[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales High Court (Family Division) Decisions
You are here: BAILII >> Databases >> England and Wales High Court (Family Division) Decisions >> Re A (Non-Accidental Injury: Medical Evidence) [2001] EWHC Fam 5 (12 January 2001) URL: http://www.bailii.org/ew/cases/EWHC/Fam/2001/5.html Cite as: [2001] Fam Law 735, [2001] 3 FCR 262, [2001] 2 FLR 657, [2001] EWHC Fam 5

[New search] [Help]

IN THE HIGH COURT OF JUSTICE
FAMILY DIVISION

		Royal Courts of Justice Strand, London, WC2A 2LL
		12 January 2001

BRACEWELL J:

1. In this judgment I propose to grant leave to publish that part which relates to medical issues, subject to the preservation of anonymity of the parties. I shall indicate which part for publication is permitted as I go through the judgment.
…
62. I now deal with the medical issues with leave for publication and I have anonymised the parties and the particular hospitals concerned.
63. A, a boy, was born on 4 June 1999. He was a healthy baby who thrived. The history of what occurred on 29 December 1999 has been given by the parents in that A was given a feed consisting of a bottle of milk and a rusk by his father shortly after 3.00 pm. The baby took his feed satisfactorily and was carried upstairs by his father to his cot in order to sleep. The baby was tired, rubbing his eyes and crying. He went to sleep in his cot. At about 4.15 pm his mother went upstairs and found the child lying prone and not moving in his cot. He was described in the earliest account as being white, limp and not breathing. His mother picked him up, shouted for the father, who came upstairs, took the child from her, slapped the baby's leg and when there was no reaction commenced resuscitation. The child posited a small quantity of milk. Mother dialled 999. The paramedics arrived at 4.28 pm and took the baby to the local hospital. During the evening the baby was transferred to a children's hospital where he died on 30 December 1999, the life support system having been withdrawn at 8.30 pm.
64. The question to be addressed is what caused the baby to collapse and then die. Issues arise whether his death was natural, unnatural or unascertained. There is disagreement between the many distinguished medical experts who have considered the cause of death. Differential diagnosis is of crucial importance in addressing the issues. However, it is to be noted that medical evidence is only one aspect and the credibility of the parents also has to be carefully considered.
65. The mother's case is that she does not know the cause of the baby's death and was in no way responsible for it and that she did not see, hear or observe anything done by anybody on the day of death nor on preceding days which may have led to his death. Her position is that she has to be guided by the medical experts and if the baby's death is judged to be unnatural then because she did nothing she would have to conclude that the perpetrator was the father.
66. The father's case is that the death of baby was a cot death and the precise cause is otherwise impossible to identify.
67. It is to be noted that none of the medical experts brought into the care proceedings nor the doctors instructed by the coroner are of the opinion that the death was a cot death, that is a SIDS. The bottom line for those experts who do not support a finding of non-accidental injury is that they term the death either unascertained or are unable to express an opinion.
68. On the third day of this hearing and almost 11 months since the baby's death the father made a further statement in which for the first time he indicated that A had fallen from a baby bouncer on a prior occasion, the date of which he was unable to specify. That explanation is discounted by all the experts as a possible cause of A's condition.
69. The local authority's case is that the retinal haemorrhages found in the eyes of baby were the result of severe shaking in an assault which no reasonable person could regard as being within a normal or acceptable range of handling a baby and that there is no plausible alternative explanation. The local authority further contends that there is a very high degree of probability that the death of A was the direct result of a severe shaking to the point of unconsciousness and a state of apnoea which interfered with the ability of the child to breathe leading to a lack of oxygen supply to the brain, which caused intra-cranial pressure and in turn caused death.
70. It is undoubtedly true that the frontiers of medical science are constantly being pushed back and that the state of knowledge is increasing all the time. That is why I find that when presented with a speculative theory based on an unlikely hypothetical base an expert will rarely discount it and will in effect never say never. Fanciful speculation is not an appropriate method of inquiry. What is needed and what the experts have done in this case is to piece together all the available information and look at the differential diagnosis. Many of the experts in this case specialise within a particular and very narrow field and by reason of being experts of referral at centres of excellence they acquire special knowledge and skill. However, concentration on a very narrow area of expertise can sometimes render it difficult for the expert to see the whole picture. In this regard I find that the pathologists are at a disadvantage when compared with the clinicians.
71. The judge has the duty of sifting the evidence from the experts, who form their assessments within their particular area of expertise, and the judge has to decide the case by reference to the various issues. Although the medical evidence is of very great importance it is not the only evidence in the case. Explanations given by carers and the credibility of those involved with the child concerned are of great significance. All the evidence, both medical and non-medical has to be considered in assessing whether the pieces of the jigsaw form a clear, convincing picture of what happened. The burden of proof in respect of disputed issues is upon the local authority and the standard is a civil burden on the balance of probabilities as propounded by Lord Nicholls of Birkenhead in Re H and Others (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, 586D-F, [1996] 1 FLR 80, 96B-G. The more serious the allegation the more convincing evidence is needed to tip the balance in respect of it. If the evidence at the end of the day is equivocal then the court cannot make a finding because the local authority would not have discharged either the burden or the standard of proof. The test, therefore, is the balance of probabilities but commensurate with the extremely serious allegations in this case. The court has to consider evidence and avoid conjecture and speculation.

Medical history

72. The two paramedics attended the home of the parents on 29 December 1999. They were informed at 4.22 pm that the baby had difficulty breathing and then at 4.26 pm that the baby had stopped breathing. They arrived at the home at 4.28 pm and found the baby on the floor at the top of the stairs. There was fluid around the baby's head. One paramedic assumed it was vomit and the other noted the vomit, which appeared to be a mixture of lumpy bread and milk, both on the floor and around the baby's mouth. The father had carried out mouth-to-mouth resuscitation. Cardiopulmonary resuscitation, known as CPR, was commenced after the baby's airway had been cleared with a hand-held suction unit. On the way to hospital in the ambulance baby was incubated and the CPR continued. One paramedic thought the baby looked waxy white and not cyanosed, as he would have expected, but he deferred to his colleague, who was holding baby and who considered baby to be cyanosed, thereby indicating a lack of oxygen.
73. On arrival at hospital Dr Rifkin, a consultant paediatrician, noted that baby was cyanosed and showed no neurological responses. Dr Rifkin and Dr Roberts, who is a locum consultant paediatrician, were in charge of baby's treatment. Dr Roberts asked the paramedics if there was anything suspicious and was told there was nothing. A was placed in intensive care and he still showed no neurological responses. Both Dr Roberts and Dr Rifkin examined A's eyes. Dr Roberts said in evidence he noted retinal haemorrhages in both eyes, he having used an ophthalmoscope, and he observed that both eyes had swelling and bleeding of the optic nerve but it was particularly in the right eye, although he stated that he did not have a good view. Dr Rifkin noted what he concluded were papilloedema in the left eye and retinal haemorrhage in the right eye. Neither consultant claimed any special expertise in eyes and they both deferred to Mr Lloyd. Fitting may have been observed by Dr Roberts, who went off duty at 6.30 pm. However, it is not clear if he observed fitting himself before he went off duty or if he was reporting what others had seen. Dr Rifkin in evidence stated that fitting was of concern by 8.00 pm and diazepam and phenobarbitone were administered. When Dr Yates arrived from the children's hospital at 10.15 pm fitting was continuing.
74. Dr Rifkin contacted the regional paediatric intensive care unit and the retrieval team headed by Dr Yates, who is a consultant in paediatric intensive care, arrived at the hospital and took A to a specialist children's hospital. He was still fitting intermittently. There was a problem with ventilation and a paralysing agent was administered. Dr Yates examined baby's eyes on the ward at the first hospital with an ophthalmoscope and he noted gross bilateral retinal haemorrhages and he drew a diagram. He discussed the CT scan with Dr Rifkin and a general radiologist, Dr Foreshaw. They consulted a text book and they considered there appeared to be a good match for a cot death. There was no discussion about the retinal haemorrhages. Dr Yates stated that he did not consider himself an expert on retinal haemorrhages and on that topic he deferred to Mr Lloyd's expertise.
75. On arrival at the children's hospital Dr McBride, who is a specialist registrar in paediatric intensive care noted that A had fixed and dilated pupils so that it was easy to see the back of the eye. His notes cover the period 2.30 am to 8.30 am. On receiving information from Dr Yates and looking at other notes he recorded 'Query near miss cot death'. He stated that he observed bilateral papilloedema and retinal haemorrhages around the disc of the right eye upon examination with an ophthalmoscope.
76. Dr Michael Clarke, who is a consultant paediatric neurologist from the children's hospital, examined A when alive at 10.25 am on 30 December 1999 and he took the history from the parents which I have set out. When he gave evidence he disagreed strongly with Dr Brian Clarke's subsequent conclusions about a connection between raised intra-cranial pressure and retinal haemorrhages. He stated in terms that he would not have expected to see diffuse axonal injury in a case of non-accidental injury and none was present. He was sure in evidence that it was an unnatural death and that A had been shaken or suffocated. There was, for him, no other scientific explanation to account for A's condition. He had no doubt the child had died unnaturally and that shaking was a component of the death. A was discovered to have a clotting abnormality but none of the experts consider that to be of any relevance. Dr Michael Clarke thought A's optic nerves were swollen with signs of papilloedema but he too deferred to his ophthalmic colleagues.
77. Papilloedema had been the subject of inquiry in this case. Four doctors found what they identified as bilateral papilloedema, namely swelling of the optic discs, during the course of admission to the two hospitals. However, Dr Smyth, now a consultant ophthalmologist, did not find any. The significance is that papilloedema are associated with raised intra-cranial pressure. Subsequently, at post mortem, Dr Bonshek did find bilateral papilloedema. On the evidence before me I do not find this apparent discrepancy surprising. I was very favourably impressed with the care and particular expertise with which Dr Smyth charted her findings. She uses a different technique from the paediatricians, who use an ophthalmoscope, which gives a magnified but very limited view. Dr Smyth found bilateral asymmetrical haemorrhages which were multi-layered around both posterior poles. They were worse in the right eye than the left. She did not think the discs were swollen but there was a haemorrhage on the edge of the disc on the right eye which blurred the margin. She supervised and approved a medical artist's representation of what she saw with one minor amendment, which I find is not of significance. She was a careful, cautious witness who has a particular speciality in examining eyes and who I accept as accurate in observation, and description. I find A did not have papilloedema in life and the other consultants who thought they saw them were wrong. She has examined hundreds of eyes and is thought of highly by her ophthalmic colleagues. She clearly explained why her stereoscopic observation technique is superior to an ophthalmoscope. From her experience she rated the haemorrhages as severe.
78. All the experts except Dr Brian Clarke deferred to her expertise and her findings coincided with those of Dr Bonshek. Papilloedema are hard to identify and Dr Smyth was not at all surprised that others thought they had identified them. As she put it, it is a common occurrence for a consultant ophthalmologist to be asked to consider others' findings and there is frequent disagreement. Professor David agreed and he too knew of her expertise. I find that those who diagnosed papilloedema prior to the examination by Dr Smyth were in error. That is in no way a criticism because they do not purport to have her expertise in that field. The subsequent finding at post mortem by Dr Bonshek is entirely consistent with the development of the swelling during the time between Dr Smyth's examination and the death of A because papilloedema take time to develop.
79. On the totality of the evidence on this topic, although I accept the expertise of Dr Smyth I agree with Professor David that whether or not there were papilloedema observed this is not a critical finding and has little effect on outcome because the crucial issue is the distribution of the retinal haemorrhages. Interpretation of these retinal haemorrhages is the cornerstone of this case. Dr Smyth consulted with Mr Lloyd, who is a consultant paediatric ophthalmologist. A's treatment was withdrawn at 8.30 pm and he died 10 minutes later.
80. The question which arises for determination is what is the significance, if any, of the retinal haemorrhages in considering the cause of death of baby? It is common ground that A did not have (i) optic nerve meningeal bleeding, (ii) far peripheral retinal haemorrhages, (iii) perimacula retinal folding, (iv) external signs of physical abuse or neglect, (v) diffuse axonal injury, (vi) damage where the brain joins the spinal cord, (vii) fractures of any limbs, (viii) stretch damage in spinal area relating to hypertension injuries, (ix) perl staining as evidence of previous asphyxia, (x) history of apnoeic episodes, (xi) acute pulmonary oedema relevant to suffocation. He did not have cerebral haemorrhages. It is the absence of cerebral haemorrhages which has in particular exercised the minds of numerous experts involved and the question addressed is whether the presence of retinal haemorrhages on its own is sufficient for an expert to say on the balance of probabilities, in accordance with Re H and Others (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, [1996] 1 FLR 80 that A suffered a non-accidental death. Four experts - Dr Rushton, Dr Geddes, Dr Brian Clarke and Dr Phillips - gave the opinion that the cause of death was unascertained. The remainder are satisfied it was non-accidental injury.
81. I turn now to the post-mortem evidence. Dr Bonshek, who is an ophthalmic pathologist, conducted the post mortem on the eyes of A. He found papilloedema bilaterally but particularly in the right eye and retinal haemorrhages in both eyes. He found evidence of retinal, pre-retinal and sub-retinal haemorrhages. His conclusions, having considered differential diagnosis and eliminating the conditions considered and identified by Professor David, were that on the balance of probability the cause of death was non-accidental injury by shaking. In reaching that opinion he took into account the unusual features, the absence of a full house of indicators, the raised intra-cranial pressure, the fitting, the CPR, but nevertheless concluded that shaking was the likely cause of death. He said in evidence he was aware of four other cases with retinal haemorrhages but no subdural or cranial bleeding where non-accidental injury was suspected but not proven. He did not know of a case in which shaking had been established with retinal haemorrhages alone. This case was unusual but not unique. He found that the evidence of raised intra-cranial pressure and the presence of retinal haemorrhages drove him to his conclusion of non-accidental injury with force used outside the normal range of handling an infant. He agreed with the description of what was observed by Dr Brian Clarke but he concluded that innocent explanations were eliminated and only non-accidental injury was left.
82. At the experts' meeting in August 2000 there was plainly some difficulty in accurately recording the views of the experts. Professor David was of the opinion that Dr Bonshek had changed his views in a manner reminiscent of a roller-coaster ride. However, in following the sequence of events I am satisfied that the reason for Dr Bonshek's apparent change of stance was the introduction by Dr Brian Clarke of what he subsequently and rather reluctantly acknowledged to be misleading information which favourably impressed Dr Bonshek at the time. Dr Clarke was contending at the experts' meeting that the literature supported suffocation as a cause of retinal haemorrhages. Dr Clarke now concedes that was wrong. I have no doubt that the false trail gave rise to difficulties between Professor David, Dr Brian Clarke and Dr Bonshek. The final opinion of Dr Bonshek having eliminated this error can be seen from the evolution of his views. In March 2000 he was stating clearly that retinal haemorrhages were extremely suggestive of non-accidental injury. On 6 November 2000 he commented that retinal haemorrhages were consistent with non-accidental injury and there was no evidence of any other explanation. The final position when he gave evidence was that the type and distribution of retinal haemorrhages were typical of shaking and highly likely to be non-accidental.
83. Dr Lawlor and Dr Newbold conducted post-mortem examinations. Dr Lawlor is a consultant pathologist. He has opined that the most likely explanation for A's demise is a combination of being shaken and suffocated. He considered the bilateral retinal haemorrhages to be non-accidental in origin. He said in evidence that he had experience of three cases in the last 2 years where infants had died with retinal haemorrhages, as the only feature, although none had been confirmed as non-accidental injury. He thought shaking likely but asphyxiation a serious possibility. He explained in evidence that one of the difficulties in medicine is that unless there is suspicion of non-accidental injury the eyes of dead infants are not routinely examined and the gold standard for a diagnosis of a non-accidental injury has become subdural haemorrhage, which for many has been the starting point and the end for diagnosis. He gave the cause of death as unascertained because he could not be sure to the criminal burden of proof but his final position was that the combination of shaking and suffocation were the most likely explanation of the haemorrhages and the cardiorespiratory collapse. He found it difficult to explain how bilateral acute retinal haemorrhages could occur without non-accidental injury.
84. Dr Newbold is a consultant paediatric and perinatal histopathologist. She considered the cause of death to be unascertained although she recognised the well-known association of retinal haemorrhage with shaking injury. Her final position was that she recognised that any association between suffocation and retinal haemorrhages as suggested by Dr Brian Clarke at the experts' meeting was flawed and the association unproven. She opined that death was not due to natural causes, on the balance of probabilities was due to a shaking injury either alone or in association with suffocation and that the timing was most probably immediately prior to the child being found. The doctor observed that this was the fourth case this year of a dead baby with retinal haemorrhages and no other injuries. She considered that the retinal haemorrhages were not the only feature to consider and there needed to be an explanation why a perfectly healthy baby suffered a cardiac arrest. This is a matter to which Professor David subsequently referred.
85. All the medical experts agree that this is an unusual case because A exhibited only one of the signs associated with non-accidental injury, namely retinal haemorrhages in both eyes, and there is nothing in the literature nor clinical experience to establish the association with non-accidental injury in the absence of other indicators such as subdural haemorrhages and other signs. Further, the experts have had to consider whether the retinal haemorrhages were secondary to the raised intra-cranial pressure, which would beg the question why baby had such raised pressure in any event, or whether the raised intra-cranial pressure was associated with the retinal haemorrhages, in which case the question arises why did baby have these haemorrhages? None of the many experts contend that this is a SIDS case. The greatest concession is to classify the death as unascertained and every one of the experts agrees that there is an association between retinal haemorrhages and non-accidental injury.
86. In addition to those experts involved in the care of the baby when living and those involved in the post mortem I have had the advantage of other experts brought into the case to give opinions on the significance of the findings. These experts are Dr Barson, a specialist in paediatric pathology, Dr Rushton, a retired consultant in perinatal and paediatric pathology, Dr Phillips, a consultant in paediatric emergency medicine, Dr Selby, a consultant in paediatric intensive care, Dr Geddes, a consultant neuropathologist, Dr Brian Clarke, a consultant pathologist, and Professor David, a professor of child health and paediatrics at Manchester University and honorary consultant paediatrician at Booth Hall Children's Hospital.
87. Professor David was commissioned to provide an overview of the medical evidence. He has analysed all the information and opinions with a view to identifying the issues for the court. He identified two key medical issues: one, what caused the retinal haemorrhages; two, what caused A to collapse and die. In his report he considered 12 suggested or possible explanations for the existence of retinal haemorrhages as identified in life and charted by Dr Smyth.
88. Dr Brian Clarke is consultant pathologist at Moorfields Eye Hospital with a special interest in paediatric eye disorders. He reviewed the post-mortem evidence. He does not have involvement with live children. His opinion was that non-accidental injury by shaking was usually a highly feasible cause of retinal haemorrhages in children under the age of 1 year. He noted that there was nothing to support an innocent or natural explanation, such as a bleeding disorder, septicemia, metabolic disease, to account for the haemorrhages and he excluded CPR and birth haemorrhages. However, he was troubled by the absence of subdural haemorrhage, which for him is an indicator for non-accidental injury. He also identified what he found as atypical features (i) no optic nerve meningeal bleeding, (ii) no far peripheral retinal haemorrhage, (iii) no perimacula retinal folding, (iv) no other signs of physical abuse.
89. He propounded two matters. One, he concluded that since the brain injury was associated with bilateral papilloedema and that most of the retinal haemorrhages were posterior and in the nerve fibre layer in location it was likely that the retinal haemorrhages were a consequence of the raised intra-cranial pressure so that the brain injury might be primary and the retinal injuries secondary. Two, he considered the retinal haemorrhages to be relatively mild in terms of the size and the area of the retina involved and that for him was an important feature of his evidence. He opined that the presence of relatively mild haemorrhages without the usual accompanying intra-cranial haemorrhages and brain injury pattern was insufficient to sustain shaking as the likely explanation in this case. He agreed that Professor David had not omitted any relevant differential condition for consideration.
90. His final position was that he seeks more than one diagnostic element in order to identify non-accidental injury. At the end of the day he could not identify any medical condition to explain the retinal haemorrhages and he acknowledged that eminent doctors felt able to diagnose non-accidental injury on this evidence, although he was in the position that whereas he could not exclude non-accidental injury he had to say the cause of death was unascertained because of (i) absence of subdural haemorrhage, (ii) the lack of severity of multiplicity and the extent and distribution of the retinal haemorrhages, and he stated that his views were strengthened by the diagram because of the concentration of haemorrhages within a tight zone at the back of the eye, no photographs being available, (iii) the tightness of the haemorrhages to the blood vessel, which he had not seen in shaking cases. He concluded he had not come across such a case of shaking and that professionals had to own up to not knowing the answer. He thought the raised intra-cranial pressure was critical and it could lead to growing retinal haemorrhages. He was of the opinion the case did not have what he regarded as the hallmarks of non-accidental injury, although he agreed that other experts diagnosed non-accidental injury more readily than he did and he conceded that he had no direct involvement with examining live children. He thought the retinal haemorrhages were compatible with but not typical of non-accidental injury. He could not exclude deliberate injury but he could not explain the cause of death.
91. In giving evidence Dr Brian Clarke did not rely on the literature which he had quoted at the experts' meeting, and which had given rise to confusion. He has never in fact produced it, despite requests. He was quite adamant that he will not identify non-accidental injury without the presence of other features. I am satisfied that Dr Brian Clarke was in error in describing the retinal haemorrhages as relatively mild. Every expert except Dr Brian Clarke defers to the specialism of Dr Smyth and I find that her description is to be preferred and that these haemorrhages are more accurately described as florid. Further, I find that Dr Brian Clarke was in error in describing the distribution of the haemorrhages. I accept the views of Dr Smyth and Mr Lloyd that the extent and distribution are extended well away from the optic nerve, as the pathologists agree, and they are not all clustered right next to the optic nerve. The significance of this is relevant to the suggestion of raised pressure inside the head, which I find on the evidence of all the other experts is co-related with haemorrhages clustered closely round the optic nerve and which are not asymmetric, as were those of this child. Dr Brian Clarke was in effect isolated in his view. In his report he had said that the alternative explanation of non-accidental injury by upper airway obstruction remained a strong possible cause because he had seen cases where asphyxiation was strongly suspected and retinal haemorrhages were present. He did not pursue that in evidence and his final position was that the cause of death was unascertained.
92. Mr Lloyd, the consultant ophthalmic surgeon and paediatric ophthalmologist, has a very wide experience. He disagreed fundamentally with Dr Brian Clarke and found no difficulty with the distribution nor the severity of the retinal haemorrhages in association with non-accidental injury. The mother has criticised Mr Lloyd's failure to produce a report. I do not find that the parents have been prejudiced in any way because Professor David recorded the detailed conversation with Mr Lloyd, who approved it and signed the record. Mr Lloyd stated that he had personal experience of children surviving non-accidental injury with retinal injuries similar to A's condition. The fact that Mr Brian Clarke did not have that experience was based on him being a pathologist and only seeing, therefore, the extreme spectrum of cases. Mr Lloyd knew of one case in literature quoted by Professor Green of an established case of non-accidental injury where only retinal haemorrhages were present. Mr Lloyd was asked about differential diagnosis. Fits as a cause of retinal haemorrhages were in his opinion highly disputed but just about possible, though it was an extremely rare association. He agreed that raised intra-cranial pressure in some circumstances could cause retinal haemorrhages and bilateral papilloedema were associated with raised intra-cranial pressure. He deferred to Professor David for the overview and opined that having considered everything he could not think of any other plausible explanation except a shaking injury. He had a high regard for Dr Smyth's expertise. He felt confident that A sustained non-accidental injury by shaking in order to cause the haemorrhages as seen and he was satisfied that Professor David had explored every possible differential. He did not expect the haemorrhages to be more severe than they were because he had seen less serious ones in cases of established shaking. He considered that Dr Clarke's specialism was death, in which event they were usually more severe. He stated that although A's appearance was not typical of cases where a child had died the retinal haemorrhages were typical of shaking injury children who have survived. He was satisfied of non-accidental injury, as he put it, to 90% certainty.
93. Dr Rushton sat firmly on the fence. He considered the cause of death as unascertained because the presence of retinal haemorrhages were for him insufficient data from which to draw conclusions. He had no difficulty in acknowledging the association between non-accidental injury and retinal haemorrhages but he had never seen a case of established shaking where only retinal haemorrhages were present and he would have expected to find other indicators. Having considered the evidence of Dr Geddes, Dr Rushton no longer relied on the absence of diffuse axonal injury. He stated he had never encountered CPR causing retinal haemorrhages and he acknowledged that it would be extremely rare for fitting to give rise to such haemorrhages. He could not think of any innocent explanation or cause to account for the haemorrhages but he thought it would be an extension of the current state of knowledge of shaken babies to make the diagnosis on the finding of retinal haemorrhages alone.
94. He together with Dr Brian Clarke considered it possible that intra-cranial pressure led to the retinal haemorrhages rather than vice versa but he was wholly unable to answer the question why then did A have intra-cranial pressure in the first place. He also gave the opinion that baby could have been a near-miss cot death, that he stopped breathing having been found limp for a sufficient period to cause brain damage and thereafter the retinal haemorrhages developed unevenly in each eye by reason of uneven brain swelling. His final position was that although the retinal haemorrhages raised questions of non-accidental injury he was unhappy with that. He thought it could be a condition new to medical science, which is developing all the time. For that reason although trauma was on his list of possible explanations he considered the death not SIDS but unascertained.
95. Dr Geddes was the only other expert who subscribed to the possibility of a near-miss cot death, which is rather a contradiction in terms because this baby did die and that begs the question why.
96. Dr Barson concluded that death was due to shaking and possible suffocation during the half hour before baby was found collapsed. At the experts' meeting he had considered death by suffocation was more likely than death by shaking because of the absence of marks on the body. He has considerable experience of cot death and he noted that he has always found that with an infant who collapses, is moribund and then dies, there is a cause of death as opposed to being a cot death. He too would have described the death as unascertained because he was not satisfied to the criminal standard of proof, although by reason of his clinical experience, which Dr Rushton did not have, he was satisfied on the balance of probabilities that baby died of unnatural causes and the presence of the haemorrhages meant he was a victim of non-accidental injury and there was good evidence for that view. He said that there were no standard constellation of features as signs of shaking.
97. Dr Selby concluded that it was most likely that A died from a shaking-type non-accidental injury, possibly with associated suffocation. He reached that conclusion despite considering the absence of various indicators. Questioned about the possibility of the retinal haemorrhages being secondary to the raised intra-cranial pressure as postulated by Dr Rushton, Dr Geddes and Dr Brian Clarke, Dr Selby did not agree because the appearances of the retinal haemorrhages are not properly described as mild and the distribution did not support such a thesis. Dr Selby said he has seen many children with raised intra-cranial pressure and they do not present with retinal haemorrhages. In his opinion that would be a very rare presentation and in any event the distribution of retinal haemorrhages would not fit. He described himself as 'well off the fence' in that he regarded retinal haemorrhages after deliberate shaking as much more likely than any other explanation. He expressed a firm belief that this was non-accidental injury.
98. Dr Phillips expressed the opinion that on the balance of probabilities baby suffered a non-accidental injury by shaking and concluded that retinal haemorrhages are very rarely the result of intra-cranial pressure. She did not find the evidence convincing enough for her to be certain of non-accidental injury. She had personally never seen a case where a shaking injury produced only retinal haemorrhages. She found it a very difficult case and her final position was that the cause of death was unascertained.
99. Dr Geddes did not support the diagnosis of non-accidental injury because a finding of retinal haemorrhages without subdural bleeding was extremely unusual and there are other causes in medicine of retinal haemorrhages. This doctor had examined the literature and in particular made reference to some 60 cases of inflicted head injury in which retinal haemorrhages were present in 70% of the cases. However, in no example was there retinal haemorrhage without subdural bleeding and the doctor had never seen that. Therefore the cause of death was unascertained, although she did state that the relevance of retinal haemorrhages and their significance were outside her particular expertise. She knew from literature that they raised the level of suspicion of non-accidental injury in infants.
100. Professor David has detailed knowledge of the Sheffield research project undertaken by Dr Geddes and had heard the research paper discussed at a conference. He points out that the group selected were of different ages and all were selected as having subdural haemorrhages. They were being considered by somebody who was not a paediatric pathologist in circumstances in which non-paediatric neuropathologists do not have experience of child abuse. Dr Geddes has subsequently produced a breakdown of the cases studied and it is to be noted that none is comparable to baby A. Dr Geddes, like Dr Rushton, thought the circumstances were compatible with a near-miss cot death and did not fit in with shaking but she was wholly unable to say what caused the swelling of the brain. Therefore the cause of death was unascertained.
101. Professor David was jointly instructed by all parties to consider the evidence and present an overview of medical issues and address differential diagnosis. He has done so, in my judgment, in a closely reasoned, detailed report which considers not only the published literature throughout the world but also draws on the extensive clinical experience which he has. I was very favourably impressed not only by the scholarship as well as the practical experience but also by the careful way in which he addressed the issues in the witness box. I find it highly significant that his knowledge and experience are not limited to a very specialised field. He has the advantage over other experts that he can draw on the narrow but extensive knowledge of highly specialised doctors and relate that to clinical experience so as to see the whole picture. I agree with Professor David that some of the pathologists in this case practice at the extreme end of the spectrum. The eyes they examine are from dead children and they do not have experience of non-accidental injury in clinics in respect of living children, which does handicap perspective. As Professor David said, the eye pathologists who are not even necessarily paediatric eye pathologists live in a limited world.
102. The position regarding the medical evidence is that the majority of the experts agree with the opinion of Professor David that on the balance of probabilities A was the victim of a non-accidental shaking incident with or without suffocation which caused retinal haemorrhages, which led to raised intra-cranial pressure collapse and death. Professor David has, I find, undertaken the most searching analysis of the matters relied on by those doctors who do not agree. Criticism has been made of Professor David by the parents and by the mother in particular in two respects: one, he failed to produce literature which he relied on as demonstrating lack of support for Dr Brian Clarke's propositions; two, he criticised in evidence the Geddes Whitwell research as being not applicable to the case of A, whereas he had not referred to it in his report and therefore did not give Dr Geddes the chance to comment in reply. These criticisms are not supported by the local authority nor by the guardian ad litem. I do not find the criticisms justified and I propose to deal with each in turn.

(1) In his extensive report, in addition to various items of research literature, there were some 411 references to articles. It was open to the parents to access those references as relevant to their case. The request to Professor David after he had concluded his evidence to produce literature relied on was in my judgment misconceived in that not only was it in the public domain already but was relied on in support of a negative, namely that Dr Brian Clarke was isolated from his peers and was not

supported by the disclosed literature.

(2) The ambit of cases referred to by Dr Geddes for the research was set out in a letter dated 27 November 2000 from Dr Geddes to the solicitor acting for mother. The five categories covered do not, in my judgment, apply to A in respect of whom no explanation of the incident was given by a carer in order to ascertain whether there was a discrepancy of history or not. These parents presented the history as a cot death with no explanation to test against medical findings.

103. I find that Professor David has considered every possible differential diagnosis. He agreed this is an unusual case. The fundamental starting point is the observation of Dr Smyth, who saw the eyes in life and who charted them. Professor David acted on her observations, whereas Dr Brian Clarke does not accept them. I find that undermines Dr Clarke's opinion. Professor David considered 12 possible conditions to account for A's collapse. No expert has suggested he has omitted any possible relevant condition. They are: (i) injury caused by shaking, (ii) injury caused by direct trauma to head or eyes, (iii) injury caused by intentional suffocation, (iv) a condition secondary to raised intra-cranial pressure, (v) caused by cardiopulmonary resuscitation at the time of collapse, (vi) caused by epileptic fits, (vii) the result of trauma occurring naturally at birth, (viii) the result of a coagulation disorder, (ix) the result of vomiting or Valsalva-type manoeuvre, (x) other medical conditions such as hypertension, tuberous sclerosis, abnormal blood vessels, inherited metabolic disease, sickle-cell anaemia, disease of the retina, (xi) SIDS, (xii) incorrect or misdiagnosis, that is that the retinal haemorrhages were not in fact present (that in my judgment is not relevant to this case).
104. Of the various conditions considered in a differential diagnosis all the experts agree that many are not applicable to A at all. Fits are theoretically possible as a primary cause but there is a very rare association and no expert has seriously advanced that as an explanation. The two remaining possibilities advanced are: (i) the retinal haemorrhages were secondary to the raised intra-cranial pressure, (ii) A was a near-miss cot death.
105. Professor David concluded, after considering all the evidence, that these could be excluded because there was: (i) florid retinal haemorrhaging, (ii) the nature and distribution of the retinal haemorrhages, (iii) it is established that retinal haemorrhaging can arise with other shaking injuries such as, for example, chest wall bruising or posterior rib fractures but without subdural haemorrhages, (iv) in a healthy baby who has taken a feed there is no viable alternative because such babies do not develop retinal haemorrhages spontaneously and cot-death babies do not have retinal haemorrhages, (v) retinal haemorrhages are associated with trauma. He therefore concluded that although shaking was the most plausible explanation, suffocation could not be excluded.
106. In respect of the retinal haemorrhages being secondary to raised intra-cranial pressure I accept Professor David's analysis that this is non-viable because:

(1) The distribution of the haemorrhages as observed by Dr Smyth demonstrates that they did not radiate from the optic nerve head

and radiation from the optic nerve head is the usual finding in retinal haemorrhages when due to raised intra-cranial pressure.

(2) There was haemorrhaging in the peripheral retina as observed by Dr Bonshek and Dr Clarke, 'haemorrhages around the equator of the globe'. One would not expect to see such haemorrhages if it was due to raised intra-cranial pressure.

(3) The finding of asymmetry of the haemorrhages. Asymmetry is a feature of child abuse that would be unusual in retinal haemorrhages due to raised intra-cranial pressure.

107. Dr Clarke and Dr Bonshek both rejected the description of the retinal haemorrhages as 'florid' and described them as 'relatively mild' in support of their evidence. I accept Professor David's conclusion that no clinician would contend they were relatively mild.
108. In respect of the suggestion of a near-miss cot death this proposition involves a sequence of events starting with a failure to breathe, then brain damage, then retinal haemorrhage as a consequence. In effect this is an explanation advanced for the differential diagnosis which I have already considered. As Professor David pointed out, and I accept, this theory presupposes that retinal haemorrhaging can be a naturally occurring feature of cot death so that many children with SIDS must have had retinal haemorrhages which have been overlooked. As Professor David commented in his report, this would entail retinal haemorrhages having an innocent explanation. The argument is advanced that because it is not routine practice to remove and dissect eyes of infants who have died suddenly and unexpectedly it is possible retinal haemorrhages could be a feature of SIDS. As Dr Lawlor pointed out, there is in my judgment a need to distinguish between absence of evidence and evidence of absence. Although pathologists do not generally examine eyes of infants, paediatricians do routinely and I accept Professor David's opinion that it would be inconceivable for any connection to have been overlooked. I agree with his conclusion that it is positively misleading to suggest retinal haemorrhages as a naturally occurring event in cot death or near cot death.

The bases for Professor David's opinion

109. (1) Some 30 years of clinical experience of paediatric child abuse with a special interest in shaking injuries.

(2) First-hand observation of the child's eyes made by Dr Smyth supported by contemporaneous notes, drawings as amended and a clear recollection of matters seen by an expert of repute.

(3) The opinion of Mr Lloyd, an expert paediatric ophthalmologist whose expertise is recognised nationally and who is a member of the ophthalmology child abuse working party and co-author of the report.

(4) Medical aspects other than retinal haemorrhages in that healthy babies do not die with retinal haemorrhages without a cause. There is not one case in medical literature of a healthy child, perfectly well, feeding a short time before, who stops breathing, has florid retinal haemorrhages and dies as a result of natural causes.

(5) The differential diagnosis excludes any medical condition as a viable explanation other than fanciful explanation.

110. The bottom line is that babies do not develop retinal haemorrhages in the course of normal handling. Experience with major traffic accidents and incidents in the home demonstrate that even a fracture of the skull in an accident or falling downstairs may well not produce retinal haemorrhages. The amount of force needed is such that it would take, I accept, for example a high speed motor car accident, to cause retinal haemorrhages. I accept Professor David's evidence that there is immensely strong evidence that the amount of force has to be high to cause retinal haemorrhages. They are strongly associated with non-accidental injury. Once every known cause is excluded what remains is the answer. The court deals in evidence not theoretical speculation. There is simply no other viable candidate as an explanation for this fit child dying with the features observed. Whether suffocation was a feature of the death is academic because shaking and suffocation are equally serious forms of child abuse. Some children die from a combination of both. The extremely low incidence of retinal haemorrhage in accidental trauma even of a severe nature and the very high incidence in child abuse is very significant and the current reasoning points strongly to shaking as the single most important and common cause of retinal haemorrhages in abused children.
111. I am satisfied that Professor David's overview is authoritative and it persuades me. I find any evidence as to the relative forces needed to produce retinal as opposed to subdural bleeding is unreliable, as he demonstrates. When all feasible explanations are excluded what remains is the truth.
112. I am satisfied, having regard to the whole of the medical evidence, that the local authority has established to the standard of Re H and Others (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, [1996] 1 FLR 80 that A sustained a violent non-accidental injury at the hands of an adult. I do not accept the contention of the parents that the Bolam test must be applied in determining whether a responsible body of opinion can be held so that if there is a respectable thesis to be advanced in contradiction to other opinions it must be accepted. The test for reasonable treatment procedure is in my judgment very different from determining what actually happened to a baby. In care proceedings the court has to test the evidence, piece together the parts of the jigsaw in order to determine whether a clear picture emerges.
113. The timing of the injury is uncertain on the evidence because I find that neither parent has sought to assist the court and it is not possible to determine which of the parents saw A between 3.00 pm and the arrival of the ambulance men. It cannot be established whether the child was shaken to death or unconsciousness with or without gasping breathing. The experts vary in their time-frame according to what may have happened. It is an uncertain science.
114. I do not find it possible to come to any conclusion to identify the perpetrator save to find that it was either mother or father or both. The local authority and the guardian ad litem do not seek to identify the perpetrator because of the confusing evidence and the lack of explanation from each parent. I have found each parent a witness who lacked credibility and who had singularly failed to assist the court. That ends the part of the judgment for publication.

Care order made as sought in respect of surviving children.

PHILIPPA JOHNSON

Barrister

The permission for BAILII to publish the text of this judgment
was granted by Jordan Publishing Limited
Their assistance is gratefully acknowledged.