England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM Coventry District Registry
His Honour Judge Cleary sitting as a S 9 Judge of the High Court
CV16C00513

		Royal Courts of Justice Strand, London, WC2A 2LL
		29 March 2017

Lady Justice King :

1. This is an appeal by Warwickshire County Council ("the local authority") against the dismissal on 20 October 2016 by His Honour Judge Cleary sitting as a high court judge, of its application for a care order in respect of a little girl, ML, who was born 7 December 2015.
2. The central issue was whether ML had suffered a shaking/shaking impact injury whilst she had been in the sole care of her father in the hours which preceded her having been admitted to hospital by ambulance on 19 April 2016. The issue before this court is whether the judge, at the conclusion of a fact finding hearing, convened to determine the cause of ML's injuries and, heard over eleven days commencing 28 October 2016, properly weighed and evaluated all the medical evidence put before him prior to concluding that he "could not hold" that ML's father (or indeed mother) were responsible for the injuries suffered by ML.

Background

3. ML is the daughter of ID ("the mother") and ML ("the father"). Both parents are Lithuanian. The parents began to live together a few weeks prior to ML's birth. They set up home in a multi-occupation property in which they had their own room with a shower, together with the shared use of a kitchen on a different floor of the building. The judge was of the view that "the true nature of the relationship has remained something of a mystery".
4. On 19 April 2016 ML had spent the first part of the day largely with her mother. In the afternoon, sometime between 15.30 and 16.00, ML was taken from the shared kitchen up to the parents' room by the father. Although ML was seen briefly on several occasions by the mother when she went up to the bedroom, ML was thereafter in the care of the father. In broad terms the father's case was that he had attempted to settle ML, who began to cry, he had picked her up, and she had "spat out" her dummy, "given a cough" and fallen unconscious. An ambulance was called and the father carried out CPR until the paramedics attended at the house at 17.45. ML was admitted to hospital, where she remained until discharged into foster care on 3 May 2016. Fortunately, given the seriousness of the brain injury she was found to have sustained, she appears to have made an excellent recovery,
5. A CT scan was undertaken on 19 April and a MRI on 22 April. The clinical picture as revealed in the hospital notes was one of ML having suffered an acute collapse requiring subsequent resuscitation including an episode of encephalopathy (abnormal brain function) and having extensive bilateral retinal haemorrhaging and subdural bleeds of different ages.
6. Although the clinicians at the hospital were not called to give evidence, the judge read and referred to their statements, which can be summarised as follows:

i) Dr Ann Callaghan, Consultant Paediatrician, referred to subdural haemorrhages of different ages and concluded that ML had sustained a "violent assault" immediately prior to becoming unwell. Dr Callaghan concluded that it was most likely that ML had been "violently shaken" on more than one occasion.

ii) Dr Khaled Sherlala, Consultant Neuroradiologist, reported subdural collections of different ages (i.e. chronic and acute). He was of the view that what he observed was "highly suspicious of non-accidental injury".

iii) Ms Dean, Ophthalmologist, found there to be acute retinal haemorrhages consistent with those found in abusive head trauma.

7. These early serious observations, made by in particular Dr Callaghan, led to the involvement of social services and the institution of care proceedings in respect of ML.
8. Within the care proceedings the court ordered expert evidence to be obtained from:

i) Dr Neill Stoodley: Consultant Paediatric Neuroradiologist

ii) Dr Patrick Cartlidge: Consultant Paediatrician

iii) Mr Peter Richards: Consultant Paediatric Neurosurgeon

iv) Mr William Newman: Consultant Paediatric Ophthalmologist

v) Dr Russell Keehan: Consultant Paediatric Haematologist

9. Each of the experts filed reports and all but Dr Keehan gave oral evidence. (Dr Keehan's evidence was accepted and he expressed no opinion as to causation). Both Dr Stoodley and Mr Richards focused on the CT scan and the MRI scans carried out upon ML's admission to hospital. Whilst the two doctors had a difference of opinion as to the precise locations of the bleeding, they were nevertheless in complete agreement with each other (and indeed with Dr Cartlidge and Mr Newman) that the likely cause of the injuries to ML's brain was a shaking/shaking impact injury sustained immediately prior to her collapse.
10. Dr Cartlidge for his part prepared a paediatric overview, going through numerous differential diagnoses and hypothetical scenarios before setting out from a paediatric perspective why, on the facts of the case, he was of the view that what was seen was an acute injury rather than a pre-existing chronic subdural collection. Whilst accepting in evidence that "the uncommon is going to happen uncommonly" at the conclusion of his cross-examination on behalf of the parents Dr Cartlidge remained steadfast in his view that, when considering the injuries both taken altogether and individually, he was satisfied that they were recent and likely to have been caused by trauma.
11. Mr Newman reporting on the injuries to ML's eyes originally observed by Ms Dean on ML's admission to hospital, identified in both eyes multiple retinal haemorrhages which were multi-layer in all four quadrants and which he said were likely to have been caused within 17 days prior to their last identification (namely since 5 April 2016, over four months after ML's birth). Mr Newman, at considerable length and in painstaking detail, examined possible causes for the retinal haemorrhages. He ruled out birth-related retinal haemorrhages and said that normal, (or even rough) handling, minor trauma, low level fall, an attempt to rouse an unconscious or sleeping child, CPR, seizures or raised inter-cranial pressure would not cause, or would be very unlikely to have caused, the retinal haemorrhages he saw in ML's eyes.
12. During the course of his oral evidence Mr Newman confirmed that nothing which had been put to him, or which arose during the course of his oral evidence, had caused him to change his opinion as to either the causation or timing of the retinal haemorrhages. He properly accepted that the presence of retinal haemorrhages should not be regarded as a "trump card".
13. Mr Newman concluded:-

"In the absence of an episode of major accidental trauma, or an underlying medical condition, and in the knowledge that not everything in medicine has a clear explanation, then the intraocular haemorrhages remain unexplained and in my opinion are most consistent with a shaking or a shaking with impact injury."

14. It can be seen therefore that each of the seven doctors who had expressed a view as to causation (the 3 clinicians and 4 of the experts), all specialists in their varying disciplines, were of the opinion that it was most likely that the injuries sustained by ML were caused as a consequence of her having been shaken shortly before she collapsed on 16 April 2016 whilst in her father's care.

The approach of the parents

15. As was rightly observed by the judge it is for the local authority to prove their case. Further, it is properly accepted by Mr Geekie QC on behalf of the local authority that the judge is entitled to reject the evidence of the experts but that in doing so he must give reasons for having reached that conclusion.
16. The mother and father had the benefit of representation by Mr Vater QC and Mr Storey QC, both counsel with considerable experience in representing parents in cases where there is an allegation that a child has, or may have been, shaken. Mr Vater in his submissions to us frankly acknowledged that given the strength of the medical evidence, he had, when embarking upon the case, an "uphill task" if he was to undermine the seemingly overwhelming medical evidence to such an extent that the judge could not be satisfied, on the balance of probabilities, that the injuries had been sustained by ML on 16 April 2016 as a consequence of inflicted injury or an undisclosed accident. To this end, whilst not putting forward a positive case that ML's collapse had been as a result of subdural haemorrhages sustained by her at birth at the hearing before HHJ Cleary, both counsel for the parents explored this possibility in considerable detail and properly exploited areas where, in particular, Dr Stoodley and Mr Richards had not been in complete accord in their evidence.

The Medical Evidence

17. The thrust of the argument of Mr Geekie QC on behalf of the local authority is that the judge failed sufficiently, or at all, to analyse the weight of the medical evidence when viewed as a whole and in particular wrongly excluded consideration of the evidence in relation to the retinal haemorrhages. In order to understand the importance of the evidence in respect of the retinal haemorrhages it is necessary to consider in basic terms the state of the medical evidence overall.
18. Given that the totality of the medical opinion was that the injuries to ML's brain were caused shortly before her admission to hospital, much of the medical evidence focused on the age of the subdural haemorrhages identified in the scans taken of ML. When reading a scan which reveals subdural haemorrhaging, the ageing of those findings is estimated by reference to their appearance on the scan; acute blood shows as 'bright' or 'white' and older bleeds as 'darker' or 'grey'. A clinician reading the scan may identify bright blood as recent/acute and darker blood as old/chronic.
19. It is generally accepted that where there is an existing chronic subdural haematoma, then a relatively minor (or indeed no) incident of a benign nature can cause a so-called re-bleed within the existing subdural effusion. Where there has been a re-bleed, both old and fresh blood can be seen in the same area within the brain with the acute blood being visible within the old bleed. If however acute blood is seen in a geographically different area of the brain from the darker blood, then that is regarded as indicative of there having been two or more separate episodes of trauma having led to the haemorrhages, of different ages, there being no source of 'old' blood to generate a 're-bleed'. Each of the treating clinicians were of the view that they were seeing both old and new haemorrhages of different ages indicating separate traumatic incidents rather than a 're-bleed'.
20. Inevitably, with an organ as complex as the brain, that is not the end of the matter; there is an alternative explanation for areas of darker fluid seen in head injury cases. When the arachnoid membrane is damaged at the time of a head injury, dark cerebral spinal fluid (CSF) can leak from the subarachnoid space into the subdural space and either collect there or dilute acute blood that is present. It can therefore be the case that what is observed on a scan is not a chronic subdural haemorrhage at all but a mixture of blood and CSF which has a similar grey, dark appearance as is seen in a chronic subdural haemorrhage. This phenomenon is known as acute traumatic effusion, and is part of an acute or recent injury.
21. If what were seen on ML's brain were acute traumatic effusions and not chronic subdural haemorrhages, then it follows that there was only one, recent, traumatic event. As chronic subdural haemorrhages resolve, membranes form within them as part of the healing process. If these membranes can be seen on a scan, then that is good supportive evidence that what is seen is a chronic subdural haemorrhage and not an acute traumatic effusion. However, the absence of membranes is not diagnostic of acute traumatic effusion; the experts must draw on other aspects of their findings and indeed their experience and expertise in order to reach a conclusion as to whether what is seen is, or is not, a chronic collection.
22. Subdural haemorrhages occur at birth in something like 50% of the population. Research is limited but seems to suggest that any such subdural haemorrhages have largely cleared by four weeks and have completely resolved by three months. D was four months old. Dr Stoodley gave his evidence by reference to four weeks and Dr Cartlidge and Mr Richards to three months.
23. Regardless of the date by which they would have expected any birth subdural haemorrhages to be resolved, both Dr Stoodley and Mr Richards were of the view that what they saw on the scans were not chronic subdural haemorrhages, (whether dating back to birth or otherwise) but acute traumatic effusions and therefore indicative of recent trauma. Further, each was clear that, regardless of whether the darker density in fact revealed chronic subdural haemorrhage or acute traumatic effusion, an explanation would still be required to account for the presence of the acute bilateral retinal haemorrhages which were found by Mr Newman to be no more than 17 days old and which, it was accepted by common consent, were unconnected with the subdural haemorrhaging.
24. From this crude outline of the issues thrown up by the medical evidence, it can be seen that in circumstances where the clinicians (as opposed to the court appointed experts) had referred to "chronic subdural haemorrhages of various ages" and where there were no membranes visible on the scars, there was ample material for cross-examination on behalf of the parents. If what was seen were chronic subdural haemorrhages, rather than acute traumatic effusions, that opened up a line of cross-examination as to whether these may have been birth injuries. Further, if Mr Richard's rather that Dr Stoodley's observations as to the precise location of blood were accepted then all the acute blood identified on the scans could be attributed to 're-bleed' into an old subdural haemorrhage, notwithstanding the clinicians' reference to more than one incidence of trauma.
25. Each of the experts was extensively cross-examined. There was a considerable amount of evidence in relation to ML's head circumference and the significance of seizures, which it is not necessary to examine for the purposes of this appeal. A number of possible alternative explanations for the head injuries were put to each of the experts (including birth subdural haemorrhage). Mr Vater described these, admittedly remote, possibilities as outliers (an expression which when used in statistics means, (per the OED), "an observation whose value lies outside the set of values considered likely according to some hypothesis, usually one based on other observations). Each of the possible outliers put to the experts was rejected by them, they used in response expressions such as 'never say never' and 'I can't exclude it but it is highly unlikely' or 'improbable'.
26. As the judge found, this was not a case where the experts had closed their minds to the possibility of alternative innocent explanations to account for ML's injuries. He described their evidence as follows:-

"[20]…..The evidence of each of these clinicians, in their written reports appeared to me to be well researched, soundly based, and well expressed. Each of them, at the stand, was refreshingly open and honest. They each acknowledged the presence of research in their respective fields, and the existence of cases of which they were aware or in which they appeared, when differences between them and their colleagues (in those cases) were exposed or when evidence at the hearing was established to provoke a change of opinion."

27. At the conclusion of the medical evidence, the judge recorded [11] that 'None of the experts changed their overall opinions'. Accordingly, at the conclusion of their evidence:

i) Each of Mr Richards, Dr Stoodley and Dr Cartlidge was of the view that the imaging was more likely to represent fresh blood and acute traumatic effusion rather than fresh blood and a chronic collection

ii) Mr Newman's opinion in relation to the retinal haemorrhages was that:

"In the absence of an identifiable medical condition or history of significant trauma, the retinal haemorrhages identified in ML remain unexplained but in my opinion would be most consistent with those found following a shaking or shaking with impact injury"

iii) It was common ground and was, and is, accepted by Mr Vater that a chronic injury could not account for the retinal haemorrhages.

iv) All of the experts remained of the view that there was nothing in the parents' accounts that explained ML's collapse, the encephalopathy, and the injuries subsequently found.

v) The court appointed experts were of the view that the combined effect of: ML's presentation, the injuries themselves and the absence of a plausible explanation, meant that it was most likely that ML had been injured by a shaking/shaking impact injury in a manner that one or more of her carers had chosen to conceal.

28. It was against the backdrop of this medical evidence that the judge went on to hear the parties give evidence to the effect that they could offer no explanation for ML's sudden and unexpected collapse whilst in her father's care.

The Judgment

29. Having set out the background the judge recorded that the standard of proof is the balance of probabilities and the burden of proof rests upon the local authority. He further reminded himself that:

"Thirdly, there is no room for a conclusion which consists of only a possibility or suspicion. The test is binary. Either the local authority succeeds in the task it sets itself or it does not. The court is not permitted to shroud its judgment in uncertainties, and thus it must either declare that it is satisfied, on the balance of probabilities, at the occurrence of an event or that it is not. And if it is not, then the assertion is rejected. There is no room for 'maybe' or 'possibly'."

30. The judge identified the various experts and recorded that none had any doubt as to their conclusions and that each was satisfied as to the outcome and to the cause of ML's injuries, namely that ML had sustained a recent traumatic injury. The judge said that:

"[23] It is necessary therefore to establish where each of them were challenged, and on what basis, and to conclude whether or not the skilful interrogation by counsel for the respondents produced doubt or uncertainty into their respective conclusions."

31. It is a feature of the judgment, and a cause of criticism by Mr Geekie QC on behalf of the local authority, that the judge throughout his judgment spoke in terms of "doubts". To approach the case by reference to "doubts" meant, Mr Geekie submitted, that if the local authority were to prove its case it was necessary to exclude all other possibilities or uncertainties but:

i) such an approach imports the criminal standard of proof, namely "satisfied so that you are sure"

ii) is a scientific impossibility in any event given that, as Mr Richards puts it in evidence, doctors "never say never"

32. The judge set out in some detail the evidence of each of the experts. He dealt with the evidence of Dr Stoodley and Mr Richards identifying the difference between them as to the location of all the blood seen on the scans. The judge set out the evidence of Dr Cartlidge noting his important observation that even if what was seen were chronic subdural haemorrhages dating back to birth (which he did not accept) an explanation would still be required for the unconnected retinal haemorrhages.
33. It was in that context that the judge turned to the evidence of Mr Newman referring to his report as having been "hugely informative, detailed and comprehensive" and his evidence as "candid and helpful". The judge briefly set out in one paragraph of the judgment the "unknowns" in relation to retinal haemorrhages but did not thereafter record Mr Newman's specific evidence that, notwithstanding the "unknowns", his opinion was that the likely cause of the retinal haemorrhages was a recent shaking or shaking/impact injury. The judge said in relation to the evidence of Mr Newman:

"[61] It appears to me to be uncontroversial to conclude that retinal haemorrhages by themselves prove nothing – they are part of the wider canvas. As Mr Newman volunteered, their existence is not a trump card and it has to be taken in the context of paediatric and radiological evidence. But I ask myself, if we look at the local authority's hand, where are the trump cards? Which cards are of low or high value, and if taken together, and there are discrepancies, should the value of one be reduced by the undervalue of another?"

34. It should be noted for completeness that the judge's recollection of the evidence is at fault there; the transcript reveals that the reference to a "trump card" did not come from Mr Newman but was put to Mr Newman by Mr Vater:

"Q… And you would be the first, I think, to accept that the presence of retinal haemorrhages should not be regarded, as it were, as the trump card. Do you understand what I mean by that? "

A… That is correct."

35. The judge did not therefore assess the impact of Mr Newman's evidence on the other evidence. Had he done so, Mr Geekie submitted, he would have appreciated that Mr Newman's evidence rendered it significantly more unlikely that the subdural haemorrhages were from a chronic injury as such a cause could not account for the acute retinal haemorrhages.
36. The judge, in pulling together his conclusions in relation to the medical evidence, speculated as to whether or not a "leak or bleed from an asymptomatic collection onto the surface of the brain would cause irritation. That, he suggested, could lead to a seizure". He went on:

"[68] I am not convinced one way or the other. Equally, however, I am troubled by some of the uncertainty presented by the clinicians whose evidence I found to be candid and open. And I remind myself that I am not exercising criminal jurisdiction and I do not have to be satisfied so that I am sure of the picture which they present, a picture which is unanimous in its conclusion, namely that ML suffered a shaking injury. Nonetheless, I am left with doubts. . . ."

37. In his first draft this paragraph concluded at this point. Having read the draft, Mr Vater drafted and proposed the inclusion of the following passage identifying the nature of those doubts:

" First, whether all of the subdural collections were recent in origin or whether they represented acute re-bleeding into chronic collections of uncertain origins; secondly whether ML's collapse was as a consequence of acute injury or alternatively a seizure caused by pre-existing subdural collection; and thirdly whether the causes of any or all of these things might simply be unknown. Notwithstanding the unanimous conclusions reached by the clinical experts, and given that Mr Newman acknowledges that retinal haemorrhages do not represent a trump card (and thus their existence does not assist me in resolving this uncertainty), I have then to turn to the lay evidence to complete the jigsaw which I am endeavouring to put together in my overall analysis of the case. "

38. The judge indicated during subsequent argument as to the appropriateness or otherwise of the inclusion of Mr Vater's addition that the paragraph as drafted by Mr Vater in fact reflected his own views. Mr Geekie in those circumstances did not pursue a separate ground of appeal in respect of this addition to the judgment.
39. The judge concluded:

"[106] The candour and assistance which all the experts have provided to this court has left me with the doubts which I have no doubt clumsily attempted to set out in this judgment. None of the experts, by themselves or collectively, persuade me that these doubts do not exist. They are not removed by the lay evidence and indeed they are reinforced."

40. The judge moved on to consider the evidence of the parents. Whilst looking at the numerous inconsistencies in the accounts given by the father in relation to the events of that late afternoon, the judge posed and answered a number of questions [99] before concluding:

"[100] It will be observed that none of these questions have been answered by me in a way which points towards the father's responsibility for the calamity which befell his daughter. My answers reveal that I am in doubt. . . . ."

41. The judge concluded his judgment by saying:

"[108] Consequently, given the binary test to which I no doubt inelegantly referred at the outset of this judgment, I conclude that I cannot hold the father responsible for the injuries suffered by his daughter. . . ".

The judge accordingly declined to make the findings sought against the father. This was a single issue case, there being no other basis for intervention by the local authority; therefore absent the judge making findings of non-accidental head injury the threshold criteria were not established and the care proceedings were dismissed. The local authority indicated to the court in submissions that it no longer pursued findings against the mother and accordingly, shortly after the judgment, ML returned to her care where she remains.

The Appeal

42. The grounds of appeal are simply put as follows:

"The learned judge's conclusion that he could "not hold [the father] responsible for the injuries suffered by his daughter" was wrong in that:

(i) He wrongly directed himself in law that he could not find the local authority case to be proved if he had "doubts" about the case or if "uncertainties" remain.

(ii) He failed sufficiently to explain why he was departing from the unanimous opinion of four independently instructed experts who advised that, on the balance of probabilities, the injuries were non-accidental or inflicted.

(iii) He failed sufficiently or at all to consider the effect of the ophthalmological evidence.

(iv) He failed sufficiently or at all to analyse the weight of the medical evidence when viewed as a whole."

43. Mr Vater submitted that:

i) The judge having heard the lay evidence was entitled to reject the medical evidence. To interfere with his conclusions at the end of a long finding of fact hearing would go dangerously close to saying that where medical evidence was unanimous it was also conclusive.

ii) The judge had not in any event rejected the medical evidence but had concluded, as he was entitled to, that the local authority had not discharged the burden of proof.

Discussion

44. Mr Vater referred the court to well-known passages in the authorities which properly emphasise the difficulty an appellant will have in dislodging a finding of fact made by a judge after a substantial hearing where he has heard and seen the witnesses give evidence. For recent iterations on the point see Re F (Children) [2016] EWCA Civ 546 and T-B-N (Children) [2016] EWCA Civ 1098.
45. Mr Vater further explained that this judge had been put under considerable pressure by those representing the mother to deliver a judgment overnight as ML was in foster care and the anticipated exoneration of the mother would enable her to be returned forthwith to her care. In any event, Mr Vater told the court the judge was driven to give judgment on the Friday (submissions having been made on the Thursday) as he was due to start another 10 day case the following Monday. In this context Mr Vater submitted that the second 'limb' of the doctrine found in Piglowska v Piglowski [1999] 1 WLR 1360; [1999] 2 FLR 763; [1999] 3 All ER 632, HL is particularly apposite:

"The second point follows from the first. The exigencies of daily courtroom life are such that reasons for judgment will always be capable of having been better expressed. This is particularly true of an unreserved judgment such as the judge gave in this case but also of a reserved judgment based upon notes, such as was given by the district judge. These reasons should be read on the assumption that, unless he has demonstrated the contrary, the judge knew how he should perform his functions and which matters he should take into account. This is particularly true when the matters in question are so well known as those specified in s 25(2). An appellate court should resist the temptation to subvert the principle that they should not substitute their own discretion for that of the judge by a narrow textual analysis which enables them to claim that he misdirected himself." [p784]

46. This case highlights the considerable pressure under which circuit judges trying these serious cases operate. The judge's desire to give a judgment overnight so as to allow ML to be reunited with her mother as quickly as possible shows him to be the humane and compassionate judge he undoubtedly is.
47. The grounds of appeal in this case however do not go to the judge's assessment of the credibility of the witnesses as such, nor do they rely on some unrealistic counsel of perfection in relation to the drafting of the judgment, (grounds of appeal which, I should say, are regarded with considerable scepticism by an appeal court conscious of the time pressure under which circuit judges work and of their particular skill in assessing the credibility of witnesses). In the present case the grounds of appeal are that the judge failed adequately to set out his reasons for failing to accept the unanimous medical evidence. More particularly it is said that the judge was wrong in having failed to take into account Mr Newman's evidence in relation to the retinal haemorrhages when considering the possibility of birth injury or outliers as having provided a sufficiently plausible alternative explanation for the head injuries such that, notwithstanding the preponderance of the medical evidence, the local authority had failed to discharge the burden of proof.
48. It goes without saying that the judge has to weigh all the evidence. The fact that the judge had 'doubts' may, but does not necessarily, translate into a finding that the local authority failed to establish on the balance of probabilities that ML had suffered an inflicted head injury. As Mr Geekie correctly submits, in assessing evidential sufficiency it is not necessary to dispel all doubts or uncertainty (see for example Re B(Non-Accidental Injury) [2002] EWCA Civ 752;[2002] 2 FLR 1133 at [25])
49. Mr Vater told the court that in putting the case on behalf of the father, he had not put forward a positive case that ML's injuries were caused at birth, rather, he said, his case was that this was a case where it was possible that the injuries were caused by either birth injuries or an outlier. The judge's doubts as to the cause of the brain injuries as expressed by him in his judgment having heard the cross-examination of the experts to that effect, entitled the judge, Mr Vater says, to conclude that the local authority had not discharged the burden of proof to show on the balance of probabilities that ML had sustained an inflicted injury shortly before her collapse.
50. I am unable to accept Mr Vater's submission. The judge, when considering the medical evidence as a whole, neither considered the inherent improbability of the case being put to the doctors by Mr Vater and Mr Storey, nor took into account the constellation of medical co-incidence which their case demanded, (that is to say the presence of chronic subdural haemorrhages together with the wholly unconnected acute retinal haemorrhages). What, however, in my judgment is fatal to the judge's judgment is his treatment of the evidence of Mr Newman in respect of the retinal haemorrhages.
51. The evidence before the court was that before it could be concluded that what was seen on the scans was, or may be, chronic subdural haemorrhages (a finding which would indicate that ML had a pre-existing injury and which appeared to be the judge's preferred option) an explanation had to be found for the unequivocal evidence of Mr Newman that ML had recent, wholly unconnected, bilateral retinal haemorrhages consistent with a shaking injury.
52. With respect to the judge, he seems to have been led down the wrong path by Mr Vater's cross-examination during which Mr Newman, readily and properly, accepted that the presence of recent retinal haemorrhages are not a 'trump card'. No one has or could suggest that the presence of retinal haemorrhages is pathogenic of a shaking injury. It would be quite wrong to do so. Their presence might however have been regarded as having been capable of providing strong support for the expert evidence of Dr Stoodley, Dr Cartlidge and Mr Richards that what was seen were acute traumatic effusions indicative of a trauma immediately prior to ML's collapse. The judge, having said that "retinal haemorrhages by themselves prove nothing - they are part of the wider canvas", in my judgment thereafter wrongly left the evidence in relation to them entirely out of the equation, failing to take the presence of the retinal haemorrhages into account as 'part of the wider canvas' or at all. Instead he concluded at [68] that:

"Notwithstanding the unanimous conclusions reached by the clinical experts, and given that Mr Newman acknowledges that the retinal haemorrhages do not represent a trump card (and thus their existence does not assist me in resolving this uncertainty), I have then to turn to the lay evidence to complete the jigsaw which I am endeavouring to put together in my overall analysis of the case."

53. Mr Vater falls back on to Piglowska and the undoubted pressure the judge was under to submit that the failure of the judge to take into account the retinal haemorrhages should not be allowed to undermine his fundamental finding that he was not satisfied on the balance of probabilities that ML had suffered a traumatic head injury shortly before the ambulance was called. Even having accorded the judge every allowance for the situation in which he found himself, I cannot agree. In my judgment the judge needed to explain why his 'doubts' led to him being unable to find, notwithstanding the medical evidence, that on the balance of probabilities ML had been shaken; such an explanation could only be provided by analysing the totality of the medical evidence which he failed to do having set aside the evidence in respect of the retinal haemorrhages
54. In my judgment the judge was wrong to exclude the evidence in respect of the retinal haemorrhages when considering whether his 'doubts' could be resolved one way or the other and whether, notwithstanding those 'doubts', he was nevertheless satisfied on the balance of probabilities that the local authority had proved their case.
55. Whilst that is not to say that a court cannot reject expert medical evidence, a court in doing so should have well in mind the case of Re M-W (Care Proceedings: Expert Evidence) [2010] EWCA Civ 12; [2010] 2 FLR 46, where Wall LJ (as he then was) emphasised the importance of providing reasons and analysis if medical evidence is to be rejected. He said:

"[39] I regard the following as trite propositions of law:

(1) Experts do not decide cases. Judges do. The expert's function is to advise the judge;

(2) The judge is fully entitled to accept or reject expert opinion;

(3) If the judge decides to reject an expert's advice, he or she:

a. Must have a sound basis upon which to do so; and

b. Must explain why the advice is being rejected;

(4) Similar considerations arise when a judge prefers one expert's evidence to that of another. Judges must explain why they prefer the evidence of A to that of B"

56. Mr Geekie rightly accepts that the judge had to consider the medical evidence alongside the lay evidence and that it is well-established that there will be cases in which a judge is satisfied by the lay evidence and rejects, even seemingly convincing, expert evidence. However, he submits, when assessing the lay evidence against the expert evidence it is first necessary for the expert evidence to have been analysed correctly and that, he says, the judge failed to do.
57. I accept the submission of Mr Geekie. By failing to take into consideration the evidence of Mr Newman the judge omitted properly to analyse and conduct an overall review of the medical evidence. In my judgment, in setting aside the evidence in respect of the retinal haemorrhages, the judge fell into the trap of addressing the head and eye injuries separately and in doing so, failed to consider if they were inter-related or to take into consideration that the court was being presented with a unifying medical hypothesis to explain the constellation of symptoms exhibited by ML upon his admission to hospital. Only by having carried out such an overall analysis of the medical evidence could the judge have then properly moved on to balance his conclusions in relation to the expert evidence against the lay evidence and, in particular, the accepted contradictions within the father's evidence as to the events leading up to the collapse of ML whilst in his sole care.

Conclusion

58. I would therefore allow the local authority's appeal and order the matter to be remitted for re-hearing. In the first instance the case should be listed before the Family Division Liaison Judge for directions and allocation. I should note in conclusion that there were a number of issues which were considered at length at trial and touched upon in this appeal but which it has not been necessary for this court to consider in any detail in order to decide the appeal. These various issues will, no doubt, feature in a re-trial and will necessarily form part of the trial judge's evaluation of the medical evidence. This is not a case where it would be appropriate for this court to substitute its own findings and fairness therefore requires that the matter is remitted for rehearing.