England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM The Hon Mr Justice Toulson

		Royal Courts of Justice Strand, London, WC2A 2LL Friday 8th June 2001

LORD JUSTICE MAY:

Introduction

This is the judgment of the Court prepared by May LJ.

1. This appeal from the judgment of Toulson J of 13^th June 2000 raises difficult questions of importance and great sensitivity.
2. The claimant, Michelle Gray, is now aged 25. She is a patient as a result of matters which are the subject of this litigation. She was born on 5^th September 1975. From the age of about 2 she had occasional fits. These were brought under control by anti-convulsant treatment. In October 1983 she was seen for general review by a consultant paediatrician, who noticed that her left arm was larger than her right arm. Investigations followed, which revealed that she had a tumour in the left fronto-parietal region of her brain. Mr Pickard, a consultant brain surgeon, told her parents that the tumour would probably prove fatal unless she had an operation. They consented to this and the operation took place on 23^rd July 1984 at Southampton General Hospital, when Michelle was aged 8. Mr Pickard performed the operation and the anaesthetist was Dr Read, then a senior registrar and now a consultant. The tumour was partly removed. It was at first thought to be malignant, but it was eventually determined as being a non-malignant ganglioneuroma, which has reduced in size over the years and has been by itself of little lasting clinical significance.
3. The operation started at 13.20 and was successfully completed by 15.45. Michelle woke normally from the anaesthetic and had the use of all her limbs. But at 16.42 she had focal right sided fitting possibly followed by a general grand mal fit. Following this, she received intensive care treatment for a number of days, the first part of which we will describe in greater detail later in this judgement. At some indeterminate stage on or after 25^th July 1984, it was discovered that she had suffered very serious permanent brain damage. She was discharged from hospital on 12^th October 1984. She could not walk, sit unsupported or take any food or fluid by mouth. Towards the end of 1984, she began to be able to walk with assistance but did not obey commands and could not talk. Her condition deteriorated and she was readmitted to hospital at the end of 1984. She had an operation for the insertion of a shunt from the brain into the abdomen. She continued to have minor seizures. During this period, her mother acted as her main carer with the help of her father and her maternal grandmother. Michelle's present condition is stable. She needs help with all functions including dressing and bathing. She is a poor sleeper and needs attention most nights on at least two occasions. She has a wheelchair and needs to use it when going out. She can walk about her home but only when supervised. She has never spoken nor responded to the spoken word. She is wholly dependent on her carers and is and will remain doubly incontinent. There will never be any improvement in her condition. It is agreed that her life expectation is approximately another 25 years. Her mother is now aged nearly 57 and her father nearly 48. The burden upon them is very considerable. They look after her with devoted care.
4. These proceedings against the health authority were begun by writ dated 21^st August 1996, twelve years after the operation. Michelle was for 10 years a minor and is a patient, and so the question of limitation does not arise. The trial took place 16 years after the operation. This unavoidably affected the recollection of factual witnesses. In addition, some hospital notes had been lost or destroyed. Quantum of damages was agreed, subject to liability, in an inevitably very high sum.
5. The issue which the judge had to decide was whether it had been established that Michelle's brain damage was caused by want of proper management of her condition during the post operative period when she was under the care of Dr Read. It was Michelle's case that a critical cause of the brain damage was systemic hypoxia and/or hypotension, which ought not to have been permitted. The defendants' case was that the effective cause was status epilepticus. The judge found in Michelle's favour as to the cause of her brain damage. But he held that as a matter of probability it had not been established that this was caused by want of proper management on the part of the defendants. He accordingly dismissed Michelle's claim. This is her appeal against that decision. She brings it with permission granted by the judge. He said that he had found it an exceptionally difficult case in which he had expressed a number of misgivings about the way matters were conducted during Michelle's stay in hospital. Lack of evidence drove him to his conclusion. He considered that there was scope for reasonable argument upon which another court might reasonably come to a different view. His critical conclusion was not a primary finding of fact but was based on an evaluation of the evidence as it stood. He said that in that exercise an appellate court would be practically as well placed as the trial judge.

The Facts and the Trial

6. The records for the operation itself are good, including a detailed anaesthetic record made by Dr Read. No criticism is made of the operation and the experts have concluded that Michelle's brain damage did not occur during that period. She was transferred from the operating theatre to a recovery room at about 15.50. Her blood pressure, heart rate and respiration were recorded on arrival there and then at approximately 15 minute intervals. Nothing untoward seems to have happened until, as we have said, she suffered a fit which may have constituted a grand mal seizure at 16.42. Mr Pickard and Dr Read decided to ventilate her to avoid hypoxic complications. Dr Read re-anaesthetised and re-intubated her, and she was then transferred to a neurological ward called "E Neuro" where she was placed on a ventilator. This ward had a small number of beds for patients requiring intensive care, with the necessary nursing staff and equipment for managing ventilated patients. She remained there under Dr Read's responsibility until 21.45, when she was transferred to the general intensive care unit. Her case is that the serious permanent brain damaged occurred during the period when she was on E Neuro ward. During this period, the record keeping was described by the judge as "frankly inadequate". The judge described the records as consisting of:

"1. An observation sheet which recorded the administration of drugs and, at hourly intervals, the setting of the ventilator, her pulse, blood pressure and condition of her pupils.

2. A nursing note, which included these entries:

"On return to ward [from recovery room] R sided focal fit reoccurred commencing at mouth and extending down arm. Intravenous phenobarbitone given. Also to begin with very difficult to sedate gagging on intubation tube … Has continued to have focal fits over evening. S/b [seen by] anaesthetist for transfer to INCU. Chest sounds clear. On suction oral secretions excessive. Catheter draining constantly."

3. A Department of Anaesthetics sheets for recording the management of intubated patients, which contained two short entries by Dr Read, the first at about the time when she was put on the ventilator and the second at about the time when she was transferred to the ICU.

4. The blood gas readings timed at 20.00 and 21.10 to which I have referred.

5. A drug prescription and administration sheet."

7. At the time of her first fitting at 16.42, Dr Read prescribed and administered 125mg of thiopentone, 500 µg of fentanyl and 25mg of suxamethonium. It is agreed that she suffered three further fitting episodes while she was on E Neuro ward. For these, Dr Read administered 50mg of phenobarbitone at some time between 17.00 and 17.30, 100mg of phenobarbitone at 19.09 and 100mg of phenobarbitone at 20.40. These times are or approximate to the times of the three fitting episodes. Ventilator readings were recorded at hourly intervals between 17.00 and 21.00. Her blood pressure was recorded at 18.00, 19.00 and 20.00 only. These recorded blood pressure readings were unremarkable. Dr Read's oral evidence was that he took Michelle's blood pressure on occasions other than those which were recorded and that on each occasion the result was unremarkable. These occasions included times after the administration of drugs to control the fitting when there was a risk that her blood pressure would drop. He was asked questions about this evidence in cross examination, but it was not suggested to him that in this respect he was mistaken. Dr Read took two blood gas samples at times which he recorded as 20.00 and 20.10, although he doubted in his evidence whether these timings were accurate because for practical reasons there must have been a greater interval between them. He had to take them to another part of the hospital for analysis and described where this was and how long it would have taken. He thought that the first of the two blood gas samples was probably taken at around 19.50. The blood gas tests showed that Michelle was significantly acidotic. The base excess was –10.0 at 20.00 and –11.6 at 20.10. In addition, the PO₂ recording for 20.00 was 23.4, whereas this reading had fallen to 9.7 at 20.10.
8. Dr Read was evidently concerned at the results of the blood gas tests. For whatever reason, his note of them is timed at 20.45. As we have already said, the fourth of Michelle's fitting episodes occurred at 20.40. Dr Read thought that she had pulmonary oedema. He feared that this, with her metabolic acidosis, might result in heart failure, especially if, as he considered likely, further fits would require him to administer further drugs. Accordingly, he decided to transfer her to the general intensive care unit which was better equipped to deal with cardiovascular problems. The handover took place at 21.45.
9. She was closely monitored on the intensive care unit, as is apparent from the records which have survived, although much of the written notes for this period have disappeared. On arrival, her systolic blood pressure was recorded as 70. This was quite low, although the evidence was that it may have fallen during the transfer. The nursing notes for this period record that Michelle continued to fit when she was not sedated. The records also include blood gas results for 22.00 and 23.30. These indicate that she remained acidotic, although the base excess figures are rather lower than those in the earlier tests. There is also a remarkably similar drop in the PO₂ reading from 23.76 at 22.00 to 6.6 at 23.30.
10. As we have said, the claimant's case concentrated on the period when she was on E Neuro ward. Various allegations were made and modified during the course of the case. A number of allegations in the amended statement of claim did not survive to the trial or were refined. It was for a long time thought that Michelle's brain damage was properly described as watershed ischaemia. But this perception changed during the trial. The judge described it as follows:

"At the outset of the trial it was the claimant's case, based on CT and MRI scans, that there was accentuated damage in the "watershed" or border zone regions of the brain, indicative of vascular hypotension (low blood pressure) as the cause of the damage. But that proposition was not maintained after the evidence of the two neuro-radiologists. They agreed that the scans did not show watershed infarcts and that the radiological picture was one of global atrophy indicative of some global insult, rather than injury concentrated in those areas which would be particularly susceptible to injury through mild or moderate hypotension. In their view, for hypotension to have caused such damage it would have to have been profound, i.e. a perfusion failure equivalent to the consequences of a near cardiac arrest.

The possibility that there was such a hypotension episode was duly explored. Professor Hull, the claimant's expert anaesthetist, suggested that the most likely time for its occurrence would have been after Michelle's intubation in the recovery area, but he accepted in his evidence that the suggestion presented very great difficulties for various reasons, which it is unnecessary to detail because Mr Faulks QC at the end of the case, in my view very properly, did not seek to argue for a finding that there had been profound hypotension.

The claimant's case on causation ultimately depended on the opinion of her expert neurologist, Dr Bates. He identified four factors which might theoretically have accounted for Michelle's general and profound brain damage. These were:

1. the fact that after surgery she was liable to dis- autoregulation of her cerebral vasculature;

2. continuing electrical activity within the brain, increasing its metabolic requirements;

3. hypotension; and/or

4. hypoxia.

Dr Bates did not exclude the first as a contributory factor, but he excluded it as the sole cause, because if there had been profound disregulation of her cerebral vasculature she would have had profoundly raised intracranial pressure, which was not the case. As to the second, he accepted that if there had been profound and continuing generalised epileptic activity with spasms of all four limbs, this in time could have led to the pattern of brain damage which she suffered. But he was not persuaded that she had generalised continuing status epilepticus and, even if she had, he would not have expected such profound damage to her cerebral neurones if her respiration, blood pressure and heart rate had been properly maintained.

Dr Bates therefore concluded that the damage could only be explained by the presence in addition of relative (ie less than profound) hypotension and/or hypoxia. He considered that although there was not good evidence for either of those phenomena, there was some evidence of both. The level of metabolic acidosis apparent from the samples timed at 20.00 and 20.10 was in his view evidence of hypoxia or hypotension. Her systolic blood pressure reading of 70 shortly after admission to the ICU would imply a mean arterial pressure in the region of 50 to 60. The normal cerebral auto-regulation would fail when the mean arterial blood pressure fell below about 45, which would be equivalent to a systolic blood pressure around 60, but in a person suffering from a level of disautoregulation the system could fail with a lesser degree of hypotension. He believed that Michelle's brain damage was probably caused by a combination of all four factors and that it would not have occurred but for relative hypotension and/or hypoxia during her period on E Neuro. He did not consider that this conclusion was negatived by the recorded hourly observations that Michelle's pupils were of normal size and reactive to light, because pupil size is an indication of brain stem activity and Michelle has retained a normal brain stem despite grave brain damage. In his experience a patient who suffers a full cardiac arrest with most severe resulting ischaemic insult may regain pupillary reaction in a short time."

11. The defendants' expert neurologist was Professor Chadwick. He considered that the cause of Michelle's brain damage was status epilepticus and that any hypoxia was a secondary consequence of seizures giving rise to a metabolic demand within the brain in excess of the ability of the cerebral circulation to supply oxygen and glucose. Status epilepticus involves continuous generalised seizures, with visible tonic clonic seizures. It is possible for a patient to be treated with paralysing agents which stop the seizures of the limbs without stopping the abnormal electrical activity within the brain. Professor Chadwick believed that this may have happened with Michelle. He considered that the acidosis in the blood samples taken around 20.00 could be explained if Michelle had suffered recent generalised tonic clonic seizures. The sense of his evidence was that "recent" meant a period up to about 5 minutes before the blood sample was taken. This did not accord with the previous recorded episode of fitting at 19.09, but Professor Chadwick thought that she may have continued in a kind of status epilepticus for an indeterminate period. He did not believe that there was any evidence to establish systemic hypoxia or hypotension as a significant cause of her injury. For that to have been so, it would have to have been present over a significant period and would have produced unmissable signs. He acknowledged that his view depended on it being found that Michelle suffered a pattern of general convulsions while she was on E Neuro ward. If she had only suffered some focal fits which had been promptly controlled, they would not account for her acidosis or the gravity of her brain damage.
12. The judge considered Dr Read's evidence. Somewhat surprisingly, Dr Read did not know that Michelle had suffered severe permanent brain damage until some time fairly shortly before the proceedings were started. He himself would have expected to have been told at the time. Not surprisingly, he had difficulties of recollection. He made his first witness statement on 9^th July 1998. He made what turned out to be a number of mistakes in this statement, and the judge was not surprised about that. Significantly however, the statement referred to Michelle having focal fits on E Neuro ward which he controlled by phenobarbitone, but he made no reference to her having any general convulsion after her fit in the recovery room recorded in the nursing notes as possible grand mal. Dr Read subsequently made two further written statements in which he gave an account of Michelle having a number of grand mal fits and he repeated that in his oral evidence. He said that Michelle's fit in the recovery room, which began as a right focal fit, degenerated into generalised tonic clonic seizures. He said that the dose of 100mg phenobarbitone which he administered at 19.09 was for a major epileptic seizure. On this occasion Michelle did not start with a right sided focal seizure but went straight into full tonic clonic seizure. The judge analysed his evidence about the blood gas samples at 20.00 and 20.10. His evidence was that the episode for which he administered a further dose of 100mg phenobarbitone at 20.40 was a further grand mal fit.
13. The judge was in no doubt but that Dr Read was an entirely honest witness doing his best to remember and reconstruct events nearly 16 years ago. However, his evidence of repeated general convulsions was in marked contrast with the contemporaneous evidence and with his own memory when he made his first witness statement. The judge said that Dr Read undoubtedly now believed in his own recollection as given in his evidence, but the question was whether that belief was accurate. The only contemporary reference to a grand mal fit was in the entry in the nursing notes relating to Michelle's period in the recovery room. The judge said that it was very difficult to believe that a nurse trained to work on the intensive care part of a neurological ward would have mistaken general tonic clonic seizures for focal fits. The judge also said that it was unlikely that the general convulsion would have resulted in serious acidosis one hour later. The judge said that he had to consider whether it was more likely that the description of Michelle's fitting made at the time by the E Neuro nurse was accurate and that Dr Read by an honest but mistaken process of reconstruction had built in his own mind a different picture; or that the nurse misdescribed the fitting and Dr Read's present memory was accurate. Of these two possibilities, the judge believed that the first was much the more likely. He accordingly did not accept Professor Chadwick's theory. He did accept Dr Bates' opinion that Michelle's brain damage was caused by a combination of factors including some degree of systemic hypoxia and/or hypotension, without which it would not have occurred. He concluded that "there must have been an episode of relative hypotension and/or hypoxia". Although Mr Grace QC, leading counsel for the health authority, submitted that the judge did not find (or did not clearly find) that the relative systemic hypoxia and/or hypotension occurred while Michelle was on E Neuro ward, in our view the plain sense of the judgment is that the judge did so find.
14. The issue of negligence therefore became whether it was established that Dr Read, exercising due care, ought reasonably to have detected that Michelle was suffering from relative hypoxia and/or relative hypotension and to have corrected this so as to prevent the severe permanent brain damage which in fact occurred.
15. The judge said that whether there was negligence was the hardest question of all. There was no doubt that Michelle's condition required intensive monitoring. That was why she was put in an intensive care bed. He recorded the argument advanced by Mr Faulks QC on behalf of Michelle as having two strands. First, he submitted that the lack of proper records, although this did not constitute causative negligence in itself, was indicative or suggestive of a lack of proper care in Michelle's management. Dr Read had failed to keep a proper record of her care on E Neuro ward and had failed to take a blood gas reading until she had been on the ward for about 3 hours. Secondly, Mr Faulks submitted that the fact that Michelle had relative hypotension and/or hypoxia itself implied a want of proper management. That evidence, taken with the evidence as a whole was sufficient to enable the court to infer causative negligence in circumstances where the health authority had failed adequately to explain how the relative hypotension and/or hypoxia came about without their negligence. Dr Read had provided explanations for the comparative absence of records and for the fact that he had not taken blood gas tests until 20.00. Of these criticisms the judge said:

"The expert anaesthetists, Professor Hull and Professor Adams, agreed at a meeting that the blood gases should have been checked within 1 hour of Michelle commencing controlled ventilation on E Neuro, i.e. not later than 18.00. Although Professor Adams at one stage in his evidence appeared to retreat from that position, he ultimately confirmed it. I accept and conclude that a blood gas reading should have been taken long before 20.00, but it is entirely a matter of speculation what it would have shown. Professor Hull and Professor Adams also agreed that blood pressure and heart rate observations should have been made and recorded at not greater than 10 minute intervals. In addition to those matters, the quality of Dr Read's notes on E Neuro (in contrast to his operation note) was below the standard acceptable in 1984. I do not believe that this can truly be attributed to the continual pressure of events unless, contrary to Dr Read's present recollection, he was having to go between different patients."

16. The judge said that these shortcomings provided legitimate grounds for criticism, but that they also presented a paradox. Dr Read had given the impression of being conscientious with a tendency in his evidence to try to dot the i's and cross the t's. The shortcomings did not fit easily with the appearance of a conscientious doctor of high standards. The shortcomings gave rise to a natural suspicion that for some reason Dr Read may not have been attending to Michelle as closely as he now recalls, possibly because of the demands of other patients. The judge pointed out that Dr Read had mistakenly assumed in his first witness statement that he would have taken blood gases at a much earlier stage than he in fact did. It might have been that it was not until Dr Read took the blood gas readings and saw to his surprise the level of acidosis that he began to become concerned about her. Until then, she had had a couple of focal seizures which he had quickly controlled by relatively modest amounts of phenobarbitone. "But," said the judge, "when so many years have passed after the event, it is very difficult to be definitive about any of these matters." The judge then said:

"As to the second strand of Mr Faulks' argument, I have concluded that there must have been an episode of relative hypotension and/or hypoxia. The difficult questions are how long it must have lasted, how severe it must have been, what signs it must have caused and, taking these matters together, whether it therefore implies negligence on the part of the defendants. The evidence does not provide any very exact answers to these questions. As to length, the duration according to Dr Bates would have to have been for several minutes, but that is a fairly imprecise range. As to severity, Dr Bates understandably did not attempt to give exact figures for the degree of hypoxia or hypotension necessary to have produced the documented acidosis when taken in conjunction with the other causative factors. The lowest recorded blood pressure was 70 systolic shortly after admission to the ICU, and that would have shown itself in pallor, but her blood pressure may well have been higher before she left E Neuro, since the process of moving her could tend to produce a lowering of blood pressure. Between the two blood gas samples taken on E Neuro the oxygen pressure in Michelle's arterial blood fell to 9.7 kPa which was low, but in the ICU it fell to 6.6 kPa at 23.30, and it was not suggested that her management in the ICU was faulty. As to signs, in an otherwise normal patient, for hypotension or hypoxia to have caused Michelle's damage it would have to have caused signs unmissable by anybody qualified and present, but Dr Bates did not suggest that the same would be so in Michelle's particular situation.

It has to be remembered that the combination of inter-active factors in Michelle's case was unusual. The more I have considered the evidence, the more I have been driven to the final conclusion that, despite misgivings about the shortcomings to which I have referred and despite profound sympathy for Michelle and her parents, I have insufficient evidence of the nature, degree and duration of the relative hypotension or hypoxia which she must have suffered to justify a finding as a matter of probability that it was caused by want of proper management on the part of the defendants."

17. The judge accordingly dismissed Michelle's claim.

Grounds of Appeal and Submissions

18. The written grounds of appeal essentially are that the judge's finding on negligence was against the weight of the evidence. It is said that he attached insufficient weight to the fact that Michelle suffered an episode of relative hypotension and/or hypoxia which Dr Read failed to observe or otherwise detect. It is said that the judge should have rejected Dr Read's evidence that he was closely monitoring Michelle, since if he had been doing so he should have reasonably observed or otherwise detected changes in her colour, appearance, blood pressure, blood gases and/or level of oxygenation. There was ample evidence upon which the judge could and should have found that Dr Read's management or monitoring of the claimant was deficient. This included the agreed evidence that Dr Read should have tested Michelle's blood gases not later than 18.00; that Dr Read should have taken her blood pressure and heart rate at not greater than 10 minute intervals; and that the judge should have rejected Dr Read's evidence that he was checking her blood pressure frequently, since it was unsupported by medical records, for the same reasons that he had rejected Dr Read's evidence that Michelle had suffered repeated general epileptic convulsions. Michelle's level of oxygenation was capable of ready monitoring and adjustment. The judge should have inferred from these facts and the fact that the records were "frankly inadequate" that Dr Read's standard of care was causatively negligent. The facts established were a sufficient basis for this inference in the absence of proper explanation by the health authority, which they did not give.
19. By respondent's notice, the health authority seek to uphold the judge's decision on the basis that there was no evidence on which the judge could have found that any failure to prevent, recognise or treat relative hypoxia or relative hypotension amounted to a breach of the defendant's duty of care. They further contend that, in so far as the judge may have found that Michelle's brain damage occurred whilst she was on E Neuro ward, there was no sufficient evidence to justify such a finding. They yet further contend that the judge's finding as to the cause of Michelle's brain damage, that is his acceptance of Dr Bates' evidence, was wrong. It is contended that the judge misunderstood Professor Chadwick's thesis, wrongly supposing that it depended entirely on it being found that Michelle suffered generalised convulsions while she was on E Neuro ward. In fact, Dr Chadwick's thesis relied (or relied also) on the fact that she had suffered convulsions after being transferred to the intensive care unit. It is said that the judge should have found that the primary cause of the brain damage was convulsions or fitting.
20. Mr Faulks submitted to the court that Michelle was damaged in circumstances which called for an explanation. The explanation provided was rejected. The only proper conclusion for the judge to have reached was that the damage to Michelle was not consistent with the appropriate standard of care to be expected at the time from an anaesthetist and skilled neurological nurses attending her on a one to one basis on a neurological intensive care ward. The judge should not have been deterred from reaching that conclusion by the lack of overt evidence of hypoxia and/or hypotension. Such an omission was itself the result of poor record keeping on the part of the defendants.
21. Mr Faulks submits that Dr Read should have made blood gas tests much earlier than he did, and not later than 18.00. Her blood pressure should have been taken at 10 minute intervals. Appropriate drugs were available to raise her blood pressure if this was necessary. In so far as there was evidence suggesting that earlier blood gas tests would probably have been normal, that evidence should have been rejected as having no basis. There must on the judge's findings have been an unrecorded episode or episodes of relative hypoxia and/or hypotension. The purpose of intensive care was to monitor and detect such episodes. They were not detected, but they should have been. If they had been detected, they could and would have been corrected. The shortcomings which were established, although they were not causative in themselves, were indicative of the poor standard of care, which more readily enabled appropriate inferences to be drawn. Although there was no direct evidence of relative hypoxia and/or hypotension, and although the timing of Michelle's brain damage was conjectural, there was one piece of hard evidence about her condition in E Neuro ward. This was the evidence of metabolic acidosis at around 20.00. It would have been perfectly permissible for the judge to have concluded on the balance of probabilities that an earlier blood gas reading and appropriate response to it would have prevented the degree of acidosis found at 20.00 and would have avoided the damage. The judge ought so to have inferred.
22. Mr Grace on behalf of the health authority emphasises what he submits was a critical shift of emphasis in the claimant's case during the trial. The case was opened, in reliance on the written report of Professor Hull and the then understanding of the CT and MRI scans, on the basis that she had suffered an episode of profound hypoxia and/or hypotension. Professor Hull's preferred view was that this may have been associated with Dr Read's administration of fentanyl following the fitting episode in the recovery room at 16.42. Profound hypotension and/or hypoxia was abandoned on about the third day of the trial after the radiological evidence had agreed that the MRI scans did not show watershed ischaemia and because it was accepted that an episode of profound hypotension and/or hypoxia simply could not have been missed by a trained intensive care team. Dr Read had already given all his evidence by this stage. Mr Grace submitted that the evidential case never properly addressed the adjusted case of relative hypoxia and/or hypotension. In the result, there was no evidence to support the claimant's eventual case. Mr Faulks was offered, but declined, the opportunity to call additional evidence towards the end of the hearing. Mr Grace further made submissions in support of the other grounds in the respondents' notice.

Discussion and Decision

23. In our judgment, the submission that there was no evidence to support the claimant's eventual case is not made out. Although the emphasis of the case went through various shifts, and although Professor Hull's written report was certainly predicated on a condition of watershed ischaemia caused by profound hypoxia and/or hypotension, Dr Bates' opinion had from the start accorded with the eventual case. The final sentence of his report dated 9^th November 1998 reads:

"It is therefore most likely that the continuing seizure activity causing an increase in the metabolic requirement of the cerebral tissues, together with relative hypoxia and hypotension and possibly affected by the altered inter cerebral circulatory auto regulation following surgery is the cause of the permanent damage to Miss Gray."

24. That this was caused by negligent management on the part of the health authority was, in our view, an available evidential case. We would therefore reject the first ground relied on in the respondents' notice.
25. Before the trial, the respective experts of each discipline met and certain agreements or disagreements were reached and recorded. These meetings took place before the change of emphasis in the claimant's case. Professor Hull and Professor Adams, the expert anaesthetists, in particular held their discussions in the context of the then case that the claimant had suffered profound hypotension or hypoxia. Nevertheless, certain of their conclusions remain relevant. They agreed that the use of fentanyl was appropriate in the context so long as the effect on blood pressure was monitored closely. 500µg of fentanyl was a very large dose for a 35kg child. If such a dose were given quickly to such a patient, who was recovering from anaesthesia and major surgery, profound hypotension could occur. This would be more likely to happen if the patient was hypovolaemic. The experts could not say whether hypotension did follow the injection of fentanyl for lack of clinical records. They suggested that Dr Read should be asked how rapidly he administered the fentanyl and what measurements of blood pressure he made following the injection. As to the first question, Dr Read answered that he administered the fentanyl in two parts and that the whole dose would have been administered within a period of five minutes. As to the second question, he said that her blood pressure would have been checked as soon as possible. This was parenthetically one of the occasions on which Dr Read said that he took Michelle's blood pressure, although it was not recorded, and that it was normal. The expert anaesthetists agreed that in the absence of hypotension there was no indication to alter fluid therapy or to administer a plasma expander, blood or blood product. They agreed that, in such an unstable patient, blood pressure and heart rate observations should have been made and recorded at not greater than 10 minute intervals. They agreed that blood gases should have been checked not later than 18.00. If the blood gases had then been normal, there would have been no indication for a further check for at least 6 hours unless abnormal clinical signs developed. They agreed that the results at 20.00 and 20.10 were highly abnormal and indicated a metabolic acidosis. They agreed that the clinical records and witness statements did not provide an explanation for the very significant metabolic acidosis. Its possible causes were circulatory arrest, gross and prolonged hypotension, hypoxaemia, or any combination of those. It is now accepted that there was no basis for concluding that Michelle suffered circulatory arrest or gross and prolonged hypotension. The same, we think, has to apply to hypoxaemia, since the sense of the agreement is that this, too, had to be gross and prolonged. In the light of their then understanding that Michelle's brain damage constituted watershed cerebral injury, they suggested that profound hypotension was a likely cause, but they agreed that there was no evidence to indicate that such an event occurred. They agreed that if Michelle did become profoundly hypotensive, it must have occurred during an undocumented period. As we have said, the oral evidence accepted that it was most unlikely that trained medical staff would have missed such an episode and the case based upon profound hypotension or hypoxia was abandoned. Even at the stage of the experts' discussions, it is evident that Professor Hull and Professor Adams were struggling to find an explanation for Michelle's brain damage. They even went so far as to emphasise a suggestion that the neurological and neurosurgical experts might be asked whether there was any possible surgical accident which could result in the same natural history and result in the same radiological appearances.
26. We have already indicated the respective opinions of Dr Bates and Professor Chadwick, which are reflected in the record of their discussions before the trial. In answer to the question by what mechanism, if at all, could the fits cause brain damage, they agreed that seizures may cause brain damage by giving rise to a metabolic demand within the brain in excess of the ability of the cerebral circulation to supply oxygen and glucose. If seizures are uncontrolled and generalised, then there may also be hypoxia and hypotension that may further exacerbate brain damage.
27. Although the judge indicated, when he gave permission to appeal, that his conclusion depended upon inferences, the grounds of appeal do not rely on inferences alone. In our judgment, parts of the grounds of appeal are unsustainable in this court on evidential grounds. This applies in particular to Dr Read's evidence about taking Michelle's blood pressure on occasions other than those which were recorded, his evidence that these were normal and his evidence that her general appearance throughout did not indicate hypotension or hypoxia. The judge made no findings of fact which rejected this evidence. The only part of Dr Read's evidence which the judge rejected as honest, but mistaken reconstruction was his evidence that Michelle suffered a number of generalised seizures. For other matters, including the possibility that Dr Read may have been attending other patients as well as Michelle, the judge went no further than saying that, when so many years had passed after the event, it was very difficult to be definitive about any of these matters. It would, we think, have been very difficult for the judge to reject Dr Read's evidence about taking Michelle's blood pressure, since Mr Faulks did not challenge this evidence in cross examination. This is not a criticism of Mr Faulks, since it was intrinsically unlikely that an experienced anaesthetist, whom the judge regarded as an honest and conscientious witness, would not have taken blood pressure readings in the way in which he described. It is even more unlikely that he took blood pressure readings which were dangerously low and failed to take steps to deal with that. In these circumstances, it is in our view not open to this court to reach conclusions of primary fact whose effect would be to reject Dr Read's evidence. We have not seen and heard him give evidence. This means that most of paragraph 3 of the grounds of appeal (except the part which refers to blood gases) and parts of paragraph 4(2) are not sustainable. In our view, this court has to proceed on the basis that Dr Read did take blood pressure readings on other occasions than those which were recorded, including at critical parts of Michelle's treatment such as immediately following his administration of fentanyl, and that those readings were normal. This may not literally measure up to the agreed requirement of blood pressure readings every 10 minutes, but it does mean that Michelle's blood pressure was apparently normal on a greater number of occasions than those which were recorded and that these included those occasions which Dr Read regarded as critical. In addition, there was no other sign of relative hypotension. This, in our view, means that a case of negligence based on a failure to detect relative hypotension is not, as the judge found, made out. If, as the judge concluded, there was or may have been one or more episodes of relative hypotension between 16.42 and 21.45, there is no more than a speculative basis for supposing that additional blood pressure readings on occasions when there was no outward manifestation of hypotension would have revealed it. This is especially so, since on every occasion when the blood pressure was taken it was found to be normal. Still less is there a proper basis for concluding when the relative hypotension probably occurred, how long it lasted and how severe it was, and whether any steps which Dr Read might then have taken would have prevented Michelle's eventual condition.
28. The case based upon Dr Read's failure to take blood gas samples two hours earlier than he did is somewhat less speculative. So far as it goes, there was a clear breach of duty here. To establish Michelle's case, it would be necessary to infer that blood gas samples at 18.00 would probably have shown serious metabolic acidosis; that her brain damage had not by then already occurred; that steps could have been taken then to reverse the condition producing the metabolic acidosis; and that those steps would have been successful in preventing her brain damage. We have struggled with these necessary inferences. In the result, we have concluded that it cannot be said that the judge's decision as applied to this part of the case was wrong. In our view, an inference that blood gas samples taken at 18.00 would probably have shown serious metabolic acidosis is by itself arguable, although even that is problematic. But we do not think that the evidence, taken at its highest in Michelle's favour, could properly enable the court to draw all the necessary inferences. The trouble is that the experts were in truth unable to provide an explanation for the very significant metabolic acidosis. They suggested possible causes but none of those causes was really available on the evidence. In these circumstances, the judge was, we think, right to conclude that it was entirely a matter of speculation what earlier blood gas samples would have shown. There would then have been difficulties in concluding that Michelle's brain damage had probably not occurred before 18.00. It had after all been Professor Hull's preferred original explanation that Michelle's brain damage had resulted from hypotension following the administration of fentanyl after the first fitting episode at 16.42. Further, there are difficulties in concluding that steps taken after a possible discovery of severe metabolic acidosis at 18.00 would have successfully overcome its cause, whatever that may have been. The steps taken after 20.00, which are not in themselves criticised, had not apparently succeeded in dealing with the problem by 22.00 or 23.30.
29. In our view, the underlying difficulty in Michelle's case was that none of the experts was really able to say with confidence what the cause of her severe brain damage was. The judge found that Dr Bates' thesis was probably correct, but that thesis still begged the question why Michelle suffered some degree of systemic relative hypoxia and/or hypotension. As to negligence, our view has in the end coincided with that of the judge. We too have been driven to the final conclusion that there is insufficient evidence of the nature, degree and duration of the relative hypotension or hypoxia to justify a finding as a matter of probability that it was caused by negligence on the part of the defendants. We do not think that poor record keeping on E Neuro ward can supply positive material which is absent and which cannot properly be inferred. Ultimately, we are satisfied that it would not be right for this court to make the inference which Mr Faulks pressed on us, but which the judge was not prepared to make, that the causatively significant period of hypoxia and/or hypotension could and should have been observed and heeded by Dr Read in time for Michelle's brain damage to have been avoided.
30. At a personal human level, we regret this conclusion. But at that same level, we conceive that in reality the cause of her severe brain damage has not been explained. It is not necessary to reach a conclusion on the second or third contentions in the respondents' notice. If it had been necessary to do so judicially, we doubt if we would have been persuaded to a conclusion that the judge was wrong as to the timing of the brain damage or in accepting Dr Bates' conclusion. At a more general level, however, we can see the force of the submission that the timing of the damage was not established. Again, there was general acceptance that prolonged, generalised seizures might cause brain damage, and Michelle did have four fitting episodes after the operation before she was transferred to the intensive care unit and she continued to fit there when she was not sedated.
31. For these reasons, we would dismiss this appeal.

LORD JUSTICE RIX:

I agree.

LORD JUSTICE ALDOUS:

I also agree.

ORDER: Appeal dismissed with costs, to be the subject of a detailed assessment; section 11 order against the Legal Services Commission; liberty to apply. Claimant's costs to be subject to detailed assessment. (Order does not form part of approved Judgment)