This is an appeal by the mother of a little boy, C, now aged rising two, against findings made in care proceedings that she inflicted serious physical injuries on the child.

In a wide-ranging challenge, the mother raises eight grounds of appeal which can be distilled down into two overarching contentions – first, that the judge was wrong to conclude that the injuries were inflicted non-accidentally and, secondly, that if the injuries were inflicted she was wrong to identify the mother as the perpetrator.

C’s mother was born in Poland and moved to this country with her parents in 2008. In 2019, she started a relationship with the father and shortly afterwards became pregnant. The pregnancy was difficult as a result of various health problems suffered by the mother. On 17 June 2020, C was born in hospital with the assistance of forceps, which resulted in extensive bruising and a fractured clavicle. At a later date, the mother falsely claimed that it had been a breech birth and that the baby had been turned by hand before delivery.

It was the parents’ case that, in the next few weeks, various marks were seen on the child by several members of the family. During this period, C was seen by the midwife, health visitor and by a GP at his 6 week check. No bruises were seen by any of those medical professionals. Although there are records of conversations between the mother and the surgery about several health issues, there is no reference to bruising in the medical records. It was the mother’s case that she left a message about bruises she had seen with the health visitor but her call had not been returned.

At a routine immunisation appointment on 17 September 2020 when C was three months old, the nurse noticed a number of bruises on the baby’s back and torso, although there was no indication that he was in any pain. The mother told the nurse that he was bruising easily. The nurse did not refer him for an immediate examination by a doctor but instead arranged for a telephone consultation with a GP the following day. In the course of that conversation, the GP asked the mother to bring the baby in for a face-to-face examination. On the following Monday, 21 September, the mother brought the baby to the surgery and after the examination the GP referred him to hospital. On arrival there, a number of marks and bruises were observed on the child which, as he was at that stage non-ambulant, gave rise to a suspicion of non-accidental injury. A radiological examination revealed that C had suffered four fractures – a fracture to the 12^th rib, close to its articulation with the spine, and three metaphyseal fractures to the proximal end of the right tibia and the distal ends of both tibiae. A suspected buckle fracture of the left ulna was subsequently ruled out by further examination. The scans showed no evidence of any bone disorders and testing of his blood clotting system revealed no abnormality.

During his stay in hospital, C was cared for at times by his parents, either together or separately. At several points additional marks and bruises were observed on his body and face, although the medical records about these marks were incomplete and inconsistent. In particular, a number of body maps were drawn purporting to show where marks were seen, but most of them were unsigned and/or undated, and the location and descriptions of the marks were inconsistent.

On 21 October 2020, the local authority filed an application for a care order under s.31 of the Children Act 1989 and was granted an interim care order at a hearing two days later. On that day, C was discharged into the care of foster carers. Thereafter, he developed bruising on several further occasions, and was briefly readmitted to hospital again in November 2020, but there is no evidence that he has suffered any further fractures since being removed from his parents’ care.

Genetic testing carried out in Poland disclosed that C carried two variant gene sequences, one being a heterozygous variant within the TGFBR2 gene which was also found in C’s father. The variant is associated with a condition known as Loeys-Dietz Syndrome (“LDS”), a rare autosomal gene disorder, which causes a range of features including some connective tissue disorders similar to those found with Ehlers-Danlos Syndrome, a condition which has featured in a number of cases of suspected child abuse. Research studies have suggested that some patients with mutations in the TGFBR2 gene associated with forms of LDS may have a propensity to bruise more easily and may have greater skeletal fragility in childhood.

A fact-finding hearing took place over 17 days in November and December 2021. The local authority invited the court to make a series of findings set out in a detailed schedule but which were, in summary, that:

During the hearing the judge heard expert evidence from three medical specialists – Dr Oystein Olsen, a consultant paediatric radiologist, Dr Stephen Rose, a consultant paediatrician, and Dr Ian Ellis, a clinical geneticist. In addition, the judge heard from a number of treating clinicians and medical staff, including the midwife who was present at the delivery of the child, the GP who examined the child on 5 August 2020 and a number of members of staff at the hospital to which C had then been admitted. The judge also heard evidence from the parents and other members of the family. A statement was produced from the foster carer and C’s current GP who had examined the later bruises sustained while he was in care, although neither of those witnesses gave oral evidence.

In his reports, Dr Olsen advised that at the date of the first radiological examination on 23 September 2021 the X-rays and scans indicated that C’s rib fracture was one to two months old and the metaphyseal fractures two to six weeks old. The rib fracture had been caused by very forceful pressure on the spine and the metaphyseal fractures had been caused by a twisting/pulling or bending of the knees or ankles. Alternatively, all of the fractures could have occurred as a result of a violent shaking incident. None of them were caused by normal handing and there was no sign of any underlying bone disorder.

Unsurprisingly, as C had sustained a fractured clavicle during his forceps delivery, Dr Olsen was closely questioned in cross-examination about the possibility of the fractures having been sustained at birth. It was his evidence that “it would be very very surprising if they could be dated back to birth.” With regard to the rib fracture, he said that only direct pressure on C’s spine could have caused the injury to the 12^th rib, which is a so-called “floating” rib unattached to the sternum at the front and therefore not susceptible to compression during vaginal delivery. Furthermore, there was no known association between cephalic (i.e. head first) delivery and fractures to the head of a rib, in the absence of rotation during delivery, which the midwife confirmed had not occurred in this case. He also ruled out the possibility of the metaphyseal fractures occurring during the forceps delivery, since metaphyseal fractures are caused by the application of two forces which would not have occurred during the delivery.

Dr Rose’s opinion was that the clinical and radiological evidence pointed to C having suffered inflicted injury causing bruising and four fractures. He did not change his opinion after seeing evidence of bruising which C had sustained in foster care. In his oral evidence he said that there was no evidence from a clinical point of view that C had a propensity to bruise.

Dr Ellis prepared a series of reports prior to the hearing at which he gave oral evidence which has been transcribed for the purposes of this appeal. In his first report, he set out in some detail the current state of knowledge about gene variants of the type from which C and his father suffer, and about the features of LDS. He advised that, subject to further inquiry, the gene variant may be of no clinical significance, or may be causative for the LDS syndrome, or may have some partial effect in which there might be some increased genetic propensity to bone fractures or bruising. Dr Ellis’ ultimate conclusion was that the identified gene variant was of no causative significance in respect of either the bruises or the fractures. He gave a number of reasons for this. Although genetic testing had identified the variant in C and his father, there was insufficient clinical evidence to suggest that either of them was suffering from LDS. Neither of them exhibited clinical signs such as the facial features associated with the syndrome. Although the father had suffered fractures, it was Dr Ellis’ opinion that the history reflected his involvement with contact sport or risk-related activities. He concluded:

The evidence about the bruising and other marks seen on C before and during his hospital admission in September and October 2020 was given by a number of witnesses, including family members, doctors and nurses. The evidence of the professional witnesses was supported by and in some respects heavily dependent on medical records. Between paragraphs 70 and 146 of her judgment, the judge meticulously recited in chronological order all the evidence about bruises and marks, from both lay and professional witnesses. It is unnecessary to recite this chronology in detail in this judgment. The following points are of relevance to this appeal.

On 5 August 2020, the mother took C for his six-week check. According to the mother, she mentioned marks she had seen on C and the GP asked if anyone was hitting C. In evidence the GP denied that she had asked that question. Her detailed record of the appointment and examination contained no reference to bruising. The judge accepted the GP’s evidence and found that the mother had lied about what she said she had said and about what the doctor had said.

At the immunisation appointment on 17 September, the nurse examining C noticed marks on his body and told the mother to make an appointment with the GP. The nurse did not draw a body map but described C as being covered in blue-grey markings across the sides of his torso and back, with associated scratch and torso marks. According to the nurse, the mother said that C bruised easily and also (untruthfully) that she was separated from the father.

According to the GP who examined C on 21 September, the mother said that he had bruised easily since birth, but that the bruises were worsening, that she and her brother bruise easily and that the father was not living with them. The GP also did not draw a body map but described seeing multiple finger-shaped bruises on C’s back. At arrival at hospital, the mother said that C had bruised easily since he was six weeks old, and again said that she was separated from the father.

The judge was faced with considerable difficulties in respect of the evidence relating to C’s time in hospital. She found that there were a number of substantial problems with this evidence, in particular with the quality of the records kept by the hospital. A number of body maps had been drawn and filed in the records during C’s hospital admission, but many of them were neither signed nor dated nor attributed to anyone. Conflicting evidence was given by two doctors about one body map. Some other medical notes were unclear. Photographs were taken by a police scene of crimes officer on 29 September after more marks had been observed, but they were only produced during the hearing and a number of them were of poor quality. All this led the judge to observe in her judgment (at paragraph 140) that it was “hard to make a cogent picture from the hospital evidence and from the clinicians’ failure to record properly”.

On 3 November 2020, when C was in foster care, he was admitted to hospital again briefly after small bruises were observed on his back.

At the conclusion of the evidence, and the delivery of written and oral submissions, judgment was reserved. A draft judgment was circulated on 17 December 2021 and a perfected version formally handed down on 3 February 2022.

Having summarised the issues and identified the relevant legal principles by reference to an agreed note prepared by counsel, the judge set out the medical and lay evidence in some detail. Her summary of the medical evidence was accurate save that she wrongly said that LDS was not associated with a propensity to suffer bone fractures. In fact, as mentioned above, it was Dr Ellis’ evidence that research has suggested that LDS may be associated with a greater propensity to fractures in childhood.

The judge concluded that the mother had lied about a number of matters during the investigation and hearing, including (1) the details of the birth (falsely claiming that it had been a breech birth and that C had been turned manually before delivery), (2) in a misleading chronology prepared for the court, (3) the fact that the maternal grandmother had returned to this country in August 2020, (4) about her conversation with the GP on 5 August 2020, (5) whether she and the father had separated in September 2020, and (6) that her mother had blood and bone disorders which caused bruising. The judge observed (at paragraphs 156-7):

In contrast, she observed of the father that he

The judge set out her findings in these terms:

The judge noted that C had continued to suffer bruising in foster care, concluding that this suggested that he is a child who develops bruises or marks easily for some unexplained reason. But she concluded:

She concluded that the four fractures had been inflicted non-accidentally. She expressed her findings as to the bruising as follows:

The judge then set out her conclusions as to the perpetrator of the inflicted injuries:

The judge was asked to clarify the judgment in several respects. These included an inquiry as whether she wished to provide further definition of the bruising found to have been sustained through heavy handling in September 2020. The judge responded:

After handing down judgment, the judge made a series of case management directions in preparation for a final hearing, now listed in August, at which a decision will be made about C’s long-term future care. It should be noted that the judge rightly directed that a transcript of her judgment be disclosed to the hospital for the purposes of carrying out a review of its practices in the light of her observations about the poor quality of the record-keeping which had emerged in the proceedings.

On 25 March 2022, the mother filed a notice of appeal out of time against the findings. On 3 May, I granted an extension of time for filing the notice and permission to appeal. At the appeal hearing on 14 June, all parties were represented. The local authority and guardian opposed the appeal. It was the local authority’s position on appeal that the judge’s finding that the injuries were inflicted was correct and that her finding that the mother was the perpetrator “was within the judge’s discretion on the evidence”. By that latter submission, I understood its position to be that the finding as to perpetrator was one she was entitled to reach on the evidence. In the alternative, the local authority contended that, if this Court concluded that the judge should not have found that the mother was the perpetrator, it was not possible on the evidence to identify the perpetrator and this Court should therefore substitute a finding that each parent was a possible perpetrator and therefore remained within the “pool of perpetrators”. The father’s representatives initially indicated that he was “neutral” on the outcome of the appeal. When the local authority’s position on the appeal became known, this Court sought clarification of the father’s position in the light of the proposal that we could substitute a “uncertain perpetrator” finding. Following further emails, the father was represented at the appeal hearing by leading counsel (who had not appeared below) and junior counsel (who had). The father remained neutral and therefore made limited submissions. We are grateful to all counsel for their assistance, in particular to junior counsel for the father, Ms Hyde, who appeared pro bono.

The mother’s appeal notice advanced eight grounds of appeal:

At the hearing of the appeal, Mr Mark Twomey QC, who appeared for the appellant in the absence of trial leading counsel who was part-heard in another matter, identified four aspects to the appeal, namely the way the judge dealt with (1) bruising; (2) the genetic variant; (3) lies; and (4) if, contrary to the appellant’s principal case, the injuries were inflicted, the identification of the perpetrator. These aspects cover all the grounds of appeal and, whilst there is a degree of overlap between the four issues, it is convenient to consider the arguments under those four headings.

First, it was argued that the judge failed to deal adequately with the bruising (and also a cut lip) found on C’s face and body while he was in hospital. The fact these occurred during his admission to hospital in September to October 2020 was described by the mother’s representatives as “the crucial point in the case”. It was submitted that the judge’s failure to deal with this issue undermined her entire analysis. The bruises that emerged in hospital were crucial because they were relevant to both the local authority and the parents' cases. On the local authority’s case, these injuries had somehow been inflicted by the parents while C was in hospital. In particular, the injuries on 27 – 28 September had been inflicted by the father. On the parents’ case, none of the bruising that emerged in hospital could have been caused by the parents. The fact that he had developed bruising during this period conclusively demonstrated that C had a propensity to bruising for unexplained reasons. This in turn supported the argument that all of C’s injuries were attributable to the genetic variant or an unknown medical cause. The judge’s failure to deal with this aspect of the case was an important omission in her analysis and therefore undermined her conclusion that the bruising was inflicted by heavy handling in the parents’ care. Furthermore, by deciding to make no findings about the injuries sustained in hospital, she failed to take into account information which might have been material to identifying the perpetrator of such injuries as were found to be inflicted. There was no dispute that C sustained additional bruising in hospital yet the judge wrongly failed to reach any findings about it. Mr Twomey submitted that she had simply failed to engage with the issue.

With regard to the evidence about a genetic variant, it was submitted that the judge made a series of errors which led her to the wrong conclusion. First, it is said, correctly, that she wrongly said that LDS was not associated with a propensity to fracture. In fact, it was Dr Ellis’ evidence that there was research evidence which suggested just such an association. Secondly, it was said that the judge’s analysis of the genetic issue did not descend into the degree of detail required, which was set out at considerable length in Dr Ellis’ several reports. Thirdly, it was said that the judge overlooked the fact that there had been no scanning of the child in foster care, so there was no evidence either way about whether or not C had sustained further fractures after being removed from his parents’ care. Fourthly, the judge failed to attach sufficient weight to the possibility of an unknown cause. The specific variant under consideration had not been identified before this case. There was inevitably an element of uncertainty about its clinical significance which the judge underestimated. Finally, whilst the judge was entitled to accept Dr Ellis’ opinion, before doing do so she was obliged to assess it in the “broader canvas” of all the other evidence, in particular the evidence of the mother’s close relationship with the child and her otherwise exemplary care.

Next, it was argued that the judge’s treatment of the issues of lies was superficial and inadequate and in some instances wrong. Some assertions made by the mother were characterised as lies when in fact they were, or may have been, no more than inaccurate statements – for example, about the circumstances of C’s birth, or mistaken beliefs about family medical history. The judge failed to distinguish between inaccurate statements that were deliberate falsehoods and those that were simply mistakes. Furthermore, she attached weight to those statements that she found to be lies without carrying out any or any adequate analysis as required by case law, in particular R v Lucas [1981] QB 720 and Re A, B and C (Children) [2021] EWCA Civ 451. Although the judge said at five points in her judgment words to the effect that she was giving herself a Lucas warning in respect of a particular statement made by a witness (including the mother), this was done in formulaic terms without engaging with the question of how the mother’s lies should be taken into account in the overall assessment.

This leads to the final issue raised on behalf of the appellant. It is submitted that the judge’s treatment of the identification of the perpetrator was superficial. It was based principally on the judge’s impression that the mother was by far the dominant partner in the relationship. To base the identification of the perpetrator on such an impression was wholly unwarranted. Judges must be careful about relying on the appearance and manner of witnesses in court and here the judge attached too much weight to the mother’s demeanour. It is also argued that the judge’s failure to engage sufficiently with the issue of the bruising suffered by C in hospital meant that an important part of the evidence which was potentially relevant to the identification of the perpetrator of any injuries found to be inflicted was left out of consideration altogether. It followed that, even if the finding that the injuries were inflicted is upheld, this Court should set aside the finding that the mother was the perpetrator. In those circumstances, it was the position of the mother and father that there should be a rehearing of the whole fact-finding hearing. The local authority argued, as referred to above, that, if the Court was persuaded that the finding that the mother was the perpetrator should be set aside, it would be open to us to substitute a finding that the injuries were inflicted by either the mother or the father.

I do not accept the assertion on behalf of the appellant that the judge’s treatment of the bruises, and in particular the bruising sustained during C’s hospital admission in September to October 2020 was the crucial issue in the case. Of greater importance was the fact that, at a point when he was non-ambulant, C sustained four bony fractures of a type which were consistent with being inflicted for which there was no explanation. For that reason, the judge rightly started her analysis with consideration of the fractures and in particular the question whether they were attributable to the genetic variant.

It is correct that the judge wrongly said that there was no evidence of an association between LDS and a propensity to fractures. But that error had no bearing on the judge’s overall analysis. It was ultimately irrelevant that LDS has an association with a greater propensity to fractures because Dr Ellis’ opinion was that there was insufficient clinical evidence to suggest that either C or his father was suffering from LDS. After an extensive analysis, he had concluded that it was well outside a balance of probability for a gene variant to be responsible for bone fractures and bruising occurring in a cluster in early childhood without either affecting a transmitting parent or recurring in the child at a later date and that the variant identified in C was therefore a coincidental finding to the bone fractures and bruising identified in September 2020. This was clear and well-supported evidence which the judge was plainly obliged to take into account and to which she was fully entitled to attach decisive weight.

There is no substance in the other criticisms of the judge’s approach to the genetic evidence. In my judgment, her treatment of this aspect of the case is sufficiently detailed to explain her reasons for accepting it. There was no need for her to descend to the detail in Dr Ellis’ reports. There was no evidence of any subsequent fractures and the suggestion that the judge should have attached significant weight to the speculative possibility that C may have suffered later fractures that went undetected has no merit.

For my part, I see no reason to conclude that the judge failed to take into account the possibility of an unknown case. At the start of her judgment, she expressly said that

Similarly at the start of the judgment she warned herself that “the assessments of witnesses should not be compartmentalised” and that the court “must have regard to the totality of the evidence as a whole”. Bearing in mind these warnings, and looking at the judgment as a whole, I am also not persuaded that the judge looked at the different strands of the evidence in isolation.

For those reasons, I reject the submission that her treatment of the evidence of Dr Ellis was materially deficient. The judge’s analysis of the evidence about the causation of the fractures was in my view unimpeachable.

Similarly, I see no merit in the appellant’s criticisms of the judge’s approach to the issue of bruising. On this issue, the judge found (1) that C sustained bruises more easily than other children and also had a tendency to rashes, possibly eczema which complicated the picture, (2) that he sustained bruises in his parents’ care in the course of September 2020, in particular on his back, (3) that the observed and recorded bruising he sustained in his parents’ care was more severe than in foster care, (4) that this was suggestive of heavy handling, but (5) in finding that it was probable that bruising may have been caused by heavy handling, some allowance had to be made for the tendency to bruise. These subtle and considered conclusions are hardly indicative of a judge who failed to engage with the evidence.

Her conclusion that she was unable to make detailed findings about the bruising that occurred in hospital was plainly one that she was entitled to reach. Given the unsatisfactory state of the evidence about that admission, in particular the poor quality of the hospital records, she was not prepared to reach any conclusion about it beyond finding that it provided added support to the finding that C had a tendency to bruise more easily. There is to my mind no basis on which this Court could find that she was wrong in coming to this conclusion. As Lewison LJ observed in Fage UK Ltd v Chobani UL Ltd [2014] EWCA Civ 5 at paragraph 114,

A lie told by a witness in a family case may undermine the credibility of the witness and in some instances be direct evidence of culpability. The principle in Lucas is that a fact-finding tribunal must bear in mind that a person may lie for many reasons and the fact that they have lied about one or more things does not necessarily mean that they have lied about other things. When assessing the forensic significance of a lie or lies told by a witness, a judge must do more than merely cite the case of Lucas. He or she must consider the probative weight to be attached to the lies in the context of the totality of the evidence, bearing in mind the Lucas principle.

In this case, looking at her judgment as a whole, including the passages from paragraphs 156-7 and 179 quoted above, I am satisfied that the judge did properly engage with the issue by considering first whether the untrue statements were deliberate lies and secondly, in respect of those statements that she found to be lies, the reasons why the mother told them. It was the pattern of lying that she found particularly compelling, in particular her lies about the details of the birth and untruthfully telling the father that she had been seeking medical assistance for the bruising by speaking to the health visitor. It was this pattern which the judge took into account in reaching her conclusion that the mother was the perpetrator of the injuries.

In my view she was entitled to reach that conclusion. Her analysis of the perpetrator issue is succinct but sufficient. In summary, her reasons for identifying the mother as the perpetrator were as follows.

Although the judge’s analysis is less detailed than one might sometimes find, it provides a clear explanation of the reasons for her findings. At the heart of it is her impression of the parents’ characters as they emerged from the evidence, including their oral evidence. That is entirely within the province of the trial judge. In Re B-M (Children: Findings of Fact) [2021] EWCA 1371, Peter Jackson LJ observed (at paragraph 25):

Does the fact that the judge was unable to reach any conclusion about the cause of the bruising and marks identified in hospital mean that she was not in a position to reach a conclusion as to the perpetrator of the injuries that she did find were inflicted? In my view, the position the judge was in was no different from that in which judges frequently find themselves at the conclusion of a fact-finding hearing when they decide they are able to make findings about some issues but not others. The fact that this judge in this case was unable to reach findings about the cause of some of the injuries sustained by C did not mean that she was prevented from making findings about other injuries, including that some of them had been inflicted and that the perpetrator was the mother. The lacuna in the findings about the bruises which emerged on C’s body after he was admitted to hospital was not so substantial as to undermine the reliability of the comprehensive evidence about the injuries sustained before the admission.

For these reasons I would dismiss this appeal

I agree.

I also agree.