Wright (A Child) v Cambridge Medical Group (A Partnership)

This is an appeal against the dismissal of a claim brought by Clarice Wright against Cambridge Medical Group, a partnership of medical practitioners, for clinical negligence. The trial judge, Mackay J, held that, although the defendants admitted that they were negligent in not referring the claimant to a hospital when they should have done, she failed on the issue of causation. The course which the proceedings took was, through no fault of the Judge, unfortunate. Much of the evidence and argument concentrated on the issue of whether and to what extent the hospital, to which the claimant was eventually referred, had been and would have been negligent in the way in which it treated her, without that hospital being a party.

The basic facts of this case, which are much more fully and very clearly set out in Mackay J’s judgment below, [2010] EWHC 1507(QB), paras 14-35, 47-53 and 56-58, are as follows.

In early April 1998, the claimant, who was then aged eleven months, contracted chickenpox. She developed a high temperature and tachycardia, and was admitted to South Cleveland Hospital (“the Hospital”) on 9 April. Within a couple of days, the claimant developed a bacterial super-infection with streptococcus pyogenes, which the Hospital had not diagnosed by the time of her discharge on 12 April. The expert evidence established that the bacteria seeded into the proximal femur (i.e. that part of the hip-bone closest to the trunk of her body), resulting in osteomyelitis (i.e. infection of the bone), on 13 or 14 April.

On 14 April, it was clear that her condition was causing the claimant significant discomfort, and her mother (“Ms Wright”) took her to see Dr Winter of the defendants that day. No complaint is made about Dr Winter, who advised Ms Wright to contact the defendants if the claimant’s condition failed to improve. On the following day, 15 April, Ms Wright thought that the claimant’s condition was “considerably worse”, as she was lethargic, feverish, anorexic and suffering from diarrhoea. Accordingly, she made contact with Dr Phellas of the defendants by telephone at about 4.45 in the afternoon. Dr Phellas failed to make arrangements for the claimant to be seen, and it is conceded by the defendants that he was negligent in not doing so, and that, had he done so, he would, or at any rate should, have referred her to the Hospital.

The expert evidence established that the seeding of the bacteria in the claimant’s femur led to a sympathetic effusion developing in the claimant’s hip joint on 15 or 16 April. This led to her condition deteriorating, with the existing symptoms intensifying, and her hands and feet starting to swell and turn blue. Unsurprisingly, Ms Wright became very concerned indeed, and, at 7.30 on the evening of 17 April, her sister spoke to Dr Robertson of the defendants, who saw the claimant an hour later. He referred her at once to the Hospital, with a letter expressing his concerns, and she was admitted to the paediatrics unit around 9.00 in the evening. The expert evidence established that the effusion on her hip joint would have become infected by 17 or 18 April.

The claimant was seen at about 10.00 in the evening of 17 April by a senior house officer, Dr Joashi, who arranged a blood culture and prescribed the antibiotic erythromycin. Next morning, she was seen by Dr Turner, a registrar, who changed the treatment, prescribing intravenous amoxicillin, a different antibiotic. However, after talking to a microbiologist, Dr McKay-Ferguson, Dr Turner changed the treatment again that afternoon, prescribing yet another antibiotic, ciprofloxacin. Although that did not seem to improve the claimant’s condition, the treatment was continued until the claimant was seen by Dr Hampton, a consultant paediatrician, on a ward round on 20 April. This was the first time the claimant was seen, or her condition was even considered, by a consultant since her admission on 17 April.

Dr Hampton decided to change the antibiotic treatment, partly as a result of the apparent ineffectiveness of the current treatment, but mainly as a result of the blood culture result, which had been available since 19 April, and had revealed the streptococcus strain, which was resistant to ciprofloxacin, but, sadly and ironically, sensitive to amoxicillin and erythromycin. Unfortunately, even then, the proper diagnosis was not made. That only occurred on the following day, 21 April, after it was realised in the morning that the claimant was not moving her hip; that led to an ultrasound scan that afternoon, which showed septic arthritis. This resulted in urgent arthrotomy (i.e. incision into the joint) and surgical drainage, which was carried out in the early hours of 22 April. However, by then, the combination of the inappropriately treated osteomyelitis and septic arthritis had led to the separation and destruction of the bone plate. The claimant was discharged from hospital on 30 April.

The claimant has a permanently unstable hip, restricted movement range, leg length discrepancy, and restricted mobility. This long term condition is a result of her not having been treated properly in time. Had the defendants referred her to the Hospital on 15 April, as they should have done, and had she then been prescribed the right type of antibiotic, she would have recovered, probably without the need for surgical drainage, and certainly with none of the permanent damage which she has suffered.

The claimant issued the present proceedings against the defendants, essentially on the basis that their negligence in failing to refer her to the Hospital on 15 April had caused the permanent damage referred to in para 8 above. At any rate by the time the matter came to trial, the defendants accepted that they had been negligent in not referring the claimant to hospital in the late afternoon or early evening of 15 April. However, they denied that they were liable for the permanent damage. Although the claimant has apparently notified the Hospital about a possible claim, she has not brought proceedings against the Hospital or joined it in this action; nor have the defendants brought contribution proceedings against the Hospital.

The case came before Mackay J on the issue of the defendants’ liability. There were six expert witnesses, most of whom gave oral evidence. Given that the claimant’s case was that she had suffered permanent physical damage through having been referred to the Hospital negligently late, the experts unsurprisingly considered the course of her symptoms and condition, the appropriate treatment and its likely effect at successive stages, and the handling of her case by the defendants. However the experts also focussed to a large extent on the treatment which the claimant actually received after she was admitted to the Hospital on 17 April.

As well as making good the conceded fact that the defendants were negligent on 15 April, the expert evidence established that, had she been referred to the Hospital by Dr Phellas on 15 April, and had the Hospital then treated her properly with appropriate antibiotics, the claimant would very probably have made a full recovery without the need for any surgery. The evidence also demonstrated that, when she was actually referred to the Hospital, on the evening of 17 April, she would have made a full recovery if she had been treated promptly with the appropriate antibiotics, although she may well also have required the arthrotomy and surgical drainage which she underwent.

As explained above, the claimant’s argument was that, if the defendants had referred her to the Hospital on 15 April, as they accepted that they should have done, she would have been properly treated by the doctors in the Hospital and would therefore not have suffered the permanent injury that she has suffered.

After considering the factual and expert evidence relating to the treatment which the claimant received after she was admitted to the Hospital on 9 April, and, more importantly, 17 April, the Judge concluded at [2010] EWHC 1507 (QB), para 62, that “[t]he compelling criticisms of the performance of [the paediatric] unit through [the] period 9-21 April 1998 make it impossible for me to find as a fact that it is more likely than not that if [the claimant] had been placed in their hands during the evening of 15 April she would have been so treated as to avoid permanent damage to her hip”.

In other words, the Judge concluded that, although the defendants were negligent in not referring the claimant to the Hospital on 15 April, this negligence caused the claimant no loss, as, even if she had been admitted to the Hospital then, she would not have been treated properly, so that she would, in any event, have suffered the permanent damage described in para 8 above. The claim accordingly failed, as the claimant had failed to establish that the defendants’ negligence had caused the damage which she had suffered: given the standard of treatment she had actually received on 9-12 April, and, more relevantly, 17-21 April, he concluded that she would have received inadequate treatment if she had been admitted on 15 April.

The claimant appeals against this decision with permission given in part in writing by Sedley LJ, and in part at an oral hearing by the Vice-President and Thomas LJ.

For the claimant, Mr Oppenheim QC (who did not appear below) put forward three arguments, the second of which is the claimant’s main argument and has two aspects:

Mr Nolan QC, for the defendant, challenged each of these arguments, and also raised a separate argument, namely that the claim should have been dismissed on the ground that the only damage for which the defendants could be held liable was not pleaded by the claimant. That damage was the pain suffered by the claimant between 15 April (when she should have been referred) and 17 April (when she was referred), plus the pain caused by the drainage operation which would not have been necessary if she had been referred on 15 April. The rest of the damage which she suffered, and in particular the permanent damage of which she now complains, ran the argument, was solely attributable to the negligence of the Hospital on and after her admission on 17 April, not to the defendants’ failure to refer her to the Hospital on 15 April.

The most convenient way to deal with these various points is, I think, to begin by addressing this argument of the defendants, which, at least as far as I can see, was not really part of Mr Nolan’s case below and was not expressly dealt with by the Judge, and then to consider Mr Oppenheim’s various arguments, which were dealt with by the Judge. However, before considering these various arguments, it is appropriate to deal with the evidence and the Judge’s findings as to the treatment which the claimant received from the Hospital after her admissions on 9 April and on 17 April, as that impinges on the defendants’ first argument, and on the claimant’s second, and main, argument.

So far as the claimant’s first period in the Hospital was concerned, the expert evidence established that she received insufficient medical attention, in that it was inappropriate for a child in her condition to have been seen only once a day by a doctor on 9 and 10 April, and not to have been seen by a consultant for two days. It was also common ground among the expert witnesses that the doctors at the Hospital should have considered post-chickenpox bacterial infection as a possible diagnosis, and that they never did. The claimant’s experts described her discharge from the Hospital on 12 April as “just acceptable” (Dr Rudd, a consultant paediatrician) and “not appropriate” (Professor Pollard, a professor of paediatric infection and a consultant paediatrician, whom the Judge described as “an impressive witness” – [2010] EWHC 1507 (QB), para 45). Professor Pollard did say, however, that the failure to diagnose bacterial infection was defensible on this first admission.

The defects in the treatment accorded by the Hospital to the claimant following her admission on 17 April, as established by the evidence, were helpfully set out in a summary in Mr Nolan’s skeleton argument, which, at least to a substantial extent, was not, as I understood it, challenged by Mr Oppenheim. The essential defects constituted the following:

By way of example, I would mention the following. It was common ground between the experts that infection in the hip is “not notably difficult to diagnose or to treat” – [2010] EWHC 1507 (QB), para 54. Dr Rudd said that the claimant had received “substandard care”, that the absence of a consultant visit on 18 and 19 April infringed “standard practice across the country”, and that the failure to examine her bones and joints was “a breach of duty of care” by the registrar and the consultant”. Professor Pollard described the claimant’s treatment as “inadequate”, and said that a full septic screening should have been undertaken at once on her admission, and that immediate and consistent administration of a broad spectrum antibiotic treatment was clearly appropriate: neither occurred. Dr Conway, another consultant paediatrician, described the claimant as having received “inadequate” care, that the hospital diagnosis on 18 April was “not reasonable”, that the decision to treat her with ciprofloxacin was “not justified … inadequate and illogical”, and that her symptoms “should have suggested the probability of sepsis”.

The Judge’s findings as to the treatment given to the claimant after she was admitted to the Hospital on 17 April are set out at [2010] EWHC 1507 (QB), paras 54-59. In reaching his conclusion, the Judge laid particular stress on the fact that “there should have been at least the consultant ward rounds” every day, whereas there were “no consultants available” on the 18 or 19 April “to advise or supervise the junior doctors”, and there was “a serious defect in the treatment which [the claimant] was receiving” - [2010] EWHC 1507 (QB), para 61.

The issue which has to be faced is whether, in the light of the evidence and the findings of the Judge, the Hospital was actually negligent in its treatment of the claimant during her two periods during which she was under its care. The Judge did not expressly make a finding on that issue.

It is difficult, and fortunately unnecessary, to decide whether the Hospital was negligent during the period 9 to 12 April, but, in the light of the facts as summarised in para 19 above, there is no doubt that, if it was not negligent, it was close to the line. The Judge described the treatment accorded to the claimant during that time as “sub-standard” and “unimpressive, to put it at its lowest” – [2010] EWHC 1507, para 48.

More importantly, the argument before us appeared to proceed on the common assumption that the Judge’s conclusion that the claimant would not have been treated promptly and appropriately by the Hospital if the defendants had referred her on 15 April, involved a finding that the Hospital was negligent in its treatment of the claimant following her admission on 17 April. In my judgment, that common assumption is justified, as it seems clear from the evidence and from the Judge’s findings summarised in paras 20 to 22 above, that the Hospital was negligent in its treatment of the claimant from the time when she was admitted on 17 April until 21 April.

The Judge said at [2010] EWHC 1507 (QB), para 45, that he would not describe the Hospital as “negligent” in his judgment, because it would be unfair on the hospital doctors involved, as they “had not been heard on this issue”, and because the claimant may bring proceedings against the Hospital. And he introduced his conclusion in [2010] EWHC 1507 (QB), para 62, quoted in para 13 above, by saying that it would be “both wrong and unnecessary for me to pass judgment on the performance of the paediatric unit at [the Hospital] … to any greater extent than this.”

Given that the Hospital doctors responsible for the claimant’s treatment between 9 and 12 April, and between 17 and 21 April, did not give evidence, and that there was a real possibility of the claimant suing the Hospital, one can understand the Judge’s concern. However, any finding of primary fact or any inference of fact which he made in these proceedings could not be binding on the Hospital. Indeed, his judgment would probably be inadmissible as evidence in any negligence proceedings against the Hospital brought by the claimant, and, even if admissible in such proceedings or in contribution proceedings brought by the defendants, I doubt that his judgment would be given much weight in the face of fresh factual or expert evidence.

Whether or not that is right, it seems to me that, because of its relevance to the point on which he rejected the claim, the Judge should have decided whether his conclusion involved a finding that the Hospital would have been negligent in its treatment of the claimant if she had been admitted on 15 April, and that would have required him to make a similar finding about her actual treatment after her admission on 17 April. Accordingly, he should have grasped the nettle. To be fair to him, I think he impliedly did so. His finding of “systemic inadequacy” in the paediatric unit (as discussed below) is hard to reconcile with the absence of negligence. And the way in which he distinguished the decision in Gouldsmith v Mid-Staffordshire General NHS Hospital [2007] EWCA Civ 397 (as also discussed below) shows that he accepted that the Hospital would have been negligent if the claimant had been admitted on 15 April, which inevitably implies that he thought she was negligently treated when she was admitted.

The defendants’ case under this head is that their duty as the claimant’s general practitioners was to refer her to a hospital to enable her condition to be properly treated. They accept that they should be liable for having failed to refer the claimant when they should have done, but they contend that they complied with their duty by referring her on 17 April, which was well in time to have her hip condition remedied: it was only because of the Hospital’s negligence after 17 April that the condition was not remedied. In other words, the argument is that the defendants’ breach of duty was in having referred the claimant to the Hospital later than they should have done, but not too late to be treated, and that, as a result, the damages she should recover from the defendants should be for the loss she suffered from having been referred later than she should have been, not for having been referred too late for her condition to have been remedied.

It seems to me that this argument raises two questions. The first is whether, on a fair view of the facts, the defendants’ negligence was a cause of, or, to put it another way, significantly contributed to, the claimant’s permanent injury. The second question is whether that injury was, to use the traditional expression, too remote, or, to put it in more modern terms, whether that injury fell outside the scope of the defendants’ duty. I shall consider these two questions in turn, although I think that the reasoning on the first question also has a part to play when discussing the second.

So far as causation is concerned, although it may very well have been that, had it been a party, the Hospital would have been held to be more to blame than the defendants, I would reject the contention that the defendants’ admitted negligence did not contribute to the claimant’s permanent injury. The defendants’ case to the contrary, as summarised in para 29 above, has obvious attraction. However, it should be examined critically, because of the obvious point that, where there are successive tortfeasors, it cannot be right that each can avoid liability by blaming the other.

Accordingly, where there are successive tortfeasors, the contention that the causative potency of the negligence of the first is destroyed by the subsequent negligence of the second depends very much on the facts of the particular case. In many cases where there are successive acts of negligence by different parties, both parties can be held responsible for the damage which ensues, so that the issue is not which of them is liable, but how liability is to be apportioned between them. The mere fact that, if the second party had not been negligent, the damage which subsequently ensued would not have occurred, by no means automatically exonerates the first party’s negligence from being causative of that damage. As Lord Denning MR said in Lloyds Bank Ltd v Budd [1982] RTR 80, 83, “the doctrine of last opportunity is gone for ever.”

Much must depend on the facts. Consider the following example. A patient, with a condition which requires an operation in ten days, visits her general practitioner, who should refer her to a hospital at once. The general practitioner fails to refer her and sends her away, but she sees him again subsequently, and he then refers her, but the hospital negligently fails to operate within the ten days. The general practitioner would plainly have a much stronger case for denying any liability if the subsequent visit had been one day after the first visit than if it had been nine days after the first visit. In the former case, the causative potency of his negligence might well be said to have dissipated by the subsequent long delay in operating, and presumably egregious ineptitude, on the part of the hospital, particularly when compared with the small delay which lay at his door. The position may be very different in the latter case, where, although the hospital was negligent in failing to operate, there would have been much more time for it to have put things right, and it was placed under a tight schedule, as a result of the medical practitioner’s negligence.

In the context of clinical negligence, the issue was illuminatingly discussed by Laws LJ in a judgment, with which Henry and Schiemann LJJ agreed, in Rahman v Arearose Ltd [2000] EWCA Civ 190, [2001] QB 351 (and which was referred to with implied approval by Lord Bingham of Cornhill in Fairchild v Glenhaven Funeral Service Ltd [2002] UKHL 22, [2003] 1 AC 32, para 12). In that case, there were successive alleged causes of the injured claimant’s damage. The first alleged cause was the negligence of the injured claimant’s employer, the first defendant, in failing to take adequate precautions to prevent him from physical assault which resulted in the injury, which was then negligently treated by the second defendant hospital trust.

As Laws LJ pointed out at [2001] QB 351, para 28, there is a respectable argument, supported by authority, in favour of the proposition that “later negligence by a doctor … would amount to a ‘new cause’ and so break the chain of causation flowing from the original accident”. However, as he went on to say, it cannot be “a rule of law that later negligence always extinguishes the causative potency of an earlier tort”, as the ultimate test is that “every tortfeasor should compensate the injured claimant in respect of that loss and damage for which he should justly be held responsible” - [2001] QB 351, paras 30-31. Laws LJ added at [2001] QB 351, para 32, that this approach should ensure that “the metaphysics of causation can be kept in their proper place”, as they “offer no hope of a solution of the problems which confront the courts in this [area]”.

In the present case, I consider that the defendants’ negligence was a causative factor of the claimant’s permanent injury. In other words, as in Rahman [2001] QB 351, para 34, I have concluded that the negligence of the defendants and the failings of the Hospital had “a synergistic interaction, in that each tends to make the other worse”, and accordingly it seems appropriate to proceed on the basis that both were causative of the damage suffered by the claimant.

I do not consider that the Hospital’s failure to treat the claimant properly once she was admitted on 17 April was of such significance that it justifies a finding that the defendants’ negligence was not causative of the claimant’s injury - or indeed a finding that it broke the chain of causation between the defendants’ negligence and the claimant’s injury. It was not such an egregious event, in terms of the degree or unusualness of the negligence, or the period of time for which it lasted, to defeat or destroy the causative link between the defendants’ negligence and the claimant’s injury. It took the Hospital less than three days more than it should have taken to make the right diagnosis, and to administer the right treatment. But it took the defendants a little over two days more than it should have done to refer the claimant to the Hospital. Further, it should have been apparent to the defendants that the claimant’s condition on 15 April was deteriorating, and that, if treatment was delayed, it would increase the long term risk to her health: the evidence established that it was notorious that ill children can deteriorate fast.

It seems to me that, even accepting the Judge’s finding in favour of the defendants as to what would have happened if the claimant had been referred to the Hospital on 15 April, the effect of the agreed expert evidence was that the two days delay in referring the claimant to the Hospital was probably of crucial importance. In reaching this conclusion, I am assuming in the defendants’ favour that the Judge was entitled to conclude that the treatment which the claimant would have received if she had been admitted in the late afternoon or early evening of 15 April (as she should have been) would have mirrored the treatment she actually received when she was admitted in the late evening of 17 April. (If that assumption is incorrect, as I think it is for the reasons discussed in paras 66-79 below, then this point becomes even more telling in favour of the claimant).

On the agreed evidence of the experts, if the claimant had been first treated with the right antibiotics on 16 April, it “would probably have prevented progress [of the infection]”; on 17 April it would “possibly” have done so; but, if such treatment had only started on 18 April, “it is unlikely to have had an effect in the absence of surgical drainage”, and if it had started on 19 April, “it would have had no effect without surgical drainage”; They also agreed that, if surgical drainage had been effected on 17 April, the claimant would be likely to have had “a normal hip”; if it been done on 18 April, there “may have been residual articular cartilage damage”, but only “a low risk of AVN [sc. avascular necrosis] and growth disturbance”; if the drainage had occurred on 19 April, there was “an increasing risk of articular cartilage damage, AVN and growth disturbance”; whereas, if drainage had been delayed until 20 April, such damage, necrosis and disturbance had become inevitable.

After the claimant was admitted on 17 April, appropriate antibiotics were administered on 20 April, and the drainage occurred on 21 April. If one assumes an admission on 15 April, and identical treatment, the antibiotics would have been administered on 18 April and the drainage would have been effected on 19 April. The antibiotics on 18 April would have been “unlikely” to have cured her on their own, but they might have done so; coupled with the drainage a day later, there would have been no more than “an increas[ed] risk” of permanent damage over the “low risk” if the operation had been carried out the day before. Accordingly, if the defendants had not been negligent, it appears to me from the Judge’s findings, that there might well have been no permanent damage, and it is very likely that there would have been significantly less permanent damage, even if the Hospital had performed as inadequately as they did following the claimant’s admission on 17 April.

In conclusion on the issue of causation, the negligent delay in treatment which sadly resulted in the claimant’s permanent injury amounted, in aggregate, to rather less than six days, of which just over two days was attributable to the defendants’ failure to refer the claimant, and rather under three days was attributable to the Hospital’s failure to treat the claimant properly. At least in the absence of special factors, I find it hard to see why the defendants’ negligence should not be held to be a causative factor of the injury.

It is right to add that there was no finding by the Judge that the defendants’ negligence could not be said to have been causative of the claimant’s permanent injury, as Mr Nolan now contends. Either the point was not taken at all (as seems to me to have been the position) or it was implicitly rejected by the Judge. In either event, it seems to me that it would be wrong for an appellate court to accept it, when it had either not been raised or it had been impliedly rejected below.

I turn then to the second question which arises under this head, namely whether the claimant’s permanent injury was within the scope of the defendants’ duty. One way of putting that question is, as mentioned by Dame Janet Smith in her judgment, namely whether the defendants “ought fairly or reasonably or justly … to be held liable” (per Lord Nicholls of Birkenhead in Kuwait Airways Corporation v Iraqi Airways Co [2002] 2 AC 883, para 70).

The defendants’ duty on 15 April was to take such steps as a reasonable general practitioner, with the knowledge which Dr Phellas had, or should have had, of the claimant’s symptoms and history, would have taken. Any physical suffering or damage undergone by the claimant that could fairly be said to have been foreseeably caused by his breach of that duty, would properly found a claim against the defendants.

In a case such as this, where the patient was in a deteriorating condition, which required her to be urgently referred to a hospital, the risk, which ought to be within the scope of the general practitioner’s duty, is that delay in diagnosis and referral may have an adverse effect on the patient’s prospects of recovery. It seems to me wrong to limit the scope of that duty only to the adverse effect which would inescapably ensue if he delayed the referral to a point where permanent damage had become unavoidable. Any medical practitioner must know that prompt referral in such a case must widen the safety margin available for the patient, and that safety margins matter, because all human processes are liable to human error.

Quite apart from this, on the facts of this case, as discussed at paras 38-40 above, it seems to me that to hold that the defendants escape liability in reality would involve resurrecting the discredited last opportunity doctrine. Furthermore, to hold the defendants liable is not unfair, in that it was always open to the defendants to join the Hospital as a party, and, had they done so, the court could have apportioned liability between the two parties.

Accordingly, for these reasons (which are essentially the same as those given by Dame Janet Smith in paras 129 to 132 below) I would reject the defendants’ contention that the damage suffered by the claimant was outside the ambit of the defendants’ duty: I consider that it is fair and in accordance with principle to hold that the scope of the defendants’ duty in this case extended to the permanent injury suffered by the claimant. It is only right to acknowledge that, while this is the conclusion I had reached before reading the draft judgments of Elias LJ and Dame Janet Smith, I have somewhat reformulated my reasoning after having had the benefit of their respective analyses of this issue.

The claimant’s case therefore stands as it did below, but she still needs to overturn or avoid the Judge’s finding that she would not have been treated adequately by the Hospital if she had been admitted on 15 April. Before I turn to that aspect, it is appropriate to deal with a different point raised on behalf of the claimant.

The claimant’s first argument rests on the fact that the defendants’ breach of duty caused her some damage: even on the defendants’ case, she suffered three days extra pain and suffering (between 15 and 17 April), and she required a drainage procedure, which she would not have done if she had been referred on 15 April. The claimant’s case is that, once it is accepted that the condition of her hip deteriorated as a result of the defendants’ breach of duty, the defendants are liable for all the loss she suffered, even though the Hospital may also be liable for the loss arising from their failure to treat her adequately after she was admitted on 17 April.

In support of this contention, Mr Oppenheim relied on an observation of Devlin LJ in Dingle v Associated Newspapers Ltd [1961] 2 QB 162, 188-9, where he said:

“Where an injury has been done to the plaintiff and the injury is indivisible, any tortfeasor whose act has been a proximate cause of the injury must compensate for the whole of it. As between the plaintiff and the defendant, it is immaterial that there are others whose acts have also been the cause of the injury …. . [T]he plaintiff … can obtain judgment for total compensation from anyone whose act has been a cause of the injury.”

I do not consider that observation assists the claimant in this case. It not only begs the question of whether the defendants’ breach of duty was “a proximate cause of the injury” complained of, namely the permanent damage suffered by the claimant. It would also mean that a defendant’sargument based on lack of causation or damage outside scope of duty, as discussed above, could rarely, if ever, be raised in a clinical negligence case, even in a very strong case, once some damage had been suffered as a result of the first act of negligence. That seems an improbable outcome.

More centrally, I do not consider that the damage suffered by the claimant from 15 April onwards can be characterised as “indivisible”, which Devlin LJ described as being “essential” if his observation was to apply. He also went on to say that “whether it is so or not is a matter of fact and not a matter of law”. The pain suffered by the claimant between 15 and 17 April was, it appears to me, divisible from any pain suffered thereafter: that is demonstrated by the point that, unlike any pain or damage caused by delayed treatment at the Hospital, there is no way in which the Hospital could be held liable for it. As for the drainage, it seems to me also to be self-evidently divisible from the long-term damage suffered by the claimant. If the Hospital had undertaken the drainage promptly when the claimant had been admitted on 17 April, it would have been the subject of a self-contained claim against the defendants; if the Hospital had failed to undertake the drainage, and had therefore been liable, it would only have been a notional loss.

This conclusion also derives a little support from the reasoning of Laws LJ in Rahman [2001] 1 QB 351, para 28, as briefly discussed by Lord Bingham of Cornhill in Barker v Corus (UK) plc [2006] UKHL 20, [2006] 2 AC 572,para 111. I would therefore reject this first argument of the claimant.

The Judge’s reason for dismissing the claim was based on his conclusion that the Hospital would have failed to treat the claimant appropriately if she had been admitted on 15 April. That conclusion was, as he explained at [2010] EWHC 1507 (QB), para 62, based on his assessment of the treatment which was actually accorded to the claimant following her admission on 17 April, although the Judge also referred to the treatment she received following her admission on 9 April.

As explained above, it is in my view inescapable that the Judge’s conclusion that the claimant would not have been treated appropriately if she had been referred to the Hospital on 15 April carried with it the implication that the Hospital would have been negligent in their treatment of the claimant if she had been admitted on 15 April. If the Hospital was negligent in not having treated the claimant appropriately in time to avoid permanent damage when she was admitted on 17 April, it would a fortiori have been negligent not to have treated her appropriately in time to avoid permanent damage if she had been admitted on 15 April.

It appears to be a generally accepted proposition that a doctor cannot escape liability for damage caused to a patient by his breach of duty by establishing that, if he had not committed that breach, the damage would have been suffered anyway because he would have committed a subsequent breach of duty. The proposition was conceded in Bolitho v City and Hackney Health Authority [1998] AC 232, 240, where Lord Browne-Wilkinson said that he had “no doubt that the concession was correctly made”.

There was some discussion as to why the proposition is correct. It would seem to be an affront to justice if a doctor could escape liability by contending that, if he had not been negligent as alleged, the same damage would have occurred because he would have subsequently committed a different act of negligence. However, moral outrage and instinctive reaction are not always the safest of guides to legal principle. A sounder basis for the proposition may be the principle that a party cannot rely on his own wrong: in such a case a doctor would be seeking to rely on his own wrong, even if it is a hypothetical notional or contingent wrong. However, while that may well be a good reason justifying the proposition, the principle, while well established, has to be applied with care (see re C L Nye Ltd [1971] 1 Ch 442, 470-1, 475), and we were not referred to any case where it was applied to a hypothetical wrong.

A further reason for this proposition, which appears to me to be correct, is that, by committing the breach of duty, the doctor has prevented the patient from the opportunity of being treated appropriately, and had the patient had that opportunity, she would have had a claim for the same damage against the doctor for the very negligence upon which the doctor is relying to avoid liability. In other words, if a negligent doctor contends that the damage would have occurred anyway, because he would have committed a subsequent act of negligence, the patient can say that, if that argument is correct, it gets the doctor nowhere: as a result of his breach of duty, the doctor has deprived her of the right to claim for damages for the subsequent (if notional) act of negligence.

Mr Nolan challenged this analysis on two grounds. The first was that it had never been advanced as the basis for what is apparently known as the second rule in Bolitho. I do not know whether this analysis has previously been advanced, but that is scarcely an objection. What is more to the point is that we have not been shown any reasoning or authority with which it is inconsistent or which suggests that it is wrong, other than Mr Nolan’s second ground of objection.

That second ground is that the reasoning is flawed because it is not within the scope of a doctor’s duty to protect his patient from financial or economic loss. As to that, I accept that, in the ordinary run of things, a doctor’s duty to his patient does not extend to protecting her from economic loss. However, an award of damages for clinical negligence is, in a sense, the legal equivalent of proper clinical treatment: it is the nearest the law can get to putting the patient into the position that she should have been if the doctor had not been negligent. That is well demonstrated by Lord Blackburn’s famous explanation of the purpose of an award of damages in Livingstone v Rawyards Coal Co (1879-80) LR 5 App Cas 25, 39, namely that it is “[to] put the party who has been injured … in the same position as he would have been in if he had not sustained the wrong”. It is also consistent with Lord Diplock’s analysis of the function of damages, albeit in a contractual context, in Photo Productions Ltd v Securicor Ltd [1980] AC 827, 849, explaining how payment of damages is a secondary obligation of a defendant whose primary obligation is to perform his contractual obligation.

Accordingly, it seems to me that, in a case where a doctor has negligently failed to refer his patient to a hospital, and, as a consequence, she has lost the opportunity to be treated as she should have been by a hospital, the doctor cannot escape liability by establishing that the hospital would have negligently failed to treat the patient appropriately, even if he had promptly referred her. Even if the doctor established this, it would not enable him to escape liability, because, by negligently failing to refer the patient promptly, he deprived her of the opportunity to be treated properly by the hospital, and, if they had not treated her properly, that opportunity would be reflected by the fact that she would have been able to recover damages from them.

I draw a little comfort from the fact that the point appears to have gone by concession in Gouldsmith [2007] EWCA Civ 397, although it was not, in the event, necessary to the decision. In these circumstances, the case, though treated by Mr Oppenheim as of central importance on this issue, is ultimately of very limited assistance.

The Judge concluded that he could distinguish what was said in Gouldsmith [2007] EWCA Civ 397, on the ground that in this case the claimant had actually been referred to the Hospital, albeit late. Unless such a referral, which was late, but not too late, can result in the Hospital’s subsequent failure to treat the claimant depriving the failure to refer of causative effect or as taking any damage out of the scope of duty of the referring doctor, as discussed above, I do not see why the fact that she was referred later should, without more, be a ground for disapplying the principle I have discussed.

It follows from this that I consider that the claimant is correct in contending that the basis upon which the Judge dismissed the claim in this case was flawed as a matter of law, albeit for reasons which were not, I think, advanced before him.

The Judge’s conclusion that the claimant would not have been competently treated if she had been referred to the Hospital on 15 April does not, in these circumstances, require to be considered. However, it should be addressed in case I am wrong on the point just discussed, and, although the issue was canvassed in some factual detail in the evidence and argument before the Judge, I can deal with it without going into much of that detail.

The ineptness of the treatment accorded to the claimant when she was admitted on 17 April is summarised in paras 20 to 22 above. The Judge made it clear that the defects in the treatment meted out to the claimant during that period were, in his view, typical of the way in which the paediatric unit at the Hospital functioned. Thus, at [2010] EWHC 1507 (QB), para 60, he said that “it was not a case of a single lapse by a single practitioner”, and that “the charge of systemic inadequacy is the most worrying criticism of [the Hospital] particularly the very late, limited and inadequate consultant involvement.”

This conclusion was an inference which was based on primary facts, those facts were technical and involved detailed expert assessment, and the whole question had been the subject of fairly extensive oral argument. Accordingly, it would require a particularly strong and cogent case before an appellate court could properly interfere with that conclusion. Despite this, for three main reasons, bolstered by three further points, I have come to the view that the conclusion cannot be supported.

My first main reason is that I consider that the Judge was wrong to reject as “speculative” the possibility that, if she was admitted on the late afternoon or early evening of 15 April (a Wednesday), the claimant would not have been treated by the same doctors, or with the same degree of negligence, as after she was admitted on the evening of 17 April (a Friday). So far as the doctors were concerned, in addition to Doctors Joashi, Turner and Hampton, there were a number of other doctors (at least another four other doctors who were not consultants and, apparently, another seven consultants) working in the paediatric unit of the Hospital during April 1998, who could well have looked after the claimant if she had been admitted on 15 April. Indeed, as the Judge recorded at [2010] EWHC 1507 (QB), para 61, Professor Pollard agreed that there would have been “more doctors around” if she had been admitted in the late afternoon or evening of 17 April rather than in the evening of 15 April, and in the same passage the Judge seems to have accepted that the claimant would have been more likely to have been examined soon after admission by a consultant if she had been admitted on 15 April.

Given that the ultimate negligence of the Hospital was the failure to treat the claimant’s condition appropriately, it seems to me that the identity or likely identity of the doctors who would have seen her, or who would have been consulted about her condition, if she had been admitted on 15 April was crucial. However, having identified that issue, the Judge does not seem to have faced up to it in terms: rather he concentrated on what he saw as the “systemic inadequacy” of the paediatric unit.

That brings me to the second main reason for doubting the Judge’s conclusion. Although he rested his case on “systemic inadequacy”, the only such inadequacy which the Judge identified was the absence of consultants – [2010] EWHC 1507 (QB), para 60 (quoted in para 66 above). As to that inadequacy, the Judge accepted in the following paragraph that “the high probability is that there were no consultants available at all that weekend [sc. 18/19 April], to advise or supervise the junior doctors”, but that clearly was not the position during the week. So, rather than supporting his view as to the likely treatment which the claimant would have received if she had been admitted on Wednesday 15 April, the Judge’s only specific example of systemic inadequacy after the claimant was admitted in the late evening of Friday 17 April, rather undermines that view.

It is true that the Judge immediately went on to say that he was “not satisfied that a weekday referral would probably have led to any significantly better result” than that which eventuated following the claimant’s admission on 17 April. However, he gave no reasons for this opinion, and it is hard to reconcile with his findings as to the importance of the availability of consultants every day, and their apparent availability in the paediatric unit during the week, but not during weekends.

These first and second points, when taken together, are particularly telling, as, following her admission in the late evening of 17 April, the correct antibiotic treatment was only consistently pursued after the claimant had been first examined by a consultant on the morning of 20 April. The fact that consultants were unavailable at weekends, but not during the week, rather suggests that the claimant would have been examined by a consultant on 16 April (at the latest) if she had been admitted in the early evening of 15 April. Given that the majority of the delay and negligence on and after 17 April appears to have been attributable to doctors who were not consultants, there is a powerful case for thinking that, if she had been admitted on 15 April, the claimant’s treatment would have been significantly better than when she was admitted on 17 April.

My third main reason is that, quite apart from these two points, it appears to me that the Judge’s conclusion did not take into account the agreed expert evidence as summarised and discussed in paras 38 to 40 above, which, in fairness to him, seems to have been overlooked in the argument before him (and, indeed, the argument before us). In my judgment, the effect of that evidence is to establish, at the lowest, that it is more likely than not that any permanent damage the claimant would have suffered due to the inept treatment, which the Judge found that she would have received if she had been referred on 15 April, would have been significantly less than that which she did suffer. Indeed, I think that this evidence established that there is a reasonable chance that she would have suffered relatively little long term damage if she had been referred in the late afternoon or early evening of 15 April.

There are three further points, which suggest that one should approach with caution the Judge’s finding that the Hospital’s negligence following the claimant’s admission on 17 April justifies the conclusion that it would have been similarly negligent if she had been admitted on 15 April.

First, I think that the Judge misdirected himself on the burden of proof. When considering the claimant’s argument that different doctors might have attended her if she had been admitted on 15 April from those who did attend her following her admission on 17 April, he said that “[t]hat is to some extent a speculative argument and the burden of proof is on the claimant at this stage of the case” – [2010] EWHC 1507 (QB), para 60. It is true that the burden of establishing causation in a negligence claim is, in principle, as with every other ingredient of the claim, on the claimant. However, once the claimant established that (i) she could and should have been referred to the Hospital on 15 April, and (ii) she would not have suffered the damage now complained of had she been so referred and been treated competently at the Hospital, she had the benefit of a presumption that she would have been competently treated thereafter. In the absence of evidence to the contrary, the court will assume that professional and other service providers would have or have performed their functions competently. Of course, provided that I am wrong in my view that this presumption is effectively irrebuttable as a matter of law in a case such as this, that presumption can be discharged by evidence, but it is, as it were, the right starting point.

A trial judge’s error as to which party has the burden of proof is often not a particularly powerful point in a civil case where there has been significant evidence from both parties. However, in this case I consider it carries the claimant some way. Not only did the Judge rely on it effectively to reject a point made by the claimant as to the possibility of different doctors attending on 15 April, but he appears to have given it some force in his final conclusion, at [2010] EWHC 1507 (QB), para 62, that the previous treatment of the claimant “make it impossible for me to find as a fact that it is more likely than not that if [she] had been placed in their hands during the evening of 15 April she would have been so treated as to avoid permanent damage to her hip”.

Secondly, even with the chapter of negligent acts, the Hospital only narrowly failed to treat the claimant appropriately shortly after she was admitted on 17 April. As the Judge explained at [2010] EWHC 1507, para 57, “Dr Turner seems to have got near reaching [the right] conclusion”, as to the correct treatment on 18 April, but was “distracted by his [mistaken] theory” that the claimant was suffering from gastroenteritis. That shows how close the Hospital got to treating the claimant appropriately the day after she was admitted, but it also means that things need not have been much better on 15 April for her to have received appropriate treatment if she had been admitted that day.

Thirdly, there was some evidence to support the contention that, because the claimant’s condition on 15 April was less complex or extreme than on 17 April, a correct diagnosis would have been easier on the former date than on the latter. Although he identified the claimant’s reasons for supporting this proposition at [2010] EWHC 1507 (QB), para 61, the Judge does not appear to have made a finding on this contention. Particularly as the contention was contested by the defendants before us, it is not one which we can sensibly rule on, but the absence of any conclusion on the point is a further ground for doubting the Judge’s conclusion on this issue.

Drawing these points together, I have reached the conclusion that the Judge’s decision that the claimant would have suffered the same damage as she did suffer even if she had been admitted on 15 April cannot stand. In the light of:

The claimant’s third line of argument was that, even if the Judge was right in his conclusion that she would not have been treated properly had she been admitted to the Hospital on 15 April, he was wrong to reject her claim. Mr Oppenheim’s point was that, even if there was a less than evens prospect that the claimant would have been properly treated, there must have been a real chance that she would have been, and that she should therefore be compensated for the loss of that chance. It is unnecessary to consider that argument, in the light of the conclusions which I have reached so far, but it is a point of some interest.

Mr Oppenheim’s argument is that, even if the Judge was right to hold that she was more likely than not to have been ineptly treated by the Hospital if she had been referred on 15 April, she still lost the chance of being treated properly and thereby avoiding the damage she has suffered, by not being referred to the Hospital on 15 April. That argument derives some apparent support from the reasoning of the Court of Appeal in Allied Maples Group Ltd v Simmons & Simmons[1995] 1 WLR 1602, a case of solicitor’s negligence in the context of a commercial property transaction. In that case, Stuart-Smith LJ explained that damages should be assessed on a loss of a chance basis because the claim was one where “the plaintiff's loss depends on the hypothetical action of a third party, either in addition to action by the plaintiff … or independently of it” - [1995] 1 WLR 1602, 1611. So, it can be said here, the claimant’s loss depends on the hypothetical action of a third party, namely the Hospital.

However, at least in this court, I consider that such an argument should probably be treated as foreclosed by the views expressed by the majority of the House of Lords in Gregg v Scott [2005] UKHL 2, [2005] 2 AC 176. It is true that the issue in that case did not turn upon what a third party, to whom the negligent doctor should have referred the patient earlier than he did, would have done, but with the extent, if any, by which the patient’s prospects of living ten years would have improved if the negligent doctor had referred him earlier. However, while Lord Nicholls of Birkenhead (who was in the minority with Lord Hope of Craighead) was attracted by the application of the loss of a chance approach, the other members of the Committee seem to have deprecated such an approach in the context of a claim in clinical negligence.

Lord Hoffmann, at [2005] 2 AC 176, paras 72-90, primarily based his reasoning on the difference between deterministic and indeterministic causes. However, in the light of (i) his dismissal of the relevance of the loss of a chance cases as being concerned with commercial loss, (ii) his emphasis on the importance of causality in clinical negligence cases, and (iii) his disapproval of novel departures from established principles in clinical negligence cases (see [2005] 2 AC 176, paras 83 to 85), it seems to me that he was effectively disapproving the importation of loss of a chance in this area. Lord Phillips of Worth Matravers appears to have taken the same view at [2005] 2 AC 176, para 170 to 174, where he said that “[i]t is always likely to be much easier to resolve issues of causation on balance of probabilities than to identify in terms of percentage the effect that clinical negligence had on the chances of a favourable outcome”, so that “[a] robust test which produces rough justice” was to be preferred. Baroness Hale of Richmond at [2005] 2 AC 176, paras 210 to 226, in an interesting discussion, appears to have rejected the loss of a chance approach when it comes to claims for personal injury, and clinical negligence in particular, for a number of reasons. I should add that Lord Hope was understood by Lord Phillips at [2005] 2 AC 176, para 175, to eschew a loss of a chance approach.

I accept that the reasoning of the House of Lords on this point does not conclusively shut out, as a matter of strict logic, this court from applying a loss of a chance approach in this case, which is concerned with rather a different point. However, certainty and consistency are of great importance in this difficult area, and, while the question would be appropriate for reconsideration by the Supreme Court, I consider that, at this level, we should probably not expand the loss of a chance doctrine into the realm of clinical negligence. It is fair to add that we had relatively limited argument on the point.

In these circumstances:

I cannot leave this case without expressing regret and surprise that neither party saw fit to join the Hospital in these proceedings. The claimant had notified the Hospital of a claim and was contemplating bringing proceedings against it in respect of the very damage which she was alleging against the defendants. The defendants were defending the claim on the basis that the Hospital’s treatment of the claimant had been negligent. It was little short of absurd to have a long hearing which involved many medical expert witnesses, a great deal of whose evidence was devoted to the treatment which the claimant underwent when she was in the Hospital, and which resulted in a judgment, a large proportion of which was given over to the actual and hypothetical treatment of the claimant by the Hospital, without the Hospital being a party.

As a general proposition, it is certainly right to keep the number of parties to a minimum, for reasons of cost and time. However, in some cases it can be more expensive in terms of cost and time to omit joining a relevant party. This, as I see it, was plainly such a case. If the claimant had lost (as the Judge concluded that she should), then the same issues as were canvassed before him would have had to have been examined in her subsequent claim against the Hospital. If the defendants had lost (as I consider that they should), then those issues may well have to be debated in contribution proceedings between the defendants and the Hospital. Apart from concern about time and costs in this case, there is a real risk of inconsistent findings. Each party took an unwarranted risk in not joining the Hospital. It also seems to me that there may well have been a better chance of a negotiated settlement if the Hospital had been a party.

As it is, I would allow this appeal and enter judgment on liability for the claimant.

I gratefully adopt the analysis of the facts in the judgment of the Master of the Rolls. To summarise the essential features in a nutshell, the claimant ought to have been referred by her doctor on the Wednesday; in fact, she was referred on the Friday evening. Had appropriate steps been taken by the hospital at that time, the permanent damage to the hip would have been avoided. However, she was not given the appropriate antibiotics and by the Sunday it was probably too late. Accordingly, by the time the consultant saw the patient on the Monday, nothing could be done to prevent the permanent injury.

I respectfully agree with the Master of the Rolls on three of the issues arising in this case, but for reasons I will explain, I would dismiss the appeal. The matters on which I agree, and on which in essence I adopt his reasoning, are the following.

First, the judge was not entitled on the evidence to conclude that if the claimant had been referred to the hospital on the Wednesday, she would on the balance of probabilities have been negligently treated and would still have suffered the permanent hip injury. The reasonable inference is that the hospital would have treated the patient properly and, as the defendants’ counsel Mr. Nolan QC accepted, the evidential burden (as opposed to the legal or persuasive burden – see Phipson on Evidence, 16^th edition, paras 6-02 and 6-03) shifted to the defendants to establish otherwise. For reasons given by the Master of the Rolls, I do not think that burden was discharged. The claimant may well have been seen by different doctors on that day, and there would have been consultants readily available, which there were not (although there should have been) over the weekend. The inference that the paediatric unit in the hospital was subject to such systemic failings that it was incapable of treating the claimant properly was in my view not open to the judge. It follows that in my judgment this case has to be determined on the assumption that on the balance of probabilities, if the claimant had been referred on the Wednesday as she ought to have been, she would not have suffered the permanent hip injury.

Second, I agree that the injury directly arising from the failure to refer, namely the additional pain and suffering over the two day period (which might also have led to the need for some surgical intervention) was quite distinct from the permanent hip injury resulting from the failure by the hospital doctors to provide the antibiotics in time. I recognise that the courts should avoid adopting too sophisticated an analysis of a claimant’s injuries, attributing different injuries to different defendants with the consequence that a claimant is compelled to pursue a number of different tortfeasors in order to recover full compensation for all the damage suffered. But here I think it would be quite artificial to treat the permanent injuries and the earlier pain and suffering resulting from the delayed referral as all of a piece. Accordingly, this was not a case where there was joint contribution to an indivisible injury, in which case both the doctor and the hospital would be liable for the whole amount (with contribution as between each other.) The principle enunciated by Lord Justice Devlin, as he was, in Dingle v Associated Newspapers [1961] 2 QB 162, 188 is not applicable. I also note that this argument was not advanced before the judge below.

Third, I agree with the Master of the Rolls that this is not an appropriate case to apply the loss of chance principles enunciated in Allied Maples Ltd v Simmons and Simmons [1995] 1 WLR 1602. I agree that Gregg v Scott [2005] 2 AC 176, even if it is not strictly determinative of this aspect of the case, effectively precludes this court from applying those principles to personal injury –and particularly clinical negligence - cases.

In my judgment, therefore, the relevant factual analysis is as follows: the doctor ought to have referred the patient on Wednesday, and had he done so she would have been treated properly and the permanent injury would, on the balance of probabilities, have been prevented. In fact, she was not referred until the Friday. At that stage also the permanent injury could have been avoided had the hospital doctors treated her appropriately, but they failed to do so. It was not a difficult diagnosis and the treatment was very straightforward. The hospital doctors were negligent in a series of ways identified by the Master of the Rolls in his judgment (para.20). Even had the appropriate treatment been provided on the Saturday the permanent damage would have been avoided, although surgical intervention would probably have been necessary. The risk of permanent damage increased on the Sunday and by the time the consultant became aware of her condition on the Monday, it was too late. The question is whether in those circumstances the doctor is jointly liable with the hospital for the permanent hip injury.

However, before considering the legal issues on the premise that the hospital would have treated the claimant properly had she been referred promptly on the Wednesday, I shall first briefly consider the question identified by the Master of the Rolls as question (ii)(a), namely whether, even on the assumption that the hospital doctors would have been negligent on the Wednesday, the judge ought to have found the doctor liable as a matter of law in view of the decision of the Court of Appeal in Gouldsmith v Mid Staffordshire General Hospitals NHS Trust [2007] EWCA Civ 397. The judge held that that there would be no liability in those circumstances on the simple basis, as I understand his decision, that the doctor’s negligence did not as a matter of fact cause or contribute to the injury and that Gouldsmith was distinguishable. I agree with that analysis and I shall explain why.

It is clear that the doctor owed a duty of care to the claimant, and that the failure to examine the patient on the Wednesday involved a breach of that duty. Thereafter the relevant question is whether as a matter of law the breach of duty caused the injury. This raises two sub-questions as the observations of Lord Nicholls’ judgment in the Kuwait Airlines case, reproduced in the judgment of Lady Justice Smith (para. 126) makes clear. The first is whether in fact the breach caused the loss, or at least materially contributed to it. If so, the second question is whether it is appropriate to impose liability for that breach. That raises issues of policy as to whether a defendant ought to be liable for the particular damage caused. This is sometimes described as asking whether there is causation in law. The first question asks whether historically there is a link in fact between the breach of duty and the damage. It is only in very exceptional cases that the courts will impose liability for negligence where the direct causal link cannot be established (see e.g the case where a claimant suffers mesothelioma resulting from exposure to asbestos dust in circumstances where he is not in a position to identify which employer from amongst a number of possible contenders caused the damage, although each had materially increased the risk: Fairchild v Glenavon Funeral Services Ltd [2002] UKHL22; [2003] 1 AC 32). This is not such a case. Accordingly, causation in fact is a necessary but not necessarily a sufficient condition before any liability can arise.

It follows that in my view on the assumption made by the judge that the claimant would still have been in the same position on the Friday even had the doctor referred her on the Wednesday, the causal link between negligence and damage does not exist. In those circumstances, the doctor’s negligence would have had no adverse consequences for the claimant.

I accept that the position will be different if the doctor fails to give the hospital an opportunity to provide proper treatment at all. In that case he alone is liable for the whole of the damage. His breach of duty, the omission to present the patient in time to receive appropriate treatment from the hospital, will have led directly to the injury. The doctor cannot then seek to transfer liability to the hospital, or to claim any contribution from the hospital, on the basis that even if he had referred the patient promptly, the hospital doctors would have been negligent and the injury would not have been prevented. The hospital never was negligent and cannot be liable for hypothetical negligence. I agree with the Master of the Rolls that the explanation for making the doctor liable in these circumstances, notwithstanding that the injury would still have been suffered even had there been no negligence by the doctor, is that the doctor will have deprived the claimant of the opportunity to sue the hospital. So even though the doctor’s negligence has not in fact caused or contributed to the injury, it has caused the patient’s inability to sue the hospital with respect to it. It would be bizarre if the claimant had no remedy against anyone at all in these circumstances. This is another exceptional situation where although the negligence has not in fact caused the injury, because it would have resulted even if the doctor had not been negligent, liability is imposed.

In my judgment this explains the decision of the Court of Appeal in Gouldsmith. In that case a hospital negligently failed to refer a patient to a specialist tertiary hospital as it ought to have done. If particular specialist treatment had been provided in time, the injury in fact suffered (loss of fingers) would have been avoided. The judge held that although there was negligence, the claimant had not shown that it had caused the injury because he had not satisfied the court that every specialist, exercising a proper standard of care, would have provided the relevant treatment. The Court of Appeal by a majority (Pill and Wilson LJJ, Maurice Kay LJ dissenting) held that this was not the right test. They held if the relevant treatment would in fact have been given, or if not, it would have been negligent not to give it and not merely a legitimate exercise of professional judgment (applying the second limb in the well-known Bolitho test), then the hospital would be liable for failing to refer. It was never argued before the court that if the hospital’s negligence had not in fact caused the injury, there should be no liability.

The claimant’s case is that the principle in Gouldsmith applies where there is delay in referral even where the doctor refers the patient in time for the hospital to provide the appropriate treatment. If the doctor is in breach of his duty to refer when he should, and the hospital in fact provides negligent treatment, then the doctor remains liable even if the hospital would have been equally negligent had the patient been presented promptly. Accordingly, Mr Oppenheim QC, counsel for the claimant, contends that even on the judge’s finding that there was no causal connection between the negligence and the injury, nonetheless Gouldsmith binds this court to conclude that the doctor is in those circumstances liable for the damage. There is no need for any causal connection in fact between his negligence and the injury. The judge, therefore, erred in law in concluding that there would be no liability in these circumstances.

In my judgment that argument is misconceived for two reasons. First, it is wrong to say that Gouldsmith binds this court since the point now advanced was never argued in that case; the court merely assumed on the facts of that case that the referring hospital would be liable if the specialist unit would have been negligent. A court is not bound by a proposition of law which the earlier court simply assumed to be correct without argument or consideration, and that is so even if it is part of the ratio of the case: see the judgment of Buxton LJ in R (Kadhim) v Brent LBC House Benefit Review Board [2001] QB 955.

Second and more importantly, Gouldsmith is in my view plainly distinguishable from this case for the reasons given above. It applies where there is a complete failure to refer in time so that the patient is denied either the appropriate treatment or, if the treatment would have been negligent, the opportunity to recover damages from the hospital for any injury suffered as a consequence of that negligent treatment. Neither of these consequences arises where the patient is referred in time to enable appropriate treatment to be provided.

In my judgment Gouldsmith does not impose liability where there is no causal link in fact between the doctor’s negligence in failing to refer and the injury, and the referring doctor has not by his negligence denied the claimant the opportunity to sue the hospital. It would in my view be contrary to principle to hold the doctor liable for delaying the referral if the delay had in fact made no difference either to the claimant’s injury or her ability to sue the hospital. I do not think that Gouldsmith has any relevance in that situation. If the damage would have been incurred in the same way even had there been no negligence, and the claimant is not deprived of the opportunity to sue the hospital, there is merely negligence in the air which has had no effect on the claimant at all.

However, I do accept that the conclusion that there is no liability is premised on the judge’s finding that the hospital would have been negligent on the Wednesday and that the patient was no worse off as a result of the delay in referring. In fact that premise is wrong for two reasons. First, even on the assumption that there would have been equivalent negligence by the hospital on the Wednesday as occurred on the Friday, there is still an issue as to whether the delay is causally relevant on the basis that it will have limited the opportunity for the hospital doctors to correct their own negligence. I consider this more fully below. Second, given that we have interfered with the judge’s findings of the facts, there is causation in fact; but for the negligent failure to refer in time, the claimant would have been treated on the Wednesday and the probability is that the damage would not have resulted if that had occurred.

I turn therefore to consider what seems to me to be the central issue in this case, namely whether the doctor should be liable in law for the permanent injury to the claimant’s hip given that the injury would not have been caused but for his negligence.

It seems to have been assumed by the judge below that if the hospital doctors would have treated the claimant properly had she been referred on the Wednesday, as we have found that they would, then the doctor would be liable. I do not believe that assumption to be correct. There is always the possibility that a hospital will act negligently in treating a particular patient, but it is fortuitous on which day that might arise. Even if unbeknown to the referring doctor, the negligence is linked to a particular incompetent hospital doctor or doctors, it is a matter of chance whether they will be the relevant doctors on duty when the patient is referred. There is no liability merely because in fact, as a consequence of the referring doctor’s negligence, the patient is referred on the wrong day and the hospital happens to be negligent on that day. In other circumstances, the referring doctor’s negligence could have fortuitously been to the patient’s benefit, such as where a prompt referral would have resulted in a negligent diagnosis whereas a delayed referral did not do so. The doctor’s negligence in failing to refer when he ought to have done has not increased the risk of injury resulting from the hospital’s negligence.

This principle was confirmed by the House of Lords in Chester v Afshar [2004] UKHL 41;[2005] 1 A.C. 134. A doctor failed to warn a patient about the potential risks of a particular operation. There was a 1-2% chance of significant adverse injury whenever the operation was performed quite irrespective of the particular circumstances of the patient. It was not possible to identify when the risk would materialise. That risk in fact materialised following the operation. If the warning had been given, the patient would have still had the operation after considering the implications, but it would have been carried out on a different day, and the very high probability is that the injury would not have been sustained at all on that occasion. Their Lordships held by a majority of 3 to 2 (Lord Steyn, Lord Hope of Craighead and Lord Walker of Gestingthorpe; Lords Bingham of Cornhill and Lord Hoffman dissenting) that the doctor should be liable. However, the basis of liability was that the obligation to warn was an important protection of patient autonomy which should not be undermined even if it is clear that the patient would have taken the risk of having the operation in any event. The important point for the purposes of this case is that their Lordships rejected the contention that the doctor should be liable on the simple basis that in fact the patient would not have suffered the injury but for the failure to warn. In failing to warn, the doctor had not increased the risk of the operation. It was quite arbitrary when that risk would materialise; the risk was the same whenever the operation occurred. The majority only imposed liability in order to give an effective remedy for breach of the right to warn.

In my judgment that is the situation here: in failing to refer on any particular date, the doctor has not increased the risk of the hospital being negligent. There was no evidence, for example, that the delay made it especially difficult for the hospital doctors to identify the appropriate treatment and that this might in part explain why such treatment was not taken. Mr Oppenheim suggested that there was, but if anything the expert evidence was to the contrary. The appropriate treatment was plain. It was a matter of coincidence that the negligence happened to be on the Friday; the hospital could just as well have been negligent on the Wednesday if the claimant had been referred then, as she should have been, or any other day. Accordingly, this factual causal link between the breach of duty and damage – the fact that but for the breach the claimant would have been treated without negligence on another occasion - was not itself enough to justify the imposition of liability.

The position here is, however, different from Chester to this extent. Although the doctor’s delay has not increased the risk of the hospital making a false diagnosis, it did restrict the time available to the hospital to correct any negligent treatment. Delay in referral will not necessarily have that effect. Sometimes the treatment is irreversible and cannot be remedied. But if the injury is linked to the gradual deterioration of the patient, then the longer the patient is in hospital before the critical point of no return is reached, the more chance there is of correcting any inappropriate treatment. Here if the claimant had been referred on the Wednesday, and if the hospital doctors had been negligent on that day as they were on the Friday, then as the Master of the Rolls and Smith LJ have suggested, the likelihood is that the doctors would have remedied the initial error before it became too late on the Sunday. To that extent, it can be said that doctor’s negligence did indirectly increase the risk of the claimant’s injury and was to that extent causally related to the injury. Is it justifiable in those circumstances to make the doctor liable for the hospital doctors’ negligence?

Whether later negligence of a third party will have the effect of relieving the earlier wrongdoer of liability for his negligent conduct depends on whether it is just that this should be so: see the observations of Laws LJ in Rahman v Arearose Ltd [2001] QB 351 para.29 which chime with the subsequent observations of Lord Nicholls in the Kuwait Airways case. Considerations of policy loom large in the analysis of whether the injury is in law too remote from the breach, notwithstanding the causal link in fact, although they are generally concealed beneath the legal concepts used to justify the result. Sometimes it is said that the liability is outside the scope of liability for breach of the particular duty, or that the act of the third party is a novus actus interveniens, or simply that the damage is too remote. The choice of concept is not, however, entirely arbitrary. Depending upon the context, one might more adequately encapsulate the court’s reasoning than another.

In my judgment the doctor should not be liable notwithstanding the fact that his act of negligence has in fact caused foreseeable loss. I would not reach this conclusion by focusing on the concept of novus actus interveniens. That is an appropriate concept to employ where the new act is not foreseeable, for example, but I would accept that it was foreseeable that the hospital doctors, like other professionals, will be negligent from time to time. Nor could it be said that the negligence in this case, although serious, deserved to be characterised as gross or egregious so as to break the chain of causation and make it unjust for that reason to impose liability on the doctor, such as was envisaged in Webb v Barclays Bank plc and Portsmouth Hospitals[2001] EWCA Civ 1141; [2001] Lloyds Rep. Med. 500 I would prefer to say that the injury caused to the claimant did not fall within the scope of the duty breached. That duty is to carry out a careful investigation of the patient’s condition and if specialist treatment is required, the patient must be referred to the appropriate body. In some cases, and this was one, the claimant’s symptoms are such that it will be necessary for the doctor to refer the patient so that he or she is seen by the specialist as soon as is reasonably practicable. The failure to do so may prolong the suffering and the doctor will be liable for that direct damage which is a consequence of the culpable delay. But if there is a referral in time for the hospital to provide appropriate treatment, and the doctor has not materially increased the risk to the patient by, for example, making diagnosis of the appropriate treatment especially difficult, then I do not think the doctor can be said to have contributed to their negligence in any way.

Focusing on the nature of the duty brings out an unusual feature of this case when compared with most situations where the claimant’s injuries are exacerbated by negligent hospital treatment, such as the Rahman case. Typically it is the negligent act of the initial wrongdoer which causes the claimant to have to see a doctor whom he would not otherwise have to see and does not want to see. It is not difficult in those circumstances to conclude that justice will often require the initial wrongdoer to remain liable for all the foreseeable consequences of the negligence, even where they include the exacerbation of the injuries resulting from the negligent but foreseeable acts of doctors. In this case, however, the very purpose of the doctor’s duty is where necessary to present the patient to the specialist in time to enable the patient to be properly treated. Where the doctor has done that, albeit with culpable delay, it seems to me unjust to make him liable for the hospital’s negligent treatment. The doctor has ensured that the patient is where he needs to be and wants to be. The doctor’s negligence is to present the patient later than he should have done, but I do not think that it is justified to treat the permanent injury which would have been avoided but for the hospital’s negligence as attributable to him. The doctor should not by his negligence take on the role of guarantor of the hospital’s negligence. He should not be held responsible for that injury.

As I have indicated, I recognise that there will be cases, of which this is one, where the delay will limit the opportunity of the hospital to put right any negligent treatment before it is too late. It seems to me that if the doctor is to be liable, this has to be the route by which he is made so. I see force in the analysis of the Master of the Rolls and Lady Justice Smith that this is enough to impose liability, but I am not ultimately persuaded that it is. I do not think that this limited additional risk, which is only triggered because of the negligence of the hospital itself, justifies the imposition of liability whenever the hospital is negligent. The hospital doctors are more specialised, hence the reason for referring a patient in the first place. But for their negligence, the claimant would not have suffered the permanent damage at all. It seems to me that on any view they would have to bear the lion’s share of any joint liability, almost to the point of extinguishing the referring doctor’s liability. I think that there is nothing inappropriate or unjust in pinning liability exclusively on the hospital alone in those circumstances.

I think there are additional policy considerations which support this conclusion, although I do not put them at the forefront of the analysis. To import liability in the circumstances might encourage doctors to play safe and refer cases to the hospital when there is no need to do so. Conversely, it might encourage hospitals (or their insurers) when sued to bring the doctor into the proceedings in the hope of being able to establish a delayed referral so as to obtain a contribution to any liability.

For these reasons, I would dismiss the appeal. I would only add that I agree with my Lord, the Master of the Rolls, that is highly unfortunate that the hospital was not made a party to these proceedings from the beginning.

I have read the judgment of the Master of the Rolls in draft and gratefully adopt his exposition of the facts of this appeal. I agree with his conclusion that the appeal must be allowed for reasons which are essentially the same as his, although I am not wholly in agreement with his reasoning. As the contentious issues are not central to the decision of any member of the court, I have not dealt with my reservations in any detail. I have also read the judgment of Elias LJ. I agree with his analysis of the law, in particular his observations relating to Gouldsmith v Mid-Staffordshire General Hospitals NHS Trust [2007] EWCA Civ 397. However, as Elias LJ has reached a different conclusion as to the result of the appeal, I will explain in my own words my reasons for concluding that the appeal should be allowed.

The essential facts are that the defendant general practitioner (GP) negligently failed to examine Clarice on Wednesday 15 April 1998. At that time, she was suffering from a bacterial super-infection which had seeded into her femur and had given rise to osteomyelitis. If the GP had examined Clarice on that day, he would or should have appreciated that she was ill and would or should have referred her to the South Cleveland Hospital (the hospital) in the late afternoon or early evening of that day. As it was, Clarice was not referred to the hospital until the evening of Friday 17 April, by which time her condition had worsened. By that time, the infection of the femur had led to an effusion in the hip joint, which was, at about the time of her admission, becoming infected. If, on admission, Clarice had been treated promptly with an appropriate antibiotic, she would probably have made a complete recovery from these conditions. She might have had to undergo the surgical drainage of the effusion; she might not. But there would probably have been no permanent damage.

The trial judge, Mackay J, heard much evidence about the treatment Clarice received at the hospital. His factual conclusions are not challenged. In summary the position was that, initially, she was prescribed an antibiotic which would, if continued, have been effective. However, in the afternoon of Saturday 18 April, the registrar in charge of her care, who was of the mistaken opinion that her illness was due to gastroenteritis, changed the antibiotic to one which was not in fact effective against her infection. Although her condition did not improve - indeed it deteriorated – she was kept on that antibiotic until Monday 20 April when she was seen for the first time by a consultant paediatrician. In the light of the blood culture result, which had been available since 19 April, she prescribed an effective antibiotic. However, the true nature of Clarice’s condition was not appreciated until the following day (21 April) when an ultrasound scan revealed the effusion and the infected hip joint. Surgical drainage was carried out in the early hours of 22 April but the appropriate treatment had been given too late and Clarice had by then suffered severe and permanent damage to the hip joint.

At paragraph 39 of his judgment, the Master of the Rolls has explained the effect of various degrees of delay. The expert evidence was that if administration of the appropriate antibiotic had been begun on 16 April (or a fortiori in the evening of 15 April), the progress of the infection would probably have been halted. As I understand it, there would have been a complete cure without the need for surgical drainage. If appropriate treatment had been commenced on the 17 April, the infection might have been halted but it might have proceeded, in which case surgical intervention would have been necessary; however, permanent damage would have been avoided. If commenced on 18 April, the antibiotic would probably not have been effective without surgical drainage, but if such drainage had taken place on that day, there would have been an effective recovery but there may have been some permanent articular cartilage damage and there would also have been a risk (albeit low) of the avascular necrosis and growth disturbance which the claimant has actually suffered. If the appropriate antibiotic had been commenced on 19 April, it would have been ineffective without surgical drainage but, even if that had been done on the same day, there would have been a significant risk of articular cartilage damage with avascular necrosis and growth disturbance. By 20 April, the severe permanent damage had become inevitable. Thus, it seems to me that the delay in treatment created a gradually increasing risk of an adverse outcome. Mackay J did not describe this gradually increasing risk but merely stated that the administration of an appropriate antibiotic immediately or soon after admission on the evening of 17 April would probably have led to a good outcome. On the appeal before us, it was accepted that prompt treatment with an appropriate antibiotic at the hospital would have resulted in recovery with no permanent damage and no submissions were addressed to the gradually increasing risk of an adverse outcome.

In the light of the evidence called at trial, and with the consequent benefit of hindsight, it is hard to understand why Clarice’s advisers sued only the GP practice and did not join the hospital in the same action. Had they done so, the claimant may have been able to demonstrate that both defendants had been negligent and could have then left it to them to fight out (or agree) the issues of causation and/or contribution. In any event, it seems most unlikely that she could have found herself in the position of losing altogether.

As it was, Clarice sued only the GP practice and alleged that the delay in referral to hospital had caused the permanent damage to the hip. The GP’s stance was that he should not be held liable for anything other than the consequences of the delayed referral. The delay had not caused the permanent damage. Clarice had been referred to the hospital in time for her to be given effective treatment. It was the hospital’s fault that her condition was allowed to deteriorate to the point where permanent damage occurred.

Thus it was that the defendant GP set about persuading the judge that it was the hospital’s failure to provide prompt and appropriate treatment which had caused the relevant damage. The judge was very critical of the standard of care provided by the hospital but decided not to categorise the failures as ‘negligence’ because the hospital was not before the court. In passing, I think the judge was right in that regard because ‘negligence’ connotes legal responsibility and it was not open to the judge to determine legal responsibility. The judge’s findings could not bind the hospital. But, within the context of the action as formulated, the effect of the judge’s findings was supportive of the defendant’s position. He was of the view that it should have been a relatively straightforward matter for the hospital staff to diagnose and treat Clarice’s condition promptly. They had failed to do so and had instead allowed Clarice to deteriorate so as to suffer permanent damage.

In the light of the expert evidence, it was plainly difficult for the claimant to deny that treatment given after admission on 17 April had been inadequate. Her case was that, if the GP had referred her to the hospital on the Wednesday afternoon instead of on Friday evening, the hospital team would have treated her competently and effectively. Staffing levels would have been different during the week and in particular there would have been a consultant on duty. So, argued the claimant, the delay in referral had caused the permanent damage. The judge agreed, somewhat reluctantly at first, to embark upon the hypothetical and highly speculative consideration of what would probably have happened at the hospital if the claimant had arrived there in the late afternoon of the Wednesday. He had formed a very unfavourable view of the way in which the paediatric unit functioned. He had been unimpressed by the care the claimant had received on an earlier admission between the 9^th and 12^th April. He was also deeply unimpressed by the care provided between 17^th and 21^st April. He concluded that there were systemic failings within the paediatric department such that he should infer that, if the claimant had arrived at the hospital on the Wednesday, the team would probably have failed to provide adequate treatment. Therefore, the delay had made no difference to the outcome and the GP was not liable for the permanent damage. It should be noted that the effect of this finding is that whenever the claimant had arrived at the hospital, the team would probably not have provided prompt and adequate treatment.

The Master of the Rolls has held that the judge was wrong to draw the inference that the hospital would have failed to provide adequate treatment on the Wednesday. I agree with him for the reasons he has given at paragraphs 68 to 78 of his judgment. Elias LJ also agrees. In my view, where the condition in question is not difficult to diagnose and the correct treatment is simple and uncontentious, there must be a presumption that a hospital will probably provide a correct diagnosis and treatment. The Master of the Rolls is of the view that that presumption cannot be rebutted as a matter of law. I am uncertain whether I agree with that conclusion but that matters not because I agree with him that it has not been rebutted on the facts of this case.

It must be accepted that there were serious failings in the provision of care over the weekend of 18/19 April. In particular, the claimant was not seen by a consultant as she should have been. But I do not think it can be inferred that she would probably not have been seen by a consultant on Thursday 16 April if she had been admitted the previous late afternoon or evening. Nor do I think it can properly be inferred that the consultant who would probably have seen her would have failed to diagnose her condition. If, as I would hold, the presumption of competent treatment is not rebutted, it follows that the claimant would probably have been properly treated if she had been admitted on the 15 April. Thus she has established a causal link between the negligence and her loss.

That does not mean however that the claimant necessarily succeeds in establishing liability for her permanent injury. Establishing a causal link between negligence and damage does not necessarily mean that liability will follow. One still has to decide whether the defendant ought to be liable for the damage in question. Lord Nicholls of Birkenhead explained the position in Kuwait Airways Corporation v Iraqi Airways Co [2002] 2 AC 883 at page 1091:

“69.

How, then, does one identify a plaintiff’s “true loss” in cases of tort? …. I take as my starting point the commonly accepted approach that the extent of a defendant’s liability for the plaintiff’s loss calls for a twofold inquiry: whether the wrongful conduct causally contributed to the loss and, if it did, what is the extent of the loss for which the defendant ought to be held liable. The first of these inquiries, widely undertaken as a simple “but for” test, is predominantly a factual inquiry. ….

70.

The second inquiry, although this is not always openly acknowledged by the courts, involves a value judgment (“oughtto be held liable”). Written large, the second inquiry concerns the extent to which the defendant ought fairly or reasonably or justly to be held liable (the epithets are interchangeable). To adapt the language of Jane Stapleton in her article “Unpacking Causation” in Relating to Responsibility ed Cane and Gardner (2001) p 168, the inquiry is whether the plaintiff’s harm or loss should be within the scope of the defendant’s liability, given the reasons why the law has recognised the cause of action in question. The law has to set a limit to the causally connected losses for which a defendant is to be held responsible. In the ordinary language of lawyers, losses outside the limit may bear one of several labels. They may be described as too remote because the wrongful conduct was not a substantial or proximate cause. The defendant’s responsibility may be excluded because the plaintiff failed to mitigate his loss. Familiar principles, such as foreseeability, assist in promoting some consistency of general approach. These are guidelines, some more helpful than others, but they are never more than this.

71.

In most cases, how far the responsibility of the defendant ought fairly to extend evokes an immediate intuitive response. This is informed common sense by another name. Usually, there is no difficulty in selecting, from the sequence of events leading to the plaintiff’s loss, the happening which should be regarded as the cause of the loss for the purpose of allocating responsibility. In other cases, when the outcome of the second inquiry is not obvious, it is of crucial importance to identify the purpose of the relevant cause of action and the nature and scope of the defendant’s obligation in the particular circumstances. What was the ambit of the defendant’s duty? In respect of what risks or damage does the law seek to afford protection by means of the particular tort? …”

The judge below did not undertake this two-staged inquiry because he concluded that there was no causal link at all between the delay in referral and the permanent damage to Clarice’s hip. That was because he was satisfied on the balance of probabilities that the hospital team would have failed to treat her properly if she had been admitted on the Wednesday, just as they did when she was admitted on the Friday. However, in the light of the conclusion that the judge was wrong about this – a conclusion on which all three members of this court are agreed – the claimant has established a causal link between the delay and the damage. She has established that the GP’s delay in referral was a cause of her permanent disability. It was not the only cause because, on the judge’s findings, it is clear that the hospital team’s failure to treat her appropriately was also a cause. Thus, the GP’s liability will depend upon the value judgment described by Lord Nicholls as the second stage of the inquiry. Is it fair, just and reasonable for the GP to be held liable for the permanent damage to Clarice’s hip?

Lord Nicholls suggests that sometimes the resolution of the issue will be a simple matter of common sense applied to the sequence of events. In other cases, it will be necessary to examine the nature and extent of the defendant’s duty and the purpose for which the cause of action against him exists. I make my value judgment on the following basis.

I have already explained how and why there was a causal link between the GP’s negligence and the permanent damage to Clarice’s hip. If the GP had referred Clarice on 15 April, as he should have done, she would probably have been treated effectively and would have suffered no long term harm. However, it was not within the scope of his duty to treat Clarice, only to refer her as expeditiously as was practicable so as to reduce her period of suffering and to give her the best possible opportunity for recovery. In the event, he referred her to the hospital just over two days later. That was a serious error. Not only did it lengthen the period in which Clarice was ill (for which unnecessary suffering only the GP could be liable) but it also had the effect of reducing the period available to the hospital to provide correct and effective treatment.

The nature and extent of the hospital’s duty was for the staff to exercise proper professional skill and care in the diagnosis and treatment of Clarice’s condition. On the judge’s findings, this diagnosis was not particularly difficult and the treatment was readily available. The hospital team had time to diagnose and treat Clarice so as to avoid any permanent harm. That being so, one might say that it was not foreseeable that the hospital team would fail to take appropriate action. On the judge’s findings, their failure to do so entailed a significant degree of incompetence and was the direct and immediate cause of Clarice’s deterioration and the resulting permanent harm. One might say that the permanent harm was so closely and directly related to that incompetence and so loosely related to the delay in referral that the permanent damage was too remote from the GP’s negligence to found liability.

I have found this value judgment a difficult one. In the end, I think that the crucial point is the extent to which the GP’s negligence shortened the period available to the hospital to provide effective treatment and increased the risk that she would suffer permanent harm. As I observed in paragraph 92 above, the judge did not consider in detail the evidence of gradually increasing risk of permanent harm which delay in treatment would cause. Nor did he consider the extent to which the GP’s delay shortened the window of opportunity available to the hospital team. The expert evidence showed that the risks of permanent harm were gradually increasing as time went by. Although the expert opinions are not crystal clear, it looks to me as though the doctors were saying that, to avoid permanent harm, Clarice needed a correct diagnosis and treatment on 18 April. Any further delay beyond that day would probably have resulted in some permanent harm although not necessarily of the degree actually suffered. Thus the delay of just over two days from, say, 6pm on 15 April to 10pm on 17 April severely curtailed the window of opportunity for the hospital team to diagnose and treat Clarice appropriately. Very roughly, the delay used up about two-thirds of the window available to effect a complete cure and about half of the window available to secure an appreciably better outcome. It was still possible for the team to treat her successfully but they had a much reduced opportunity and a much reduced margin for error. That being so, I think one can properly say that the GP’s breach of duty had an ongoing effect on events which was not completely wiped out by the hospital’s failure to act correctly.

I regard this case as being very close to the line but I have come to the conclusion that, because of that ongoing effect, it cannot be said that the permanent damage was too remote from the delay. The failures of the hospital team did not wipe out the effect of the delay in referral. In my judgment therefore it would be fair, just and reasonable for the GP to be held liable for the permanent harm. I have little doubt that justice will require that the hospital makes some contribution to that liability, probably a large one, but I have come to the conclusion that the GP cannot altogether escape.

For those reasons, I would agree with the Master of the Rolls that the appeal should be allowed.