S (Children: Findings of Fact), Re

This appeal concerns findings of fact in care proceedings.

The advantages possessed by a judge making findings of fact after hearing evidence are well understood. This court will not intervene unless there has been some clearly demonstrated error of approach.

When a court is considering a body of evidence in order to reach findings of fact it must take into account all the evidence and consider each piece of evidence in the context of the other evidence when reaching overall conclusions.

These principles are so well known that they do not require the citation of authority.

In the present case, the court was faced with evidence of a sequence of injuries to a small child over time, some minor in themselves, some serious. It found that the child's father had injured the child on one occasion and that he and the mother had covered it up. It also found that other injuries were the result of accidents and that some were in any case insufficiently serious to amount to significant harm: these conclusions are challenged by the local authority in an appeal supported by the Children's Guardian.

I have reached the reluctant conclusion that the fact-finding process in this case went wrong. The judge made findings against the parents in one respect concerning their treatment of the child and their credibility, but he did not take account of the implications of those findings when considering the other allegations of mistreatment.

At the end of the appeal hearing we informed the parties that the appeal would be allowed and that the judge’s findings would be set aside. As the matter must be remitted for rehearing, I will give my reasons for supporting this outcome as briefly as possible.

This is the second set of care proceedings concerning two boys, J (5) and B (3). The first set took place in 2020/2021. In February 2020, B was admitted to hospital aged five weeks, when he was found to have several skull fractures. The local authority issued care proceedings and the children were placed with their grandparents under an interim care order. The proceedings came before His Honour Judge Wicks, who conducted a fact-finding hearing that lasted for 10 days in September/October 2020 and February 2021, with judgment being given on 18 March 2021. The judge dismissed the proceedings, finding that the local authority had not established that the injuries were inflicted by either parent, and that the most likely cause was an unfortunate accident, unwitnessed by either parent, involving J jumping onto a baby bouncer in which B was lying, forcing it onto the floor and bringing B’s head into sharp contact with the floor. None of the experts thought this probable, but they could not exclude it as a possibility. The judge also found that, although there had been arguments between the parents and some domestic abuse from the father towards the mother in the past, by the time B suffered his injury, their relationship was on a stable footing. His order included these recitals:

“AND UPON the court finding that the most likely explanation for the injuries to B is that J jumped from the sofa onto the baby bouncer in which B was laying bringing his head into contact with the floor;

AND UPON the court finding that there is no blame attached to the parents for the injuries suffered by B;

AND UPON the court recording that the children’s medical and Local Authority records should be updated to reflect the absence of any wrongdoing on the part of the respondent parents so to avoid the suggestion that any ‘Child Protection markers’ exist in respect of the subject children.”

The children returned to the parents’ care. The local authority attempted to work with the parents under a child in need plan, but the parents refused and the case was closed at the end of March 2021.

The present care proceedings began just five months later. On 16 August 2021, B was brought to hospital with soft tissue injuries to his head and face. He was found to have a large and bulging subgaleal haematoma over the left parietal and left temporal area, bruising on and behind the left ear, bruising on his left temple and right parietal area, a bruise and swelling on the forehead between his eyebrows, bruising around and below his right eye and bruising on his left cheek. No injuries were seen on J. The parents were arrested and the children were taken into police protection. In due course, interim care orders were made, under which the children have for the most part lived with their maternal grandparents.

Examination of the mother’s phone revealed a large number of photographs taken between the middle of April and August 2021 and showing B with numerous facial injuries. These included photographs taken on 21 and 22 July 2021 that showed very severe facial bruising.

On 14 July 2021, the mother had cancelled a home visit by the health visitor, saying that the family was unwell.

A paediatric opinion was obtained from Dr Cartlidge in December 2021. He described the list of bruising to B as extraordinary for a child of that age. He advised that further evidence was required to exclude excessive bruising caused by Ehlers Danlos syndrome or a clotting disorder. If these were excluded, he considered that at least some of the frequent and extensive bruising sustained between April 2021 and August 2021 was most likely caused non-accidentally.

On 12 April 2022, a hearing took place at which the judge allowed the local authority’s application to reopen his findings of fact concerning the 2020 skull fractures and directed that the causation of those injuries should be redetermined alongside the investigation of the 2021 bruising.

Dr Saggar, a clinical geneticist, provided a report in August 2022. He found no clinically significant gene mutations and excluded Ehlers Danlos syndrome, but he found evidence that both children had inherited a connective tissue disorder, hypermobile spectrum disorder (“HSD”), from their father. This conclusion did not alter Dr Cartlidge’s opinion.

The local authority sought findings that the parents were responsible for the 2020 skull fractures and the 2021 bruising. It alleged that by May 2021 the parents were struggling to cope with B and that the father was using cannabis regularly. The parents denied mistreating B and said that his condition when admitted to hospital in August 2021 was the result of him walking into a stair gate on one occasion and falling on a laminated floor on another.

The second fact-finding hearing began on 31 August 2022. Evidence was given by eleven witnesses. After five days the hearing had to be adjourned during the mother’s evidence after she became too distressed to continue. She completed her evidence on 11 January 2023, with counsel for the Guardian finishing his cross-examination through written questions. Submissions were delivered in writing.

Judgment was given on 8 June 2023 and a final order was made on 26 June 2023. The judge made these findings of fact:

“94.

I set out in summary my findings:

a.

B suffers from a hypermobility spectrum disorder (HSD) which makes him clumsier and more susceptible to bruising.

b.

Between April and June 2021, whilst in the care of his parents, B suffered a number of bruises and cuts. None of these were caused by M or F, whether deliberately or carelessly. The supervision of the children may not always have been adequate. However, none of the injuries were sufficiently serious to amount to significant harm.

c.

In July 2021, B suffered significant bruising and swelling to the left side of his head and face. These injuries were caused by a blow or slap with an open hand, administered by F, after he lost his temper with B.

d.

M was aware of what had happened to B. She cancelled a visit by the health visitor and refused to allow even a doorstep visit, so that the health visitor would not see the injuries to B. The reason she gave for the cancellation, namely that the family were suffering from sickness and diarrhoea, was untrue. M has, to that extent, failed to protect B.

e.

B suffered further significant injuries to his head and face in August 2021. These were caused accidentally by (i) B colliding with the stairgate, on or about 9 August 2021; (ii) B slipping and falling on a laminate floor and hitting his head, a day or two before his admission to hospital on 16 August 2021.

f.

B’s head injury in 2020 was caused accidentally, for the reasons set out in the judgment delivered in the 2020 proceedings.

95.

On the basis of these findings, I conclude that the threshold for making orders under section 31 of the 1989 Act is crossed.”

In his judgment, the judge summarised the evidence and the law, and then set out his analysis. He was critical of how the local authority had pleaded its case. He considered B’s HSD. He assessed the evidence of the parents as well as their reliability and character. In particular, he said this:

“72.

In my judgment in the 2020 proceedings, I set out my impression of both parents at that time. I stand by that impression; no party has invited me to revisit it, in the light of the evidence I heard in the current proceedings. Indeed, neither parent was cross-examined on the events which were the subject of the 2020 proceedings.”

He then turned to consider the various injuries and pleaded allegations. His essential reasoning was as follows:

“84.

The parents do not dispute that B has suffered a considerable number of bruises and cuts, mainly to his head and face, during the period between April and August 2021. It is impossible to determine, on the evidence, when each of these bruises or cuts occurred; nor is it possible to determine accurately the number of bruises. What can be said is that B sustained more bruises than might ordinarily be expected for a child of his age, and this is what particularly struck Dr Cartlidge. B and J are both active children. I accept, on the evidence, that B’s hypermobility syndrome makes him clumsier and more prone to bruising. For the reasons I have already given, I am loath to draw an adverse conclusion from a simple comparison between the bruising documented between April and August 2021 and any bruising sustained after August 2021, when B was in protective care. Throughout this period (save for a few days in July 2021, which I shall come to) there have been regular visits to the house by MGM and by professionals, such as the health visitor. Many of the bruises described as having been first noted in the period between April and June 2021 are, in themselves, relatively minor. None excited the suspicions or concerns of professionals. There were no safeguarding referrals made during this period. I am unable to conclude that these injuries of themselves amounted to significant harm, even if attributable to inadequate supervision. I keep in mind the observation of Hedley J in Re L (above) that a threshold finding must be ‘at least something more than commonplace human failure or inadequacy’. Whilst I accept that frequency of bruising might be an indicator of significant harm caused by a caregiver, it does not follow that frequency, of itself, is probative of significant harm caused by a caregiver.

85.

I accept Dr Cartlidge’s view that the most significant injuries are those sustained by B in July 2021 (the extensive bruising to his head and face) and August 2021 (the head injuries that led to B’s admission to hospital). I am entirely satisfied, having seen the photographs of the injuries, and having read the descriptions of them, that these do amount to significant harm. Furthermore, it is inherently unlikely that either of these parents repeatedly beat B over a period of three to fourth months, to cause the thirty or so injuries set out in the schedule of allegations. Such a sustained period of abuse would be likely, in my view, to have made B fearful of the parent responsible, and yet all the evidence points to B having a good relationship with both parents. Nothing of concern has been noted by contact supervisors: no apprehension, or fearfulness, on B’s part, towards either M or F. Similarly, no such behaviour was noted by the health visitor.

86.

The parents say that the July 2021 bruising was caused when B fell out of his bed. However, on the parents’ own evidence, B’s bed had been dismantled by this time, following an earlier fall from the bed in April 2021, and B was sleeping on a mattress on the floor. Further, the injuries are not, in Dr Cartlidge’s opinion, consistent with such a fall; rather, the injuries are consistent with something which has wrapped itself around the face.

87.

The health visitor did not see this injury at the time, as her planned visit was cancelled by the parents, who said that the whole family was suffering from a sickness and diarrhoea bug. Had she seen it, she would have made a safeguarding referral.

88.

I am satisfied that (i) this injury was caused by a slap or blow to the face with an open hand; (ii) it was inflicted by F; and (iii) M was aware that it had been inflicted by F. It is likely that F lost his temper with B and that on this occasion he lashed out. F is right-handed; the injuries are to the left side of the face, consistent with a blow from a right-handed person. A blow from an open hand would, to adopt Dr Cartlidge’s description of the likely mechanism, wrap itself around B’s face, causing the bruising and swelling so graphically illustrated in the photographs. M must have been aware that this is how B sustained this injury; there is no other explanation for her decision to postpone the health visitor’s planned visit and to refuse to allow even a doorstep visit. Further, I find that the reason given by the parents for this cancellation – the stomach bug – was untrue. It is unlikely that either parent was in fact suffering from such a bug, given that, at the time when they were supposed to be ill, they were texting each other about having sexual intercourse and about ordering takeaway food. This was the only time that M cancelled a visit from a professional, and I am satisfied that it was because she wanted to conceal B’s injury.

89.

As to the August 2021 injuries, I am satisfied on balance that these were caused when B collided with the stairgate on or around 9 August, and when B slipped and fell on the laminate flooring of the living room at the family home a day or two before his admission to hospital. The experts accept that these mechanisms are possible – although not, in their view, probable – causes of the soft tissue head injury and associated bruising. There was some focus on the bruising around the ear, one of the so-called ‘protected areas’ which would not generally be injured in an accidental fall. However, Dr Cartlidge accepted that such bruising was possible if, for example, B had struck the side of his head on a toy as he landed on the floor. The contrast between this and the July injury, in terms of likely causative mechanisms, is striking. There is a plausible explanation for how the July injury could have been inflicted, namely a slap or blow; there is no such corresponding explanation for how the August injury could have been inflicted. I refer to what I said in my judgment in the 2020 proceedings: for the LA to prove its case, it must persuade me that, in some unknown and undescribed way, either M or F caused a soft tissue injury to B’s head. A slap administered in anger could not, as I understand the medical evidence, account for such an injury.

90.

M made no attempt to conceal the August 2021 injury. I accept that she had raised concerns in the past about the safety of the laminate flooring, and had asked that the landlord fit a carpet. This fits with her decision – which I accept with hindsight was questionable – to take B to an appointment with the letting agent, presumably to reinforce her request for a carpet. However, B had clearly suffered a significant fall on his head, and another blow to his head a week earlier when he ran into the stairgate. M’s focus should have been on securing prompt medical attention. However, there is no evidence that B’s health was compromised by any delay in taking him to hospital, and the LA does not invite me in terms to find that the parents delayed unreasonably in seeking medical attention.

91.

The LA invites me to conclude, in the light of all the evidence, that the head injury sustained by B in 2020 was attributable to either M or F. The LA’s case, in essence, is that if I conclude that one or more of the injuries suffered by B in 2021 was inflicted either deliberately or carelessly by either M or F, I should go on to conclude that the injury suffered in 2020 was similarly inflicted. However, this is precisely the sort of fallacious reasoning that Theis J warned against in Surrey CC v E (above). There is no new evidence about the state of the parents’ relationship in 2020 which would lead to me reconsider what I said about the relationship then. The LA is still unable to describe how precisely either M or F inflicted this head injury, or to account for the absence of other injuries (e.g., rib fractures) which might be more strongly indicative of abuse. Having considered all the evidence that was before me in 2020, alongside all the evidence in these proceedings, I reach the same conclusion, namely that the LA has not persuaded me that the injury was inflicted either deliberately or carelessly by either M or F; rather, it was an unfortunate and unforeseeable accident.”

The Guardian and the local authority sought clarification of the judge’s reasoning at paras. 88, 89 and 91, and also asked why the finding against the father was not relevant when reconsidering the 2020 head injury. As to that, the judge responded:

“I was not specifically addressed on the issue of propensity on behalf of the LA or the Guardian, although the relevant authorities are cited in [counsel]’s submissions on behalf of M. I suspect this is because the one proven event of inflicted injury (the July 2021 injury) cannot on any view amount to a propensity. Thus, in so far as it is now submitted that I should conclude from my finding about the July 2021 injury that F had a propensity to inflict injuries on B, I reject that submission.”

Permission to appeal was granted by King LJ on 17 August 2023.

The local authority challenges the judge’s exculpatory conclusions in relation to the 2020 head injuries and the 2021 bruising. The grounds of appeal are:

1.

In relation to the 2020 injury, the judge having directed the reopening of the factual enquiry, the judge:

a.

incorrectly treated his decision as a review of the original material;

b.

wrongly relied on the fact that there was no appeal against the original findings when that was irrelevant as the judge had re-opened the enquiry into that issue.

c.

Failed to consider the injury in the context of his finding that the father injured B in July 2021 by slapping him and that the parents lied

i.

in relation to the causation of that event.

ii.

the father’s cannabis use.

iii.

the tensions in the parties’ relationship including in respect of the parenting of the children.

d.

Wrongly placed the burden on the local authority of disproving a theoretical, speculative unwitnessed event (that J jumped onto his brother).

e.

Misdirected himself in relation to:

i.

the issue of propensity and

ii.

the relevance of hindsight or outcome bias.

2.

In relation to the bruising the judge was wrong:

a.

not to consider the overall patterns of bruising (described by the independent expert paediatrician as an extra-ordinary extent for a child aged between 15.5 and 19.5 months) before, during and after the period when B was in the care of the parents between March and August 2021, instead focusing on each bruise as an individual matter;

b.

to disregard the evidence of the social worker that B was reluctant to be with his parents at the beginning of some contacts.

3.

In relation to the injury to the head shown on presentation to hospital on 16 August 2021

a.

Failed to consider the injury in the context of his finding that the father injured B in July 2021 by slapping him and that the parents lied in relation to the causation of that event.

b.

Wrongly placed the burden on the local authority of disproving a theoretical, speculative unwitnessed event (namely that there must have been a toy on the floor which B fell onto).

c.

Did not consider the relevance of the evidence of both parents and others that B (a child they said marked and bruised easily) displayed no signs of injury when he was checked by them both immediately after the reported fall on 15 August 2021 and later that same day at bed-time.

d.

Misdirected himself in relation to:

i.

the issue of propensity and

ii.

the relevance of hindsight or outcome bias.

For the local authority, Mr Twomey KC and Mr O’Brien referred to Re CTD (A Child: Rehearing) [2020] EWCA Civ 1316, [2020] 4 WLR 140, at [8]-[13], which affirms that the court’s task at a rehearing is to conduct a fresh investigation into the whole of the evidence. Here, the judge had lost sight of the reasons why he had ordered a rehearing concerning the 2020 injuries and his very limited reference to them (found only in paragraph 91) shows that he did not engage with the local authority’s case. The lack of an appeal from the earlier findings was irrelevant. The judge then misdirected himself by considering propensity. The proper approach is that a finding that injury A was caused by a parent is relevant when assessing the probability that injuries B or C were caused by that parent, though it is not determinative. In the present case, the judge did not take proper account of the fact that the assault in 2021 reduced the improbability of an assault having occurred in 2020. Instead, he directed himself that it was not determinative and seems to have treated it thereafter as immaterial. The judge also misdirected himself as to hindsight bias by interpreting the need for caution about hindsight as a ban on using later events to shed light on earlier ones. As to the bruising between April and July 2021, the judge failed to consider the significance of the overall pattern in the light of Dr Cartlidge’s evidence and he overvalued the evidence about HSD. He did not explain why what was in effect constant bruising for six months did not amount to significant harm attributable to inadequate parenting, whether through active harm or a failure to supervise. The same problems beset the judge’s reasoning in regard to the August 2021 injuries. He did not take proper account of Dr Cartlidge’s evidence and relied on speculation that B must have landed on a toy left on the floor, even though the mother, who was present, did not describe this. The local authority was left with the burden of disproving a speculative account of an undescribed event. The errors in self-direction concerning propensity and hindsight bias were carried through into this area too.

For the Guardian, Mr Watson supported these submissions.

For the mother, Mr Bagchi KC and Ms Turner submitted that the judge’s findings in relation to the 2020 skull fractures were correct or, at least, open to him. Critically, he assessed the parents as not being of the character to inflict injuries on a baby. The baby bouncer theory was unlikely but not impossible. The successful application to reopen the findings was not based on new evidence as there was none. The only variable from the first proceedings was the character and credibility assessment of the parents. The judge’s task was to assess whether his evaluation of the parents’ character and credibility had altered in a way that justified a different conclusion on the 2020 fractures. His clarification responses show that he did not consider the finding that the father had inflicted the July 2021 injury of sufficient evidential weight to render him the probable cause of the 2020 fractures; nor did it amount to propensity. The judge gave reasons for his conclusions about the 2021 bruising. As to the final injuries, it is important that none of the experts suggested that they could have been caused by a slap; conversely, they said that a fall onto a laminated floor could account for the subgaleal haematoma. While there was no positive evidence about a toy or similar protruding object on the floor to cause the ear injury, the judge was entitled to look at the whole of the evidence and to decide on the balance of probabilities that the injury was not inflicted. His conclusions about this injury are unassailable, partly because a key plank in his reasoning was made up of his assessment of the parents’ credibility. Overall, the judgment is sufficient for the local authority to know why it had not persuaded the judge to its view and the high hurdle for disturbing findings of fact is not cleared.

For the father, Mr Garrido KC and Ms Collinson emphasised the unique position of a judge who had seen so much of this family over such a long period. The judge was entitled to find that there was no, or no sufficient, link between two different injuries with different causal mechanisms occurring in different family circumstances, 17 months apart. As to the later bruising, he did not treat each bruise in a separate compartment, but simply preferred the evidence that the bruises were relatively minor, and that some were found on parts of the body susceptible to accidental injury in a boy with HSD. He distinguished between these more superficial injuries and the more significant injuries suffered in July and August 2021. It is not perverse to find that a child has suffered both inflicted and accidental injuries. As to the final injuries, he was entitled to accept the parents’ accounts, which were not inconsistent with the expert evidence. These submissions were strongly enhanced by oral argument from Mr Garrido, who persuasively identified ways in which it might be possible for this court to uphold the judge’s decision.