Holdsworth, R v

This case arises from the death of Kyle Fisher on 23 July 2004 at the age of 2 years 8 months. Suzanne Holdsworth was convicted of his murder on 8 March 2005 at Teesside Crown Court before Grigson J. No criticism is made of the conduct of the trial, but she applies for leave to appeal and an extension of time for doing so on the basis of fresh medical evidence. Her application was referred to the full court by the registrar. The hearing before us took the same form as a full appeal.

Kyle lived with his mother, Claire, at 23 Troutpool Close, Hartlepool. She was aged 19 at the time of Kyle’s death.

The applicant lived across the road at 3 Millpool Close with her partner and their daughters, Lesley and Jamie Lee, who were then aged 14 and 10. The applicant’s partner was a long distance lorry driver.

Claire and Kyle saw quite a lot of the applicant and her daughters. Lesley was much closer in age to Claire than was the applicant, who was then aged 35, and Lesley and Claire became friends. Jamie Lee developed a bond with Kyle. At the time of Kyle’s death the applicant and Claire were on good terms, although there had been periods of hostility between them.

Claire’s mother lived nearby but at the time of Kyle’s death she was away on holiday with her partner.

On Saturday 17 July Claire went with a friend in the evening to a nightclub in Hartlepool. She had been unable to arrange a babysitter and she left Kyle on his own asleep in bed at her mother’s house. The applicant learned about this and told Claire that she would tell her mother when she returned from holiday. Claire was upset and begged her not to do so but the applicant was adamant.

On Tuesday 20 July Claire and Lesley went to the cinema leaving Kyle at the applicant’s house in her care. Claire and Lesley left around 5 pm and got back at about 8.30 pm. At the trial there were partially conflicting versions of what happened after that. Claire’s version was that when they arrived back at 3 Millpool Close the applicant was in the garden. Claire asked her how Kyle had been. The applicant told her that he was fine and was in bed. Claire went up stairs to look at him. She opened the bedroom door but did not put on the bedroom light. Kyle was asleep on the floor. She picked him up. He was drowsy but spoke to her. She could not see anything wrong with him and so she put him in bed and went downstairs. She told the applicant that Kyle must have fallen out of bed and then went round to her mother’s house to feed the dog. Within half an hour of leaving she received a phone call from the applicant saying that Kyle had bruises on his head and wanting to know why she had not mentioned them. She said that she had not seen any bruises and asked the applicant how Kyle was. The applicant replied that he was fine and she could hear Kyle laughing in the background. Claire did not go back to look at him and left him there for the night.

According to the applicant’s version, when Claire and Lesley arrived back from the cinema she (the applicant) went out for about half an hour. On her return Lesley was in the garden and told her that Kyle had fallen out of bed. She went in the house and asked Claire if Kyle was alright. Claire said that he was. Claire stayed for another hour or two. When later the applicant went up to bed she saw that Kyle was sitting up on his bed. She put the light on and saw that he had bruising on both sides of his head. She immediately rang Claire and was annoyed with her for not having seen and mentioned the bruising. Claire had said that she had not turned the lights on. Lesley gave Kyle a drink. He was fine at this stage, talking and giggling, so she put him back to bed.

Lesley’s account supported that of the applicant. She said that after they had returned from the cinema, and her mother had gone out, Claire went up to check on Kyle while Lesley was putting away some gardening tools. Claire was up stairs for about 15 minutes. Eventually Lesley went up to see her and Claire explained that Kyle had fallen out of bed. Lesley supported her mother’s account of later going up to bed and at that stage seeing bruises on Kyle’s head.

It was common ground that next morning Claire came round and looked at Kyle. It will be necessary to come back to the witnesses’ description of his condition after whatever happened to him on the Tuesday evening, but all agreed that his behaviour on the Wednesday was normal and that he showed no sign of discomfort, apart from complaining of a poor head when touched with a soft toy in the course of play.

The applicant spent the evening of Wednesday 21 July at bingo with a friend. It had been arranged that when she came back she would look after Kyle so that Claire could go out. Claire gave Kyle his tea at her mother’s house. Sometime about 9 to 9.30 pm a witness, Ellen Ashley, saw Claire and Kyle coming out of Claire’s mother’s house and there was conversation between them about Kyle having banged his head a few times.

The applicant arrived back from bingo not long after 10 pm. Claire brought Kyle round to her house in his pyjamas and left soon after 10.15 pm. From that time on the applicant and Kyle were alone in the house.

At 11.34 pm the applicant made a 999 call. The conversation began with the applicant saying “Can I have an ambulance straightaway please? I am babysitting for a 2 year old child and he has just gone all floppy…He is not breathing…His eyes are rolling and everything.”

The conversation continued for a little over 11 minutes until the ambulance arrived. Kyle was then breathing at a rate of 22 inhalations per minute, which is slightly faster than normal. His pulse was 83, which was normal for a child of his age. His oxygen saturation rate was satisfactory but he was unresponsive and his Glasgow Coma Score was assessed at 3. His eyes were dilated and unresponsive to light. Kyle was taken to the accident and emergency department at Hartlepool Hospital. His condition was plainly very poor and he was transferred to Newcastle General Hospital. There he was seen by Mr Todd, a consultant neurosurgeon, and Dr Devlin, a consultant child neurologist. A CT scan and later a MRI scan were carried out. He was placed in the Paediatric Intensive Care Unit (PICU). The consultants concluded that he had suffered severe brain stem injury from which he had no chance of survival. On Friday 23 July life support was withdrawn. Subsequently a post mortem examination was conducted by Dr Sunter, a Home Office pathologist.

The applicant has been consistent in her account of events during the period of more than an hour on the Wednesday evening between Claire leaving Kyle with her and her calling 999. Her account is that during this period Kyle was downstairs with her and was behaving normally. He had some food and a drink and they watched television. He identified characters on the programme. He then became sleepy and started to nod off. She asked if he wanted to go to bed and he said “Susie knackered”, which was a phrase that he had picked up from her. They got up and at that point he had a sudden fit. She carried him to the kitchen and splashed water on his face in the vain hope that this would bring him round. He continued to fit and jerk. She was in a panic and had no idea what was wrong with him. She then laid him on the sofa and dialled 999.

Externally Kyle had a large number of minor marks of a kind which are common in an active toddler. As to these Dr Sunter considered that it would be “unreasonable to read too much into what are after all just multiple trivial bruises”. The bruises which he identified as significant were to the top of the head, both sides of the jaw and both shoulders.

He described the bruising to the top of the head as “a complex pattern…consisting of a large area of nondescript deep bruising (measuring 145 mm by 185 mm on the deep aspect of the scalp) and within which there were evident at the skin surface curved linear marks to the right of the mid line and straight linear marks to the left of the mid line.”

He considered this to be a very peculiar complex of injury but, having looked at the banisters at 3 Millpool Close, the injury could be accounted for by Kyle’s head being impacted several times against them.

He considered that the bruises to the right and left jaw margins and to the tops of the right and left shoulders formed a symmetrical pattern of further peculiar injury and would be consistent with gripping. Putting these features together, he considered that they were indicative of the child being gripped by the neck and shoulders when in a crawling position and having the top of his head struck several times against the banisters.

Internally, there was no evidence of any primary injury to any particular part of the brain, such as a bruise or laceration of the brain substance or a fracture, but there was diffuse cerebral odoema and a subdural haemorrhage on the right side. The latter would not have been life threatening, because the amount of blood accumulated (i.e. 38 grams) was not large, but it was an indicator that there had been a blunt force head injury.

The prosecution called as witnesses Dr Sunter, Mr Todd and Dr Devlin. The combined effect of their evidence was that Kyle died as a result of a blunt force head injury causing acute cerebral odoema, increased intracranial pressure and irreversible damage to the nerve cells of the brain through disruption of the supply of blood and oxygen. To cause the degree of damage the trauma must have been severe and the symptoms would have become apparent immediately or very soon after the impact, i.e. within minutes. The child would have been immediately unwell and it would have been apparent that the injury was catastrophic within an hour at the outside.

If this evidence was accepted by the jury (as it evidently was), it necessarily followed that the trauma must have been inflicted by the applicant, since he had been in her sole custody for over an hour before the 999 call, and her account of Kyle behaving normally until he suddenly had a fit for no apparent reason could not be truthful. The applicant herself gave evidence and was naturally cross-examined on that basis.

No medical evidence was called on her behalf. Mr Blaxland QC told the court that reports had been obtained from a number of experts whose specialities included forensic pathology and paediatric neurology, but none of their reports provided the defence with any basis for challenging the prosecution’s medical evidence.

Prior to the hearing the applicant issued notices of application to call 5 expert witnesses whose reports were disclosed. In the event Mr Blaxland decided to rely on only 3 of them and we heard evidence de bene esse from those witnesses. They were Professor Guerrini, who is a clinical professor of paediatric neurology at the University of Florence (and formerly professor of paediatric neurology and consultant paediatric neurologist at Great Ormond Street hospital); Dr Anslow, who is a consultant neuro-radiologist at the Radcliffe Infirmary, Oxford; and Dr Squier, who is a consultant neuro-pathologist at the Radcliffe Infirmary. The prosecution called evidence from Dr Rittey, who is a consultant paediatric neurologist at the Ryegate Children’s Centre (part of Sheffield Children’s NHS Foundation Trust), and Professor Milroy, who is a professor of forensic pathology at the University of Sheffield and a consultant pathologist to the Home Office. (Professor Milroy was instructed by the prosecution because Dr Sunter has died). Nobody has questioned the integrity of any of these witnesses or their expertise in their respective fields.

The theory of the causation of Kyle’s death advanced on behalf of the applicant contains a number of strands, but Mr Robertson QC on behalf of the prosecution argued forcibly, and with justification, that the prosecution has been faced with a moving target when addressing the question how these different strands cohere into a unified theory. As the appeal hearing went on there was a process of intellectual debate between the experts, and the final presentation of the appellant’s case only became fully apparent when Professor Guerrini was re-called (at the court’s suggestion) after Dr Rittey had given evidence. The defence made the point that this was a case with a number of complex factors and that it may not be possible to be precise as to the interaction of the various factors. Be that as it may, with the benefit of hindsight it would have been better if experts of like discipline had been in direct contact with one another before the hearing in order to consider together the points advanced by one another and to produce a summary of points of agreement and disagreement with a summary of their reasons for disagreement.

The explanation for Kyle’s death suggested as a possibility by the applicant’s expert witnesses can be summarised as follows:

1.

Kyle suffered from a number of pre-existing brain abnormalities which predisposed him to an epileptic seizure.

2.

Kyle may also have suffered from a subdural haemorrhage as a result of whatever had happened prior to the Wednesday evening, although this is not an essential part of the theory. If so, that would have been an additional predisposing factor for an epileptic seizure.

3.

The description given by the applicant during the 999 call was a description of an epileptic seizure.

4.

The seizure may have led to the hypoxic ischaemic brain damage (i.e. cerebral odoema, rise in intracranial pressure and starvation of blood and oxygen to the brain cells).

The applicant’s case is that if that is a credible medical explanation, her conviction must be unsafe, because the medical evidence was the foundation of the case against her. Mr Robertson submitted that the fresh medical evidence is irrelevant because there was an unanswerable case that Kyle suffered some trauma while in the custody of the applicant on the Wednesday evening, independent of the opinion of the prosecution’s medical experts that the injuries from which he was suffering on his admission to hospital must have been caused within an hour prior to his collapse. Mr Robertson made it clear that this was his primary submission. He also submitted that the medical theory advanced on the applicant’s behalf is not credible for reasons explained by Dr Rittey. Mr Robertson wove his arguments together in submitting that even if the medical theory were intellectually possible, in deciding whether it affords a credible explanation for Kyle’s death it is necessary to look at the case as a whole.

Professor Guerrini’s report also raised the possibility that the injuries observed on Tuesday evening and Wednesday morning may have been caused by Kyle suffering an epileptic seizure in which he was thrown out of bed on the Tuesday evening. Mr Blaxland did not rely on this suggestion for the purpose of the present appeal, because for that purpose it did not matter how Kyle’s Tuesday evening injuries were sustained. Although the prosecution had put it to the applicant in cross-examination that she had injured Kyle on the Tuesday, the murder charge was based on the allegation that the applicant inflicted fresh trauma on the Wednesday evening and Mr Robertson told the jury in his opening speech (according to his written note) that “whatever happened on the Tuesday night is totally irrelevant when considering how Kyle’s fatal injury was caused and by whom”. Mr Robertson properly maintained that position on the appeal.

Mr Robertson’s submission that there was an unanswerable case that Kyle suffered trauma while in the custody of the applicant on the Wednesday evening, independent of the opinion of the prosecution’s medical experts that the injuries from which he was suffering on his admission to hospital must have been caused within an hour prior to his collapse, depends on circumstantial evidence. The first point to note is that if this evidence had been put to the jury as a foundation for inferring that the applicant injured Kyle on the Wednesday evening, the judge would have been required to give the jury a full and careful warning about the proper approach to circumstantial evidence. However it is clear from the opening note of the prosecution for the trial that the case was not put this way. That is reinforced by the fact that the very experienced judge did not give such a direction. He clearly (and rightly) did not see this circumstantial evidence as a material, let alone important, part of the prosecution’s case for finding that the applicant inflicted the fatal injury.

In the early part of his summing-up the judge referred to evidence which the jury had heard about the character of the applicant and Claire before summarising the issues which the jury had to decide. There had been evidence that the applicant was a mature woman and experienced with children but also evidence of a loss of temper on occasions. The judge turned from that general evidence to the matters which the prosecution had to prove. He said:

“The first crucial issue in this case is whether the prosecution have proved that it was Mrs Holdsworth who inflicted the fatal injury, and this evidence which lawyers called propensity evidence, evidence as to bad temper on the one side, evidence as to her patience and skill on the other, goes to that issue. She is on the face of it an unlikely candidate, but at the end of the day, and this is a matter for you, you may think the important evidence comes from the doctors.”

That evidence from the doctors was all one way.

At the trial both sides proceeded on the basis that there must have been some additional trauma after the Tuesday night, for which the applicant and Claire each denied responsibility. In summarising the evidence the judge reminded the jury of the two accounts of events over the preceding days which formed the background to the events on the Wednesday evening. The jury may have formed an adverse view of the applicant’s credibility in relation to those events or they may have disbelieved the applicant simply on the basis of the medical evidence. The judge gave the jury a Lucas direction but nowhere in the summing-up did he suggest that the jury might rely on inferences from evidence relating to the period prior to the handover of Kyle on the Wednesday evening as circumstantial evidence from which they might infer that the applicant inflicted the fatal injury.

The principal matter on which the prosecution relies in support of its argument that there was an unanswerable case against the applicant quite apart from the medical evidence is the alleged appearance of fresh marks after Kyle’s admission to hospital as compared with his appearance earlier on Wednesday at the applicant’s home. It was accepted by Mr Robertson that if there were such fresh marks, there is no direct evidence when and where they were inflicted. But he submitted to us that the inescapable inference must be that those marks were caused by injuries caused by the applicant between 10.15 and 11.34 pm on the Wednesday evening.

This topic brings us back to the descriptions of Kyle’s condition after the events of Tuesday evening.

Mr Robertson relied primarily on answers given by the applicant when first interviewed by the police and in cross-examination. When first interviewed by the police, the applicant said that after the incident on the Tuesday evening she saw that Kyle had purple bruising in a straight line on the left side of his head. She made no mention of any mark on the right side. On 25 July she was interviewed after she had been arrested for Kyle’s murder. During the interview she was shown photographs of the bruising on Kyle’s head and was asked if those were the injuries which she had noticed on the Tuesday. She said that they were and that looking at the photographs made her remember that there was bruising on both sides. She had only seen the bruising on the right side when Lesley had lifted Kyle’s hair up and drawn her attention to it.

Mr Robertson also relied on the following answers in cross-examination:

Q. Now, are you saying, Mrs Holdsworth, that the injuries after the Wednesday night were exactly the same as on the Tuesday night?

A. Not on the photograph when she showed me it, no.

Q. No. What you saw on the photograph was different from what you noticed on Kyle’s head from the Tuesday night? That’s right, isn’t it?

A. On the Tuesday night when I looked at Kyle’s head they looked straight.

Q. Yes. They were different from the bruises that you saw on the photographs…

A. Yes.

Q. …taken after the Wednesday night?

A.

The photographs were larger than…

That evidence cannot fairly be taken as an admission or basis for a finding that Kyle had more marks after he was handed over to the applicant on the Wednesday evening than at the time when he was handed over. The applicant’s evidence on that was to the contrary. The change in her answers to questions from the police before and after being shown the photographs could be regarded as reducing her credibility, but it does not go further than that. The fact that bruising on the photographs appeared larger than she remembered might have various explanations.

Claire Fisher said in her witness statement that on the Wednesday morning she was worried about Kyle’s bruises and so went to see him. She was shocked by what she saw. There was dark purple bruising on both sides of his head. On the right side there was a straight line as if he had hit his head against something straight. On the other side there were a number of bruises, dark purple in colour, not in a line, lower down than those on the right side of his head and nearer the front. When she saw Kyle at Hartlepool Hospital she could not see any other marks or bruises on him than she had seen on the Wednesday morning. Later in the PICU at Newcastle General Infirmary she noticed that he had a little bruise under his chin which she had not previously seen. She confirmed that statement in her evidence at the trial.

If the scallop shaped bruising on the right side of the head had been new bruising, caused by the applicant repeatedly striking Kyle’s head against the banisters, it is a pertinent question why Claire failed to recognise this as fresh bruising at a time when she identified the small mark on the chin, which the prosecution say was a grip mark. Moreover it was the prosecution’s suggestion that the scallop shaped bruising on the right and the straight line bruising on the left were caused simultaneously by Kyle’s head being thrust between the banisters (and it was their combination which Dr Sunter considered to be a particularly unusual pattern). That is hard to reconcile with Claire’s evidence of seeing the straight line bruise on the Wednesday morning. Mr Robertson properly acknowledged that this was something which weakened the force of Dr Sunter’s opinion that those marks were caused together.

Ellen Ashley’s evidence was that when she met Claire outside Claire’s mother’s house on the Wednesday evening Claire told her that Kyle had banged his head a few times. She did not say where or when it happened but said that he had done it that day. As Claire spoke she indicated with her hands to each side of Kyle’s temples.

The prosecution also relied on evidence of lies by the applicant. When first interviewed by the police, the applicant, Claire and Lesley all lied about where Claire had spent Tuesday night. They said that she had spent the night at the applicant’s home. The idea of telling this false story first came from the applicant. She claimed that her motive was to protect Claire. The medical staff at the hospital had said that social services were going to be involved. The applicant said that she thought that it might look bad for Claire that on the Tuesday night she had not come to look at Kyle and take care of him after he had hurt himself. The prosecution submitted that the object of the lie was to enable the applicant to accuse Claire of causing injuries to Kyle on the Tuesday evening, which had in truth had been caused by the applicant. The prosecution also suggested that the applicant and Lesley were continuing to lie in saying that on the Tuesday the applicant had gone out after Claire and Lesley had returned from the cinema, and in Lesley saying that Claire had been on her own upstairs with Kyle for 15 minutes, thereby creating a window of opportunity for Claire to have injured Kyle. If that were right, it would provide material to support a case that the applicant had injured Kyle on the Tuesday evening; it was not evidence of the infliction of fresh trauma on the Wednesday.

Put shortly, we have concluded that none of the circumstantial factual evidence provided a basis independent of the medical evidence for a finding that fresh trauma was inflicted after the applicant took charge of Kyle on the Wednesday evening. Even if we had concluded otherwise, the weight of such evidence would have had to be evaluated along with other factors (including the likelihood of the applicant behaving in such a way and the likelihood of Kyle showing no sign of fear or anxiety in her presence on the Wednesday if she had inflicted bruises on him on the previous evening). It would not have been possible for this court satisfactorily to carry out that kind of primary evaluation simply from a reading of the transcript and the court documents. It would have been an exercise for the jury on a proper direction.

In this case the effect of the medical evidence was to turn what was otherwise no more than a case of opportunity to commit a fatal assault into a certainty, if the jury accepted (as they plainly did) the opinion of the prosecution’s medical experts about the conclusions which could be drawn from Kyle’s injuries. In these circumstances the safety of the applicant’s conviction depends on whether there was a credible alternative medical explanation consistent with the applicant’s account of events.

At the outset it is worth noting certain negatives which were apparent from Dr Sunter’s report and were highlighted by Dr Anslow. There was no evidence of any brain bruise. There was brain swelling, but that is a different matter. There was also no evidence of any scalp haematoma or scalp swelling. There was evidence that there had been trauma, because there were the external bruise marks identified by Dr Sunter and there was a subdural haematoma. But medically neither the bruise marks nor the subdural haematoma could in themselves be dated to any particular time. Dr Sunter accepted at the trial that he could not exclude the possibility that the subdural haematoma had been caused on the Tuesday. He said in cross-examination:

“With a subdural haemorrhage you can’t say the injury took [place] so many hours ago, so many days ago. It’s a guess”.

In the case of a subdural haematoma the accumulation of blood may be slow and in the meantime the individual may behave quite normally.

So it was not possible to identify primary medical evidence of a necessarily fresh injury. The cerebral oedema was secondary to something, and the question was what was that something. The prosecution’s case was that the only sensible inference was that it was secondary to significant trauma, and, if so, that the trauma must have been very recently inflicted, enabling the jury to find that the applicant’s account of events must be false.

It is in that context that the alternative medical explanation falls to be considered.

We return to the four propositions set out in paragraph 26 by way of summary of the alternative medical theory.

As to the first, it is accepted that Kyle suffered from a number of pre-existing brain abnormalities which predisposed him to an epileptic seizure. He had an abnormally large brain, nodules of ectopic neurones capable of generating epileptic activity and a scar involving the cerebral cortex as a result of previous trauma. These factors would have caused him to have an over 80% risk of exhibiting epileptic seizures.

As to the second, it is accepted as medically possible that the subdural haemorrhage resulted from whatever happened on the Tuesday and, if so, that this would have been an additional pre-disposing factor to epileptic seizure.

As to the third, the seizure described by the applicant during the 999 call may have been either clonic or tonic. Dr Rittey considered that it was likely to have been a tonic seizure resulting from brain stem compression, but he accepted the possibility that it was an epileptic seizure, which was Professor Guerrini’s opinion.

That leaves the critical issue whether it is a credible medical theory that the seizure may have led to hypoxic ischemic brain damage.

Professor Guerrini stressed that we do not know as a fact what were the dynamics of the relevant factors. Epileptic activity may affect a lone area of the brain or a combination of networks or the whole brain. During an epileptic seizure the vastly increased neural activity means that the patient will require a greatly increased amount of oxygen and blood in the brain. If Kyle already had a subdural haemorrhage, the drainage of blood and CSF may have been imperfect. 11 minutes or more was a long time for a clonic seizure during which he may have suffered a period of apnoea. In Professor Guerrini’s opinion it was possible that this caused cerebral oedema resulting in severe hypoxic ischemic brain damage.

Dr Rittey considered this explanation to be untenable. When proper account is taken of the recorded facts about Kyle’s condition at the time of the arrival of the ambulance and on Kyle’s admission to hospital, Dr Rittey did not consider that the suggested train of events and explanation had a sound theoretical basis. Further, it was unsupported by any empirical evidence to be found in the medical literature.

As to the theoretical argument, Dr Rittey argued that if the severe brain damage occurred during the seizure while Kyle was not breathing, it was inconceivable that he could have then started breathing again. Yet his rate of breathing and pulse were normal when the ambulance arrived. The fact that Kyle was unresponsive and his eyes were fixed confirmed that he had suffered very severe brain damage, which must have been a result of the trauma which pre-dated any seizure. Dr Rittey also argued that on the alternative theory Kyle must have been severely acidotic, but that blood taken on his arrival at hospital showed that he was not acidotic. As to the empirical evidence, Dr Rittey emphasised that there was no case reported in the medical literature of a seizure as short as 11 minutes causing damage of the severity suffered by Kyle. Insofar as Professor Guerrini had referred to papers in support of the suggestion that Kyle could have died as a result of the seizure, the cases reviewed in those papers were quite different from the present case, and Dr Rittey was critical of the way in which Professor Guerrini had sought to rely on that literature.

On the matter of breathing and pulse, Professor Guerrini responded that the centre which regulates respiration is at two levels in the brain, mid and lower. You could therefore have a situation in which the seizure activity in the brain stopped the individual from voluntary control of breathing, but at the end of that activity normal, but involuntary breathing and heartbeat could resume, because the oedema had not yet had fully affected the brain stem. As to the lack of response of the pupils to light, Professor Guerrini agreed when first cross-examined that if Kyle was in a post-convulsive coma when the ambulance arrived, his eyes should have been responsive to light. When re-called to give further evidence, he said that the response to light could have been affected if the oedema at that stage had been severe enough to affect the mid part but not the lower part of the brain stem. This was a new development and Mr Robertson was critical of the way in which it was advanced. As to acidosis, Professor Guerrini said that there is a chemical barrier such that the acidosis in the brain would not have been matched by corresponding acidosis elsewhere in the body. He made the point that on the prosecution’s own theory there had been acute hypoxic ischemic injury to the brain and that on either theory the acidosis would have been the same. The issue which divided the experts was what may have caused the hypoxic ischemic injury.

On the matter of empirical evidence, Professor Guerrini observed that in a clinical situation a patient would never be allowed to continue fitting for more than 5 minutes without medical intervention and therefore you would not have medical studies of people undergoing seizures of the length in the present case. Where patients died as a result of such a seizure, it would be away from hospital. He emphasised that although death during a seizure is rare, it does happen and there will not necessarily be a post-mortem. Speaking in general terms, Professor Guerrini said that his theory is one of the main theories advanced to explain sudden death in epilepsy. In this case the possible co-existence of a subdural haemorrhage would have been an unusual complicating feature. Dr Squire made the point tersely when she said that she had never seen a paper which considered the effect of seizure in the context of subdural haematoma.

Dr Rittey acknowledged that Professor Guerrini is an internationally acclaimed expert in epilepsy and that any theory advanced by him on this subject is entitled to very careful consideration.

Conclusions of medical experts on the cause of an injury or death necessarily involve a process of deduction, that is inferring conclusions from given facts based on other knowledge and experience. But particular caution is needed where the scientific knowledge of the process or processes involved is or may be incomplete. As knowledge increases, today’s orthodoxy may become tomorrow’s outdated learning. Special caution is also needed where expert opinion evidence is not just relied upon as additional material to support a prosecution but is fundamental to it.

With these considerations in mind, we have concluded that the interests of justice make it necessary that we should admit the fresh medical evidence adduced by the applicant under s 23 of the Criminal Appeal Act 1968. Having listened to the oral evidence called on both sides and evaluated it in the context of the case as a whole, we do not consider that we can safely dismiss the medical scenario advanced on the applicant’s behalf as definitely incredible. In R v Pendleton [2002] 1WLR 72 Lord Bingham commended it as wise for the Court of Appeal in a case of any difficulty to test their own provisional view by asking whether the evidence, if given at the trial, might reasonably have affected the decision of the trial jury to convict (paragraph 19). In our judgment it might, and therefore the conviction must be judged unsafe.

For those reasons we grant permission to appeal and quash the applicant’s conviction for murder. We add as a postscript that any re-trial will require a high level of case management. We draw attention, in particular, to the observations of this court in R v Harris [2006] 1 Cr App R 5, at paragraph 273, regarding the powers of the court to make provision for experts to consult together and, if possible, agree points of agreement or disagreement with a summary of reasons.