just and reasonable” that my claim must inevitably fail on the issue of causation. Particularly that may be thought unfair given that in cases where the claimant can prove on the balance of probabilities that, but for the negligence, he would have escaped the adverse consequence of his condition, his damages will nevertheless be discounted to reflect the possibility that this was not so - see, for example, Smith v Leech Brain and Co Ltd [1962] 2 QB 405 (discussed by Ms Stapleton at pages 397-398 of her article) and, subsequent to Hotson, Judge v Huntingdon Health Authority [1995] 6 Med LR 223.

  102. I recognise that there are powerful policy arguments for not extending the scope of liability within the field of health treatment (see for example Lord Hoffman’s speech in Fairchild at paragraph 69) and I recognise too the difficulty of defining sufficiently clearly and narrowly the circumstances in which it is sufficient for a claimant to prove merely the loss of the chance of avoiding physical injury rather than causation of the injury itself - a definition which would need to distinguish such cases as Wilsher. All I suggest is that the loss of a chance argument in this context deserves to be confronted more squarely than perhaps it was in Hotson and resolved in a clear and principled fashion.

  103. Since writing the foregoing, I have had the advantage of reading Mance LJ’s judgment in draft. It seems to me, with respect, to go a very long way towards justifying the existing state of the law as I conceive it to be. On the facts of the present case I find both its reasoning and its conclusion compelling.

  104. With considerable regret, I too would dismiss this appeal.


Hotson v East Berkshire Health Authority [1987] AC 750; Fairchild v Glenhaven Funeral Services Ltd [2002] 3WLR 89; Mallett v McMonagle [1970] AC 166; Tahir v Haringey Health Authority [1998] Lloyds LR: Med 104; Herskovits v Group Health Co-operative of Puget Sound (1983) 664 P 2d 474; Lawson v Laferriere (1991) 78 DLR (4th) 609; Chappel v Hart [1999] Lloyd’s Law Reports: Med 223; Environment Agency (formerly National Rivers Authority) v Empress Car Co (Abertillery) Ltd [1999] 2 AC 22; Rahman v Arearose [2001] QB 351; Bonnington Castings Ltd v Wardlaw [1956] AC 613; McGhee v National Coal Board [1973] AC 1; Chaplin v Hicks [1911] 2 KB 786; Spring v Norwich Union [1995] 2 AC 296; Allied Maples Group Ltd v Simmons & Simmons [1995] 1 WLR 1602; Thompson v Smiths Shiprepairers Ltd [1984] QB 405; Holtby v Brigham & Cowan (Hull) Ltd [2000] 3 AER 421 (CA); Hatton v Sutherland [2002] EWCA Civ 76; [2002] 2 AER 1; Wilsher v Essex Area Health Authority [1987] QB 730 (CA); [1988] AC 1074; Kitchen v Royal Air Force Association [1958] 1 WLR 563; Davies v Taylor [1974] AC 207; Bolitho v City and Hackney H.A. [1998] AC 232; Judge v Huntingdon Health Authority [1995] 6 Med LR 223; Summers v Tice 199 P 2d (1948); Sindell v Abbott Laboratories 26 Cal. 3d 588 (1980); Lawson v Lafferrière (1991) 78 DLR 609; Seyfert v Burnaby Hospital Society (1986) 27 DLR (4th) 96; Smith v Leech Brain and Co Ltd [1962] 2 QB 405; Chester v Afshar [2002] 3 AER 552; Doyle v Wallace [1998] PIQR Q146

Authors and other references

Jane Stapleton, "The Gist of Negligence", The Law Quarterly Review Vol. 104 July 1988

McGregor on Damages (18th ed.)

McGregor on Damages, 2d Suppl. to 16th Ed

C. Nesson, The Evidence or the Event? On Judicial Proof and the Acceptability of Verdicts (1985) Harvard. L.R. 1357

D. Shaviro, Statistical Probability Evidence and the Appearance of Justice (1989) Harvard L.R. 530


Simeon Maskrey Esq, QC & Julian Matthews Esq (instructed by Messrs Park Woodfine) for the Appellant

William Edis Esq (instructed by Messrs Tamsin Thomas, The Medical Defence Union) for the Respondent

all rights reserved