B (children) , Re

LORD JUSTICE WARD: A wedding party on 6 November 2004 should have been a happy occasion. Instead it set in play the launching of care proceedings in respect of three young children, and an order made in the Plymouth County Court on 24 June 2005 by HHJ Tyzack QC that:

“… on a balance of probability, on a date unknown but 2-3 weeks before 10.11.04, either the mother or the father (but likely to be the mother) caused [J] to suffer bilateral subdural haemorrhages probably by shaking her on one occasion. The retinal haemorrhages probably occurred at the same time.”

As a result of that finding, supervision orders in respect of the three children were made on 2 February 2006, placing the children under the supervision of the local authority for 12 months. The local authority have, however, taken the sensible view that they will apply for the discharge of those orders in July; that is to say, after they will have been in force for only five months. The parents now appeal to this court for permission of the court against those findings of fact.

The parents are a married couple. They have three children, a boy born on 30 July 2001 and twin girls born on 25 November 2003. A guest at the wedding was holding the twin, J, when, in circumstances the judge was satisfied were completely accidental, J fell backwards onto the floor, a distance of about 3.5 feet. The father described the scene of his arrival in these terms:

“… she was on her back, her hands were straight with her thumbs up; her eyes were rolling around in her head; she was coughing; I turned her into the recovery position since it is part of my job [at B&Q] to know about first aid; I thought she was going to die she looked so poorly …”

As the judge found, entirely appropriate steps were taken to call an ambulance and J was taken to the local hospital, where she was admitted overnight for observation and was discharged home at about 1.00pm the following day, 7 November 2004. She was 11 months old at the time. The following day, 8 November, J was taken back to hospital by her parents, concerned at vomiting and a “vacant” episode. Head injury advice was given and she was discharged home. On 9 November, the mother was so concerned about J, who had been vomiting and had a floppy episode, when she could not raise her head, that the general practioner was consulted. He recommended calling an ambulance and J was admitted back to hospital about 6.00pm. On 10 November a CT scan was taken and the provisional report was that it was normal, but a later report found increased frontal spaces and unable to rule out subdural haemorrhage. An MRI scan was carried out on 11 November and the hospital notes record:

“… abnormal fluid in the subdural space on both sides … The signal is that of old blood, i.e. chronic subdural haematomas.”

The appearance was consistent with haematomas of at least two weeks old. There was an initial opthalmological examination on 13 November revealing retinal haemorrhages and the consultant ophthalmologist expressed the view that the lesions appeared acute, i.e. less than two weeks old and no evidence of long standing haemorrhage. Another ophthalmologist examined J the following day and noted that the retinal haemorrhages looked quite fresh, were small and should absorb very quickly.

The concern was therefore raised that J had suffered non-accidental injury. The police became involved and interviewed the parents on 16 November 2004. They could offer no explanation for the injuries, apart from the wedding fall. Later on 7 December the mother mentioned to a social worker that she had recollected that in a part of July, she had fallen whilst holding J in her arms. I shall refer to this as “the July fall”. On 17 December the local authority issued applications for care or supervision orders in respect of all three children, no doubt basing their proceedings on the consultant paediatrician’s report of 30 November, that:

“‘Head scanning (CT and MRI) showed no major fresh bleeding, but chronic (more than 2 weeks’ old) haemorrhages in the subdural space. These are typical of shaking type injury, imaging showing no suggestion of antenatal or perinatal brain damage.’”

At a preliminary fact finding enquiry held by the judge the local authority conceded that, apart from the subdural haematomas and the retinal haemorrhages, there would be no basis for the order they sought. The medical evidence was therefore of great importance in this case.

The judge correctly directed himself in accordance with Re U(a child) (Serious Injury: Standard of Proof) Re B [2004] EWCA Civ 567, [2004] FLR 263, that the judge:

“… surveys a wide canvas, including a detailed history of the parents’ lives, their relationship and their interaction with professionals. There will be many contributions to this context, family members, neighbours, health records, as well as the observation of professionals such as social workers, health visitors and the children’s guardian …”

The judge again correctly directed himself that the burden of proof was on the local authority and that a standard of proof is the civil standard of a balance of probability, as that has been explained by Lord Nicholls of Birkenhead, in Re H and R (Child Sexual Abuse: Standard of Proof) [1996] App Cases, 563(8)(6):

“The balance of probability standard means that a court is satisfied an event occurred if the court considers that on the evidence the occurrence of the event is more likely than not. When assessing the probabilities the court will have in mind as a factor, to whatever degree is appropriate in a particular case, that the more serious the allegation the less likely it is that the event occurred and hence the stronger should be the evidence before the court concludes that the allegation is established on the balance of probability … deliberate physical injury is usually less likely than accidental physical injury … built into the preponderance of probability standard is a generous degree of flexibility in respect of the seriousness of the allegation. Although the result is much the same, this does not mean that where a serious allegation is the issue the standard of proof required is higher, it means only that the inherent probability or improbability of an event is itself a matter to be taken into account when weighing the probability in deciding whether on balance the event occurred. The more improbable the event, the stronger must be the evidence that it did occur, before on the balance of probability its occurrence will be established. Ungoed-Thomas J expressed this neatly in Re Dellow’s Will Trusts [1964] 1 WLR 451: ‘The more serious the allegation the more cogent is the evidence required to overcome the unlikelihood of what is alleged and thus to prove it …’”

The judge observed that the allegation made by the local authority against each parent is, as he described it, a most grave one, namely that one or other is responsible for inflicting upon their daughter a deliberate shaking-type trauma causing serious injury. The structure of the judgment, and the conclusions of which the judge arrived, can be explained as follows.

In paragraph 19, he dealt first with the wider canvas, into which context he would be setting the medical evidence. He concluded that the picture presented was “almost entirely favourable to the parents.” The one matter of concern related to a period in the mother’s life when she suffered from depression, from which however she had made a full recovery without there being any suggestion of its recurrence. The conclusion was that the evidence overwhelmingly points to:

“The mother and father being good and loving parents and part of a close-knit family.”

The health visitor described the children as:

“Delightful, confident and sociable young children, who smile readily as they engage with each other and family members and explore their environment.”

She observed the mother’s:

“… obvious enjoyment of her babies … (she) … engages with J and C in a gentle and affectionate manner.”

Her handling of the children was confident. As the judge found in paragraph 22, in all she presented a very positive picture of this family and of the high standard of care given to the children.

The judge then addressed the medical evidence referred to the report of Mr Cole, the consultant ophthalmologist and added in paragraph 27 that:

“In his oral evidence he said that such haemorrhages could have been caused by the ‘wedding fall’, or an episode prior to that. He expressed difficulty in being able to date them and said that they would have occurred within weeks or months of presentation.”

Next he dealt with the evidence of Dr Stoodley, a consultant neuroradiologist at the Frenchay Hospital in Bristol, called by the local authority. He gave a summary of his evidence, the salient features of which were, to read from paragraph 28 of the judgment:

“… this ‘always has to be a diagnosis of exclusion’. [C19] What he is saying in this case is all other possible causes can be excluded which means we are left with non-accidental injury -- since a child as young as [J] could not have caused the injuries herself. The mechanism of the injury is his opinion ‘likely to be one of shaking/impact’. In cross-examination he said that ‘… shaking is the most significant mechanism in head trauma, the repetitive backwards and forwards; I mean shaking in the sense of acceleration/deceleration … The degree of force required is not known, he said, but is likely to be at least such that an independent observer would recognise that it would harm a child’[C20].”

In paragraph 33 he added:

“In terms of changes of behaviour, his evidence was that if a carer had shaken [J] there could have been a brief period of unconsciousness followed by a range of possible symptoms: crying inconsolably; vomiting feeds; droopy mouth; lethargy; not sleeping through the night. The perpetrator would realise that any such symptoms were caused by their actions [C20] but could easily be unaware that any intracranial injury had occurred.”

As for the July fall, he said in paragraph 31:

“He said it was possible for subdural haematomas to occur following an impact such as a fall onto a concrete paving stone but he would have expected such ‘impact blood’ to disappear or resolve after about 2 weeks (Dr Johnson gave a slightly longer time frame -- up to four weeks).

He excluded the wedding fall, saying in paragraph 34:

“His answer is that whilst both falls are within the ‘window’, it is unlikely that either was responsible. The reason he gives for excluding the wedding fall is of course that he saw no evidence of acute -- that is fresh blood. What he saw was chronic blood which means that it has to be dated at least 2 weeks prior to the scans. In his oral evidence he said there was no evidence on the scans (i.e. acute blood) of a fall within 2-3 weeks hence his ability to say it was unlikely that the wedding fall is responsible.”

As for the parents’ evidence, the judge recorded in paragraph 36 how the father in interviews with the police could give no real explanation for the injury, though he mentioned the possibility of the young son causing the injury by rocking J in a bouncy chair. No party initially mentioned the July incident to the doctors at the hospital. That only arose later. As for the mother’s account, he set out the account in the mother’s first statement, in which she gave details about what had happened in July 2004, when she took the twins to see her mother:

“… how she was walking up a concrete slab path when she suddenly missed her footing and fell down holding [J] in her left arm. She says she heard a thud and then [J] screaming. ‘I believe she hit her head on the ground. I am not sure whether she hit her head on a paving slab. However I believe she could have done so as I cut and scarred my left elbow.”

That is in paragraph 39 of the judgment.

In paragraph 40, the judge records that in her oral evidence the mother explained how she had suddenly remembered the July incident, but observed that:

“The picture I have about how this came to light is a confusing one.”

He added in paragraph 41 that the grandmother was also very confused about timing. The father in his oral evidence had said that the July incident was first remembered in January, sometime before his birthday on 26 January. He then explained how the mother had woken him in the night and asked him if the July fall could be relevant. He said that they should discuss it in the morning – sounds a typical response – and in the morning she told her father, who was staying with them, and he advised her to go to the police.

The judge said there was a surprising lack of clarity about how this all came to light and added in paragraph 43:

“This incident in July is put forward by the parents as a possible indeed highly credible explanation for subdural haematomas – hence the need for Drs Stoodley and Johnson to deal with it. However as I have already indicated I am quite satisfied the incident has been greatly exaggerated. I shall set out my reasons in a moment, but I pause to observe that if Dr Stoodley and Dr Johnson had been aware of that, the likelihood in my judgment is that neither would have attached much significance to it. Both doctors were of necessity proceeding on the basis that the mother’s account was accurate.”

He then reverted back to the medical evidence. Observing of Dr Stoodley that he could not completely exclude either fall, but did not think either accident had led to the bleeds that he saw, he turned to Dr Johnson, who was a consultant radiologist at the Birmingham Children’s hospital, originally instructed by the mother, but in fact called by the guardian for the children. He confirmed Dr Stoodley’s findings of chronic subdural haematoma, greater than two to three weeks’ old and that there was no evidence of acute, i.e. recent, haemorrhage.

The judge recorded in paragraph 46 the opinion of Dr Johnson that:

“… it is unlikely (but not impossible) that the fall described in July 2004 caused the subdural haemorrhages.’ The wedding fall he finds even less likely to be the cause of it because of the difficulty over timing.”

He pointed to a difference in the evidence of the two experts, saying in paragraph 48 that Dr Johnson was:

“… less confident than Dr Stoodley in the MRI procedure and said that he would prefer to rely on CT scans, finding the latter more accurate. He would not be confident of dating using MRI alone. He concluded his evidence by saying that the wedding fall was an unlikely cause; the July cause was possible but unlikely; and he was not aware of any other cause. Therefore the likelihood of this being a case of non-accidental injury was significant; it was more probable than not, he could not be 100% [certain].”

The judge’s conclusions needed to be clarified as a result of a query raised by Wall LJ, when he dealt with the application for permission to appeal on paper. In answer to that query, Judge Tyzack wrote back and said this:

“I find that on a balance of probability on a date unknown, but two to three weeks before 10 November 2004, 1) either the mother or the father caused J to suffer bilateral subdural haematomas, probably by shaking her on one occasion; 2) it is not possible on all the evidence I have heard and on a balance of probability to decide which of the two were responsible or whether both were responsible; 3) however, and with a view to assisting the parties and indeed the professionals in the future planning and conduct of this case, including the carrying out of any risk or other relevant assessment, as between the two parents I find there to be a rather greater likelihood that the mother was the perpetrator than the father because, as I have said above, she was the primary carer and I found her evidence to be far less credible than the father; 4) but having said that, I ought to make it quite clear that although there to be rather more evidence against the mother than the father, [it] is still insufficient in my judgment to be able to make any more precise finding.”

May I begin by expressing the court’s gratitude to Mr Tolson QC and Miss Vincent, who had appeared on behalf of the mother, and to Mr Ward, who appeared on behalf of the father, for providing the services pro bono. Our thanks are also due to Mr Braithwatie, who appears for the local authority, for the restrained and fair way in which he presented a difficult case on behalf of his client, the local authority, which itself has acted with commendable propriety and fairness in the exercise of its public duty to protect the children within the county.

The submissions of Mr Tolson have been refined by his responses to the probing and prompting of the court. In the end, the attack on this careful judgment is developed on two lines. The first relates to the judge’s finding that the mother had deceived the court by giving a false account of the July fall. Mr Tolson complains that her honesty was never challenged and that to make this finding without her having been given the opportunity to deal with it is procedurally unfair.

Secondly, it is that in all the circumstances of this case, which include the wider canvas, the evidence was not cogent enough to compel the conclusion that the subdural haematomas were caused by some form of shaking whilst the child was in the care of her parents.

As to the mother’s dishonesty, the judge made these findings about her. In paragraph 52 he said:

“ I regret to say I did not find the mother a particularly impressive witness. What really troubles me is not only the way the July fall came to light but also that the alleged incident itself lacks credibility. I am prepared to accept that [J] may have been involved in an incident of some kind, at the grandmother’s home in July but in all the circumstances I am driven to the very clear conclusion that it was nowhere near as serious as has been painted by the mother. If it had been I am quite satisfied that this mother (who has specifically described herself as being rather over-protective) would have mentioned it to the staff at the hospital and also to the police. It is also likely that, irrespective of her mother’s word of caution, she would have taken [J] to the doctor and would have been likely to have mentioned it when she next saw the health visitor. A fall of a kind she describes (in which her own elbows were cut and grazed leaving scarring and her baby falls onto her head) would have been a most memorable incident in the life of this family and therefore it is highly unlikely that she would have simply forgotten about it. It only apparently happened 4 months before. I find it to be an unusual event for any mother to drop her baby daughter in such a way that she falls onto her head.

“53.

I am therefore driven to conclude that whatever it was that happened in July it was a comparatively minor and trivial incident and the mother was well aware that it was trivial … I can well understand how any parent faced with this situation would cast around for any explanation which might come up with theories which were unmeritorious, but the problem in this case is that the mother is putting forward a possible cause or explanation for [J]’s injuries which I find is inconsistent with her behaviour at the time and subsequently.

“55.

This inconsistency of behaviour together with the fact that inexplicably [J] suffers no visible injury at all, drives me to the conclusion that the mother exaggerated the incident […]

“56.

I am concerned that the parents and grandmother have not told me the whole truth about this incident and this in my judgment casts a shadow over their evidence.”

Later in setting out his conclusions, the judge said this of the incident, paragraph 69:

“[J] was an 11 month old baby who was unable to injure herself. I am satisfied that some event has happened to her about which I have not been told. I am satisfied that the parents have not told me the complete truth. In reaching that conclusion I am more concerned about the mother’s evidence and credibility than that of the father who on the whole I think was truthful, although the way they both remembered the July incident was not credible.

“70.

The impression I have is that the mother has used a minor incident to explain these serious injuries and the father has gone along with it. He was not present at the incident in July and therefore could not have known what had happened, but if he had been told in July when it is alleged that it happened that the mother accidentally dropped [J] in such a way that she had fallen on her head, then the likelihood is that she would have remembered this in November 2004 and said something.

“71.

By late December 2004 or early January 2005, the mother must have realised that if the wedding fall was discounted some other credible explanation was going to be required. I am driven to the conclusion on all the evidence that I have heard that she, probably knowing perfectly well what had happened to [J] (in the sense of one episode of shaking), decided to place reliance on an innocuous event that occurred 4 months before by embellishing it to give it authenticity. The difficulty is that for the reasons I have given the event lacks credibility; it simply will not bear the weight of forensic analysis.

“72.

I am satisfied on the balance of probabilities the reason why the mother has done this is to try and provide a credible explanation for [J]’s injuries and thereby exculpate both herself and the father. However despite this finding I have not found it possible on the evidence on a balance of probability to inculpate the mother.”

The finding, as set out in paragraph 53, is that the fall was comparatively minor and trivial, but the judge goes onto to say “the mother was well aware that it was trivial.” In paragraph 71, he finds the mother:

“… probably knowing perfectly well what had happened to [J] (in the sense of one episode of shaking), decided to place reliance on an innocuous event …”

Thus he was satisfied, as he said in paragraph 72, that the reason the mother had resorted to an exaggerated account of the July fall was to try and provide a credible explanation for J’s injuries and thereby exculpate both herself and the father.

Those findings amount to a finding that the mother knew that she had shaken the child and was lying to protect herself. Mr Braithwaite for the local authority concedes very fairly that he did not suggest to her that she had fabricated this evidence. He did not challenge her honesty; neither did the judge. It is, in my judgment, wrong that serious findings of this kind could be made without the witness having had the opportunity to deal with them. It is unfair and it amounts to a serious procedural irregularity, which would compel this court, in my judgment, to send the matter back for further hearing, so that the family have a fair opportunity to defend themselves.

But I turn to the second round of the appeal, the cogency of the evidence. The judge’s conclusions in this respect are set out in the following paragraphs of his judgment, under the heading “Conclusions”:

“62.

I was impressed by the compelling evidence of Dr Stoodley and Dr Johnson. Both are highly skilled experts and experienced in their field. Although I have been referred to certain perceived differences between them I have come to the clear conclusion that the substance of their evidence is clear, consistent and really both are saying much the same thing. There is moreover in my judgment an irresistible logic in their evidence. They are both clear in their reading of the scans that what they are looking at is chronic not acute blood. Both say this dates the haematomas as at least two to three weeks old but they could be older. This therefore makes the wedding fall as unlikely to be causative. Both also say (Dr Johnson perhaps expressing more doubt than Dr Stoodley) that the July incident was unlikely to be responsible. Had both been aware of my findings on the July incident, namely that the mother has exaggerated it, both would have probably excluded it altogether.

“63.

In any event Dr Johnson said that if there had been a bleed in July he would have expected it to have dispersed in 4 weeks, which of course is well before November. Too much can be made in my judgment of little differences between experts; is there much difference between ‘unlikely’, ‘very unlikely’, ‘rather unlikely’ and ‘extremely unlikely’? They are all, in my judgment, degrees of unlikelihood in which there is a sliding scale of confidence that an event is unlikely; but what the word unlikely means in my judgment is that on a balance of probabilities it is more likely than not that the event did not happen making it therefore improbable.

“64.

In any event I accept Mr Braithwaite's submission that the July incident was on all the evidence that I have heard, including the unsatisfactory evidence of the parents, an innocuous event which is much less serious than the mother maintained. This means in my judgment that it must render this incident as even more unlikely to have been the cause of subdural haematomas still visible on the scans in November, such that it can be excluded even as a possibility.

“65.

Neither doctor can be sure to the criminal standard that these subdural haematomas were caused non-accidentally, which is an entirely understandable position, but I am satisfied to the civil standard that that is what they are. In my judgment the evidence of the two radiologists is sufficiently compelling and cogent to prove on a balance of probability that [J]’s injuries were probably caused by an adult applying an excessive and unreasonable amount of force on one occasion, and that whilst this was sufficient to cause the injures they were at the lower end of the spectrum.”

So the judge’s stark conclusion, as he expressed it, was that J was probably injured whilst in the care of her parents.

Upon analysis it is obvious that there are three possible explanations for this child’s injury. The first is that the injury was sustained in the July fall, the second is that it resulted from the wedding fall and the third, which only arises if the other two could be discounted, is that it must have been caused by shaking. Each of those accounts needs careful analysis.

As for the July fall, the medical evidence excludes this principally because both Dr Stoodley and Dr Johnson would not expect the bleeding from that fall still to be present four months later. There was also doubt about the mechanism of the injury being severe enough to cause the harm in the first place. If the judge is correct that the mother’s account of the fall was exaggerated, then I would agree with his conclusion that it could well be excluded even as a possibility. As for the wedding fall, this undoubtedly did cause encephalopathy and there was clear evidence in the hospital notes of the history, with the dramatic picture painted by the father of the baby with her eyes rolling around in her head, of her irritability, of several bouts of vomiting and her general floppy condition. There is also some evidence of the retinal haemorrhaging being occasioned by this fall.

Mr Tolson took us to his cross-examination of the ophthalmologist, who is a very careful clinician and who would not be much moved to give any precise dating for the injury. This I would say encapsulates his evidence:

“I cannot tell you when. As I said, I cannot place a date for when these haemorrhages took place, so theoretically it could be related to a recent fall or it could be related to an episode that occurred prior to that.”

He felt it was inappropriate to try to time an event based on the appearance of a haemorrhage, because in his view, the time could last for weeks, even months. So at best this evidence is neutral.

As for the subdural haematoma, both experts were firm in their view that if one were viewing an injury three or four days old, then there would be a greater brightness in the blood than was present on the CT and MRI scans. In their view, the wedding was an unlikely cause, though a possible one.

Looking at shaking as a cause, Dr Stoodley was much firmer in his view that this was the likely explanation, Dr Johnson not being able to “say definitively whether it was accidental, or non accidental”. If the baby had been shaken two or three weeks before the scans, then that was possibly within the time frame covered by the retinal haemorrhaging, but of course shaking more than three weeks before the scan renders the retinal haemorrhaging less probative of such shaking.

If then one stands back and analyses the evidence so far, it is plain that the parents’ reliance on the July incident becomes very unconvincing if the judge was entitled to find that the mother had exaggerated her account of it. Mr Tolson attacks that finding as perverse. He draws attention to the account of it being confirmed by the grandmother. He draws attention to the dramatic way in which both parents recollect waking up at night and so forth.

But in my judgment Mr Tolson cannot sustain this challenge. The judge saw the witness and he was entitled to form a view about her, even if he did in the event go further then fairness demanded. He would have been entitled to find exaggeration, which puts the July fall at the speculative end of the possibilities in this case. Even if the wedding fall is not entirely convincing because of the old blood seen on the scans, it nonetheless seems to me that the real question is whether it was so unconvincing that shaking is the only conclusion to which one is properly driven. Here the wider canvas is very important. Although the judge did make his findings favourable to the parents early in the judgment, he did not refer to them again at all in his conclusions, and one is left most uncertain as to whether or not, and if so to what extent, he took that into account in reaching those conclusions.

The findings in favour of the parents are important. There is no history of bad parenting here at all. That does not exclude the possibility that the mother might, in a moment of weakness, have unintentionally shaken the baby hard enough to cause injury, but nothing suggests that it is probable. On the contrary, the parents in this case have revealed themselves to be totally protective of the children. They took J back to hospital on three occasions because of their concerns for her in November. They took J to hospital on 27 October when she was diagnosed to be suffering from no more than an upper respiratory tract infection, a common cold. They worried whether to take her to hospital or to the doctor after the July incident. They are a closely-knit family. If there had been shaking two or three weeks earlier, that would have produced some observable distress to the child, and it seems to be inconsistent with their pattern of conduct in taking the child for medical attention that they would have ignored her distress on that occasion.

The burden of proof is on the local authority, and with a serious allegation like this cogent evidence is required to overcome the unlikelihood of what is alleged and thus to prove it. When the family history is put in the balance, as it should be, then the court cannot in my judgment be satisfied that the November fall was so improbable a cause as to leave shaking the only acceptable explanation.

In my judgment, the local authority did not discharge the burden of proof in this case and the proper answer to the issue the judge was trying was to find that non-accidental injury of J was not established.

I would therefore allow the appeal and discharge the Supervision Order that had been made consequent upon it.

LORD JUSTICE LAWS: I agree that the appeal should be allowed for all the reasons given by my Lord, Lord Justice Ward.

LORD JUSTICE LONGMORE: There is, as it seems to me, an evidential void at the heart of the judgment in the present case. The judge started by outlining the wider canvas and concluded that it was favourable to the parents. He then examined the medical evidence. He summarised the evidence of the consultant neuroradiologist, Dr Stoodley, called by the local authority, by saying, “Since other causes could be excluded, J must have suffered a single episode of non-accidental injury, which was likely to have been caused by shaking and or impact.” It was unlikely that either of the known events in July or November 2004 could account for what the judge called “chronic blood”, viz. the chronic subdural haematomas which he had observed in the scans, particularly the MRI scan, in which he had more confidence than the CT scan.

The judge then summarised the evidence from Dr Johnson, the consultant paediatric radiologist, called by the children’s guardian but instructed by the mother, who says that although he found it difficult to analyse the MRI results, he agreed with Dr Stoodley that on the basis of the CT scan neither of the known incidents in July and November 2004 was a likely cause of a chronic subdural haematomas. So the likelihood of the present case being a case of non-accidental injury was ”significant”; it was more probable but not 100%.

In my view, that puts Dr Johnson’s evidence too high. What he actually said was, as cited at page 56 and confirmed at pages 63 to 64 of the transcripts that were before us:

“If we do not have any suitable history then they [by which I assume he means ‘we’] do become suspicious of non-accidental injury. But I do agree the presence of subdural haemorrhages is not an absolute sign of non- accidental injury and the appearance of the subdural haemorrhages could be the result of other factors.”

The judge did not evaluate this differing evidence of the doctors; indeed, he seems to have thought, see paragraph 62, that they were saying much the same thing. But in any area of medicine where, as this court said, when it gave permission to appeal in paragraph 9 of its judgment, based on R v Harris and Others [2005] EWCA Crim 1980, it is normal to rely on a triad of indicators, viz encephalopathy, retinal haemorrhaging and subdural haematomas, and only one such indicator was certainly present, the difference of emphasis between the doctors did in my view require to be evaluated, if Dr Stoodley was to be preferred to Dr Johnson.

The judge then proceeded to set out the reasons why he found the mother’s evidence unsatisfactory; particularly that she had exaggerated the July incident. He asked himself why the mother had done that and whether she was using the incident as an attempt to divert attention from what had really happened to J. He inferentially answered the latter question “yes” (although, on proper consideration, it was never part of the local authority’s case that the mother was being dishonest or seeking to deceive the court) because he concluded in paragraph 65 that the subdural haematomas were:

“… probably caused by an adult applying an excessive and unreasonable amount of force on one occasion, and that whilst this was sufficient to cause the injuries they were at the lower end of the spectrum.”

In paragraph 68:

“The stark conclusion I reach on all the evidence that I have heard, is that J was probably injured whilst in the care of her parents.”

The reason why I said initially that it is an evidential void, is that when the judge came to make his critical decision in this case, there is 1) no apparent appreciation of the not insignificant difference of emphasis between Dr Stoodley and Dr Johnson; 2) no apparent caution exercised, in the light of the fact that only one of the three indicators on which reliance is usually placed for a finding of non-accidental injury in this class of case was present; and 3) no evaluation, against the background of the medical evidence, of the favourable impression made by the parents as a family, not only on the witnesses such as Miss Bennett and the health visitor, Mr Jarrett, but also on the judge himself.

In these circumstances, while it is perfectly possible to acknowledge there may be suspicion, I cannot think it was right for the judge, with respect, to have found on the balance of probabilities that J was non-accidentally injured while in the control of her parents. The evidence does just not have that degree of cogency which is required to overcome the unlikelihood of the serious allegation that is being made, to adapt the words of Ungoed-Thomas J, approved by Lord Nicholls in Re H [1996] App Cases 563 at 586H. As Lord Nicholls also said, a judicial suspicion is not a proper factual basis on which the court’s jurisdiction can be exercised “… because that is no more than a judicial state of uncertainty about whether or not an event happened”; see 519E.

For these reasons and the reasons given by my Lord, Lord Justice Ward, I agree that this appeal should be allowed.