This is an appeal in public law care proceedings against findings of fact made by HHJ Weston KC (‘the judge’) on 26 November 2023. The judge found that fractures to both the tibias of a little girl W, then 10 months old, had been inflicted either deliberately or recklessly by either the Appellant mother (‘the mother’) or the Respondent father (‘the father’). That being the case, the judge held that the threshold criteria for the making of orders under section 31(2) of the Children Act 1989 (‘CA 1989’) were satisfied.

In my judgment the judge fell into a number of errors of principle and the appeal must be allowed. This Court has a full picture of the background and surrounding circumstances. Other than the findings of inflicted injury, there is no other basis upon which the threshold criteria can be satisfied; accordingly, the matter will not be remitted for a retrial. In order to avoid delay while the judgment was finalised the parties were informed of the Court’s decision last month, and following the submission of a draft by them an order allowing the appeal was made on 10 April. The following are my reasons for allowing the appeal.

This approach accords with the view of the Guardian who, at first instance, expressed the view that absent findings in relation to the tibial fractures sustained by W, there were no remaining circumstances which would support a finding that the threshold criteria were satisfied and that the proceedings would therefore come to an end.

Following communication of the Court’s decision a careful process of rehabilitation of W to her parents has now commenced.

The mother and the father have been in a long-standing, stable relationship for over a decade. The father has worked nights for many years and the mother is a teaching assistant as is her own mother, the maternal grandmother (‘the grandmother’).

W was a much wanted baby who was born prematurely at 33 weeks towards the end of November 2020. By reason of her prematurity, W remained on the neonatal intensive care unit for a period of a little over three weeks and for the first four days of her life was in receipt of Total Parenteral Nutrition (‘TPN’), a process of administering highly specialised forms of food intravenously. W was discharged back home from the hospital on 21 December 2020 after which she continued to experience gastroesophageal reflux symptoms in the form of persistent vomiting and problems with feeding. She was initially prescribed Gaviscon, but from 11 February, until the end of September 2021, the consultant paediatrician prescribed Omeprazole, a proton pump inhibitor (‘PPI’) used in the treatment of indigestion and acid reflux.

In due course, the mother returned to her part time work as a teaching assistant. Arrangements were made whereby W was taken to the home of the grandmother and maternal step-grandfather (‘step-grandfather’) three mornings each week at about 8.00 am. Shortly after W’s arrival, the grandmother herself would leave to go to work leaving W in the capable care of her step-grandfather. At this time W was crawling and pulling herself up (although not standing unsupported), she could be seen on video footage taken in August 2021, bouncing vigorously in her ‘Jumperoo’.

On 30 September 2021, the step-grandfather moved a few steps away from the sofa where he was changing W in order to dispose of a dirty nappy, when W rolled from the sofa and landed awkwardly. W was distressed and her leg was clearly hurt.

The family acted entirely appropriately. The step-grandfather rang W’s parents and within an hour W was taken to the Walsall Manor Hospital. Following an X-ray, W was found to have sustained a fracture to her right femur.

The medical professionals, both at the hospital and subsequently instructed as independent experts in these proceedings, accepted that, whilst unusual for a baby to sustain such an injury from a low-level fall, the femoral fracture was an accidental injury. The local authority did not seek findings in respect of this fracture and accept the family’s account that this was caused by the accidental fall from the sofa.

It follows therefore that an important part of the history is the fact that W sustained an accidental fracture of her femur at ten months old.

Whilst at hospital, a skeletal survey was carried out which revealed that W had, in addition to her femur fracture, fractures to her left and right tibias which radiologically were shown to be a little older than the femur fracture. Neither the parents nor the maternal grandparents (“the grandparents”) were able to give any history or explanation as to how these fractures had occurred. This inevitably raised concerns about the possibility of inflicted injury. The local authority issued care proceedings and W was made the subject of an interim care order on 8 October 2021. W was placed into foster care where she remains to date.

On 22 November 2021, within the care proceedings, the court directed the instruction of a consultant paediatric radiologist, Dr Olsen, and a consultant paediatrician, Dr Shenoy. Each expert filed a report in April 2022.

Since that time, parental assessments have been carried out. The quality of the contact between W and her parents is exceptional as is the dedication shown by them throughout the time W has been in care. The court was informed that there is no question of W being placed for adoption. The alternatives being considered prior to this appeal having been allowed are rehabilitation of W to her parents, placement with a family member who had been identified or, if all else had failed, the unusual course of the making of a special guardianship order in favour of the current foster carers in order to ensure that W would not lose her relationship with her parents.

Section 32(1)(a) CA 1989 provides that when an application is made in care proceedings, the court must draw up a timetable with a view to disposing of the application without delay and “in any event within 26 weeks beginning with the day which the application was issued”. It follows that care proceedings having been issued on 7 October 2021 these proceedings should, under Section 32 (1)(a), have been concluded around the week ending 8 February 2022 when W would have been approximately 16 months old. I am conscious that post pandemic, the 26-week requirement has, to many over-worked courts and judges and hard-pressed local authorities, presented a simply unachievable target. In this case, there have also been case management challenges resulting from the complex medical issues which have arisen. That being said however, it cannot be acceptable that care proceedings having been issued in respect of a ten month old baby of otherwise exemplary parents, by the time the finding of fact hearing took place, W was no longer the baby who was received into care, but a toddler of rising three.

At the conclusion of the hearing on 28 July 2023, it was agreed that written submissions would be filed within seven days. The judge indicated to the parties that judgment would be delayed by “a few weeks” due to a period of annual leave and a medical issue with the judge’s hand. Unfortunately, there was a further significant delay of four months before judgment was handed down. The judge gave advance notice of his decision on 6 October 2023 to the effect that he intended to find the fractures were inflicted, but the judgment itself was not then delivered until 27 November 2023, seven weeks later. The judge rightly expressed his apologies for the delay.

Permission to appeal the judge’s order was granted by Baker LJ on 25 January 2024 and listed as a matter of urgency allowing the appeal to be heard on 28 February 2024. The result of this overall delay is that W had been away from her parents for 2 years and 5 months when the appeal was heard.

The trial took place over three days. The judge heard evidence from a pharmacologist Dr Sharp, Dr Shenoy, both of the parents, the grandmother and step-grandfather. The local authority submitted that the two tibial fractures were inflicted injuries and that each of the four family members was a potential perpetrator. The local authority’s case therefore was that the judge should make a finding that this was an uncertain perpetrator case and that each of the four adults should remain in the pool of perpetrators.

The judge identified what he regarded as the significant issue at para.[52] as follows:

“the significant issue that I need to consider and arrive at conclusions on arise sic from whether [W] had bone fragility arising from her prematurity, (including low birth weight), her extended period in NICU, her TPN feeding and problems with feeding on discharge from hospital.”

The judge went on at para.[54] to say that another issue he was going to have to consider was W’s “likely reaction to having sustained tibial fractures.”.

As already indicated, and at paras.[125] and [127], the judge exonerated the grandparents and found that the tibial fractures had been inflicted by either the mother or the father.

It is useful to set out the grounds of appeal upon which permission has been granted at this stage before moving on to analyse the approach of the judge which had led to him making his findings.

The grounds of appeal can be briefly summarised:

Mr Bowe KC on behalf of the mother, supported by Mr Sampson KC on behalf of the father, accepted without reservation the challenge an appellant faces in seeking to undermine a finding of fact made by a judge at the conclusion of a trial. He referred the court to Re S (Children: Findings of Fact) [2023] EWCA Civ 1113 that:

“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”.

I remind myself also of the judgment of Lewison LJ in Volpi v Volpi [2022] EWCA Civ 464 at para. [2].

It is submitted on behalf of the parents, and accepted by the Court, that the judge did make an error of principle in that the grounds of appeal together demonstrate that whilst the judge referred in his judgment to his having considered “the totality of the evidence”, having reminded himself at para.[76] that medical evidence, “even if compelling, is but one part of the evidential jigsaw”, in reality the evidence was considered by him in its component parts without putting together and considering the whole evidential picture. Had he done so, Mr Bowe submits, it would have been clear that the evidence did not establish on the balance of probabilities that these were inflicted injuries.

The principal evidential features considered further later in the judgment included the following:

At no stage in the judgment was this whole evidential picture put together. As will be explained, the judge reached his finding that the tibia fractures were inflicted injuries based largely on his conclusion that W did not suffer from bone fragility and on the failure of the parents to give any sort of history as to how the fractures occurred against the backdrop of what he found to be W’s likely reaction to the fractures. Only having made the finding that the fractures had been inflicted by one of her parents, did the judge move on, for the purpose of identifying the likely perpetrator, to consider the parents’ evidence, their credibility and whether they were likely to have caused the injuries.

Dr Olsen (the paediatric radiologist) who was not required to give oral evidence, made the following key findings:

Dr Olsen regarded it as “particularly important” in W’s case “to explore the possibility of bone weakness” as although there was no specific sign of bone fragility in the radiological examinations “the leg fractures per se in a child who has recently started weight-bearing raises significant suspicion (that the bones were insufficient).”

Dr Olsen explained that W’s injury pattern was quite unusual in that the fractures seen in the tibias were all commonly caused by axial loading i.e. weight bearing. Fractures of this type seen in tibial bones are often, he said, called “trampoline fractures” because of the typical mode of causation in older children. It was for this reason that Dr Olsen thought it important to explore the extent to which W had been weight bearing by standing whether supported or not.

So far as Dr Shenoy (the paediatrician) was concerned, he too was of the view that the femoral fracture may have been caused by the fall from the sofa and said in his report that:

Dr Shenoy explained in his written report that premature infants who spend a prolonged period in the neonatal unit do have bone fragility. Following the fractures, Dr Shenoy would have expected W to be in significant pain and distress and any change in clothing, including a change of nappy would lead to discomfort and pain. In his oral evidence, he elaborated, saying that not every fracture results in swelling or heat, that each child’s pain response is different, and that W’s response may have been different over time.

The parents subsequently provided photos and videos of W weight bearing and jumping on her Jumperoo. Dr Shenoy in his oral evidence agreed that W bouncing in her Jumperoo would be sufficient to cause the tibial fractures (described by Dr Olsen as trampoline fractures) if there was bone fragility.

In the light of both experts regarding it as important to consider whether W had an increased risk of fracture, the court on 27 May 2022 directed the instruction of an expert pharmacologist, Dr Stephanie Sharp, and a clinical geneticist, Dr Ellis. Dr Ellis reported that W had no genetic predisposition to bone fractures. The views and evidence of Dr Sharp require more detailed consideration and examination, particularly as they were dependent in large part on research material which did not directly relate to a baby in W’s position.

Dr Sharp is a forensic pharmacologist. In her brief report she explained that Omeprazole is a PPI indicated for, amongst other things, gastrointestinal reflux disease. From a pharmacological point of view, she explained, the rates of absorption of Omeprazole are erratic and do not follow pharmacokinetic principles with studies showing a marked individual variability in both the rate and extent of absorption.

Infants, Dr Sharp explained, are prone to gastro-oesophageal reflux and as a result PPIs are commonly used in neonates and infants. PPIs are associated with an increased risk of facture in adults but the information is not completely clear on the likelihood of effect on bones of infants.

Dr Sharp’s evidence relied substantially upon a paper: Early Acid Suppression Therapy Exposure and Fracture in Young Children by Laura Malchodi et al (‘the Malchodi paper’). This paper looked at a cohort of 851,631 children. Only children without known serious medical conditions were included in the study. It was common ground that W would have been excluded from the study as: she was premature, had been in the neonatal unit for three weeks, was of a much lower birth rate than the babies in the study, had been prescribed Omeprazole for far longer than the children in the study and had suffered a fracture before one year of age. Her prematurity of itself gave rise to an increased risk of bone fragility.

The Malchodi paper’s results did not establish a causal relationship between PPI exposure in respect of the healthy children in the control, but did show a positive association between PPI and childhood fracture incidence. Longer duration of the treatment and earlier age of first exposure were associated with increased fracture hazard. The median prescription length in the study was 60 days with initiation of PPI between 0–6 months which was associated with a 23% increased fracture hazard with the median first fracture age being 3.9 years. The study said that the use of [PPIs] “should be weighed carefully against possible fracture”.

The Malchodi paper concludes:

“AST used in the first year of life, especially PPIs, are associated with increased fracture hazard in children. Results should not be interpreted to suggest that PPIs… alone explain fractures, which is important in suspected cases of nonaccidental trauma. Results indicate longer AST use and earlier initiation may increase fracture hazard. Practitioners should be aware of the potential for fracture when considering treatment with AST verses lifestyle changes and watchful waiting. If AST use in necessary, providers should limit prescriptions to a single drug and limit their duration when possible.”

At the date of her admission to care, when W was aged 10 months, she had been prescribed Omeprazole for seven months.

Dr Sharp referred to two additional reviews, one in 2018 (Dermyshi et al) and one in 2021 (Binti Abdul Hamid et al), which concluded that there is insufficient evidence as to the side effects of Omeprazole in preterm infants and that caution should be used when prescribing such medication.

Dr Sharp excluded from consideration a more recent paper: The Clinical Characteristics of Fractures in Paediatric Patients Exposed to Proton Pump Inhibitors by Fleishman et al (‘the Fleishman paper’) on the basis that it related to a small cohort of only 32,001 children. Whilst one respects that expert view, it is of note that the Fleishman paper which was included in the bundle, highlights the lack of research in this area, but, as in the Malchodi paper, notes the “significantly higher rate of fractures among paediatric patients exposed to PPI”. The Fleishman paper in addition found that in those patients, the location of the fractures is statistically different from those patients without exposure, with the PPI cohort more likely to suffer from lower extremity fractures.

Dr Sharp concluded that there is a small risk of all fractures over a period of time in children treated with PPIs such as Omeprazole and that the risk may be increased with increased duration of treatment. The Malchodi paper, she said, indicated the median age for fracture to be 3.9 years, which would suggest that any risk of fracture would occur in an older child and not imminently in a baby in receipt of the medication.

In her oral evidence, Dr Sharp accepted that the Malchodi paper shows that there is a 23% increase in fracture risk when a child is given PPI in the first six months of life and that the children included in the study were at a lower risk of fracture than W.

Whilst she did not discount the possibility that Omeprazole might have been a contributory factor in relation to W’s fracture risk, and she accepted that W having had Omeprazole for over seven months placed W in a higher risk category than the children in the study, Dr Sharp maintained her opinion that on the balance of probabilities Omeprazole was not causative of the fractures by reason of bone fragility in W’s case. The judge said that “Dr Sharp repeatedly pointed to the fact that the fractures in the Malchodi study were over the full term of the study so not necessarily when the child was receiving omeprazole or a PPI”.

Dr Shenoy in his oral evidence, perhaps unsurprisingly, told the court that care has to be taken in applying the research literature to W’s circumstances.

The judge began his discussion and analysis at para.[72] by noting that in this case there was not a “raft of ‘risk factors’ or ‘red flags’”. The judge said that the “wide canvas” included 10 matters which were set out at para.[73]. These included the parent’s exemplary previous history with no problems with drink, drugs or domestic violence, their commitment to W and their positive engagement with the local authority and that there are none of the risk factors identified by Peter Jackson J (as he then was) in Re BR (Proof of Facts) [2015] EWFC 41 (“Re BR”). At para.[74], the judge recorded that he had in mind that W had sustained an accidental femoral fracture whilst in the care of the step-grandfather who acted appropriately when the accident occurred.

The judge went on to consider at para.[79] “bone fragility”, one of the two issues he had identified as significant (see para.[19] above). The “potential strands” he identified were (i) her prematurity and low birth weight; (ii) feeding issues and (iii) her having been prescribed Omeprazole. It should be born in mind that both Dr Olsen and Dr Shenoy had each independently flagged up the possibility of bone fragility in this baby given her prematurity and the administration of Omeprazole.

The judge considered each aspect in relation to bone fragility separately as a ring-fenced issue. In relation to bone fragility consequent upon prematurity, whilst recognising that a long period of time on a neonatal unit and prolonged periods of TPN do increase the risk of bone fragility, Dr Shenoy had not considered her prematurity was likely in her case to have caused bone fragility but accepted that she was “more vulnerable than a heathy child to fracturing”. The judge was therefore satisfied that, in isolation, W’s prematurity had not caused her to have bone fragility.

So far as feeding was concerned, the judge held that even though W was still being prescribed Omeprazole, as she still had episodes of posturing relating to feeds, she was on the whole a well-baby, gaining weight and was not failing to thrive. Her feeding issues therefore did not contribute to bone fragility.

Dr Shenoy had indeed said that by the time of her injuries, W was a healthy child. The Guardian had put it that “beyond her prematurity and reflux, the court should start from the position that by the time the fractures occurred she was in essence a healthy child”. In my view that rather begs the question, W was undoubtedly making excellent progress, but that does not assist with a determination as to what hidden impact there may have been by way of future risk of fractures presented by the combination of W’s prematurity and the extended period of administering of Omeprazole upon what was by now an otherwise healthy baby?

Having closed off those two avenues the judge turned to Omeprazole. The judge at para.[102] set out twelve extracts from Dr Sharp’s evidence, a list taken substantially from the Local Authority’s closing written submissions. He said that Dr Sharp had made a number of important points in her evidence. In particular, he physically underlined in his judgment that the data referred to was for “all fractures over the term of the whole study”. He then quoted Dr Sharp as saying that: “the fracture does not have to have occurred whilst the child was receiving PPI, but over the course of the study, that is the chance of fracture occurring”. The judge further included in the list that the 23% increased likelihood of risk of fracture to W was over the period up to when W was five years old and it did not mean that the 23% chance was purely for the period when Omeprazole was being administered.

The judge included in his list that, although no study was directly comparable to W, Dr Sharp was able to draw conclusions by way of a “meta-analysis”. I feel a measure of unease at this being regarded as a basis for concluding that, whilst the Malchodi paper would not have applied to W, conclusions can nevertheless be drawn by way of “meta-analysis”. Meta-analysis is a statistical process that combines the data of multiple studies to find common results and to identify overall trends. Whilst a true meta-analysis provides the ability to extrapolate to the broader population, the research and evidence of Dr Sharp, relying as it did almost exclusively on the Malchodi paper, does not appear to me to provide a quantitative summary of the findings of multiple studies investigating a similar phenomenon which would form the basis of a meta-analysis.

With respect to the judge, it would have been of assistance if there had been some analysis or summary in which this miscellany of points had been put together and an overall view taken of Dr Sharp’s evidence against the backdrop of the other evidence. Instead, the judge said that he was satisfied that Dr Sharp’s opinion that Omeprazole did not cause or contribute to W sustaining fractures was “well made out on the totality of the evidence”. He said:

“107.

For all the reasons identified, in terms of answering of answering the overall question as to whether W did have bone fragility, my clear conclusion, having had the benefit of having heard the evidence (both lay and expert) is that she did not have bone fragility.

108.

I am satisfied on the totality of the evidence that notwithstanding her difficult start in life, including a premature birth, a short period of TPN feeding and 22 days in the NICU and a lengthy period of Omeprazole use, W was a generally healthy child. I do not identify any significant feedings issues. She appears to be a child who was gaining weight, developing normally and there is no evidence that she was a child who was failing to thrive.

109.

Accordingly, I do not consider that her prematurity, her health and feeding and her Omeprazole use, when considered individually and collectively was such as to cause or contribute to her fractures by reason of bone fragility. I also note, in the context of Dr Sharp’s evidence that any increased risk of fracture was not at the time of taking the Ofmeprazole but in the later years and over the period up to when W was five years old, that since W’s removal from parental care in October 2021 that she had not sustained any further fractures.

110.

I am very clearly of the view that W did not have any bone fragility at the time when she sustained her fractures. I have had well in mind the well established principles that medical science is constantly evolving and that in the context of allegations of inflicted injuries that there are many examples of the medical science evolving and medical experts now accepting what in the past they firmly denied (e.g. that subdural bleeding might plausibly be caused as a result of childbirth). However, on the present medical evidence and knowledge I am satisfied that W did not have bone fragility. In arriving at that conclusion I am not ignoring the potential for this being an outlier case and for unknown causes of bone fragility and/or fractures to exist.”

In my judgment the judge fell into error in analysing the medical evidence in closed off compartments and as a result, as is asserted in Grounds 1, 2 and 4 of the Grounds of Appeal, he failed to consider the relevance of the accidental femur fracture and the cumulative impact of the various possible causes of bone fragility. As a consequence, the judge in his analysis set out above did not factor in that, whilst the prematurity would not in itself have caused bone fragility, that feature should not have been discounted and looked at in a vacuum. This was a premature baby with increased vulnerability to fracture, who had been prescribed Omeprazole for many months and had already sustained an accidental fracture. These features should have then been put into the equation together with the significant association of increased risk of fractures seen in children who, unlike W, had not been born prematurely and who had been prescribed Omeprazole for a significantly shorter period of time than W.

With respect to the judge, his findings turned substantially on the evidence of the forensic pharmacologist who was saying that, based on the limited and somewhat contradictory research, she had concluded on the balance of probabilities that the Omeprazole had not contributed to the bone fractures with heavy emphasis being laid on the fact that W was so young when she sustained the fractures, but without the benefit of any research covering a cohort of babies who had been born prematurely and consequently had an increased risk of bone fragility.

The judge’s acceptance of Dr Sharp’s central reliance on the fact that the healthy babies in the Malchodi paper sustained their fractures at a median age of 3.9 years singularly failed to take into account the fact that this baby had had what was accepted to be a highly unusual accidental fracture to her femur at 10 months of age when she was still being prescribed Omeprazole.

Ground 3 of the Grounds of Appeal submitted that the judge placed excessive weight on the evidence regarding pain response and had reversed the burden of proof onto the parents.

Dr Shenoy gave evidence as to the likely presentation of W after she had sustained the tibial fractures. He said that he would invariably have expected W to demonstrate pain at the time and to have shown some distress thereafter. There may or may not, he said, have been heat and swelling as fractures present differently. No one child, he told the court, is the same as another and pain tolerances vary. A clinically silent fracture was a possibility.

The judge said that whilst Dr Shenoy had entertained the possibility of the tibial fractures being clinically silent he had not said that it was probable and “on [his] assessment of the totality of the evidence” the judge did not consider a silent fracture to have been probable and held that W would have reacted, at least in the very early stages after sustaining the tibial fractures, as expected by Dr Shenoy and in a similar way to that demonstrated when she sustained the femoral fracture.

It would, the judge said, have been reasonable to expect the carers to be able to give a history.

In Re BR, in relation to pain reaction, Peter Jackson J said at para.[15] that:

“…It would of course be wrong to apply a hard and fast rule that the carer of a young child who suffers an injury must invariably be able to explain when and how it happened if they are to be found to be responsible for it that would indeed to be to reverse the burden of proof. However, if the judge’s observations are understood to mean that account should not be taken, to whatever extent is appropriate in the individual case, of the lack of history of injury from the carer of a young child then I respectfully consider that they go too far.”

It follows therefore that whilst the judge was perfectly entitled to take into account the lack of history “to whatever extent is appropriate”, against the backdrop that a carer will not invariably be able to explain an injury. The failure however to provide an explanation must be considered against the backdrop of all the evidence including an assessment of the lay evidence.

Unfortunately the judge then moved on, without any reference to the evidence given by the parents or grandparents, to make his findings of inflicted injury. At para.[120], the judge said that he set against the medical evidence the fact of the “very positive” wider canvas of evidence and the absence of risk factors. He concluded at para.[121] that he was satisfied that “having regard to the totality of the evidence”, that the local authority had discharged its “burden in proving that the tibial fractures were inflicted injuries”. It follows that Ground 3 is made out in so far as it asserts that excessive weight was placed by the judge on the issue of W’s pain response.

Having found the threshold criteria to have been met, the judge moved to a second section of the judgment entitled “Identification of perpetrator”.

Ground 5 of the Grounds of Appeal goes to the weight the judge placed on the wider evidential canvas and in particular the absence of risk factors surrounding the parents and wider family.

The judge early in his judgment had, over a number of pages, set out the proper legal approach to be adopted in a finding of fact hearing. In particular he referred, by reference to Re BR, that whilst the medical evidence is important and must be carefully assessed, it is not the only evidence. The judge said at para.[41] that “evidence of parents and carers is of utmost importance. It is essential for the Court to form an assessment of the credibility and reliability and the Court is likely to place significant weight on the evidence and impression it forms: Re W and another (non-accidental injury) [2003] FCR 346”.

Peter Jackson J had said at para.[7] in Re BR that the court’s evaluation had to take account of the fact that unlikely events occur all the time although the probability of them arising in any individual case is extremely low. In the present case an unlikely event had happened, W had sustained an accidental fracture of the femur from a low-level fall.

Peter Jackson LJ returned to this theme in the very recent case of Re R (Children: Findings of Fact) [2024] EWCA Civ 153 which was not before the judge. In that case the medical evidence was that the baby had sustained a head injury which had all the features of the well-known triad of injuries. The medical evidence was that the injury could have originated from a low level fall, but was more probably attributable to a single shaking event. It was submitted on appeal that the medical evidence left open the possibility of accidental causation and that the judgment was so flawed in respect of the assessment of credibility and probability as to be invalid.  Peter Jackson LJ in allowing the appeal said at para.[34] that “medical and non-medical evidence are both vital contributors in their own ways to these decisions and neither of them has precedence over the other.”

Under his heading of “The Evidence”, the judge said that he had re-read the papers and that “it is impossible to mention all the evidence” and so he focussed on the issues he regarded as necessary to determine the disputed issues. As noted above, the two issues identified in the judgment by the judge were whether W suffered from bone fragility and W’s likely reaction to having sustained the tibial fractures. The judge highlighted that he had “specifically considered the medical evidence” but “had in mind that it is one part of the evidential jigsaw and is not in itself determinative”. Finally, regarding his approach to the evidence, the judge said at para.[71] that he would comment in brief terms later in the judgment his “impression of the lay parties and their respective evidence”.

Unfortunately, despite having reminded himself that the evidence of the parents was of the utmost importance, nowhere in the judgment does one find any summary or understanding of the evidence given by the parents. All that is to be found is a general observation that there was no evidence from anyone who had cared for W of an occasion when she cried out or was distressed, other than when she sustained the femur fracture and also that the father had accepted that W would have been distressed.

The judge concluded his analysis in relation to W’s likely presentation upon sustaining the tibia fractures by saying at para.[120] that he had well in mind the “very positive wider canvas of evidence and the absence of the type of risk factors as identified in Re BR”.

It is trite law that this Court will only rarely go behind a finding of fact made by the trial judge. In the present case, however, in relation to his finding as to the pain W would have suffered when she sustained the tibial fractures and the finding of inflicted injury, the judge in my judgment gave precedence to the medical evidence which so far as the finding of inflicted injury was concerned was itself largely based on research which was not directly applicable to this child.

Nowhere, before going on to make the critical finding that the tibial fractures were inflicted injuries, did the judge factor in the fact that W had sustained an accidental femur fracture at ten months old when a cornerstone of the evidence of Dr Sharp and the judge’s findings was that any fracture would have occurred when the child was significantly older. The judge failed to put that fact together with the undoubted association of an increased risk of fracture in children who were not born prematurely and had been prescribed Omeprazole for a limited period. The judge’s only analysis of the significance of the femoral fracture was in respect of W’s likely pain reaction.

One of the difficulties with the judge’s conclusion at this stage of the judgment (which in this, a single issue case, marks the finding of the threshold criteria) is that there is no understanding on the part of the reader as to the content of the evidence of any of the four principal protagonists and no attempt has been made to put their evidence into context. It is of course the case that no judge can be expected to set out all the evidence; however, simply referring to a “very positive wider canvas” or “the totality of the evidence” is not in my view enough. There is no assessment of the credibility of the parents notwithstanding that in the judge’s lengthy citation of authority he included a reference to Re W and another (Non-accidental injury) [2023] FCR 346 which said at para.[41] that the evidence of the parents and carers is of the utmost importance and that it is essential for the Court to form an assessment of their credibility and reliability.

Ms Meyer KC on behalf of the local authority said that “the judgment could have been structured better” and accepted that the judge had not addressed the issue of the credibility of the parents but, she submitted, the failure to set out any detail or summary of the oral evidence of the parents or grandparents should not undermine his ultimate conclusion that these were inflicted injuries.

This was not, however, simply a matter of structure. It is not possible in this case to conclude that on a reading of the judgment overall the judge had conducted a full and sufficient analysis. Even when the judge went on to consider the issue of perpetration, he did not set out, even in summary form, the evidence of the family and there is no consideration of the parent’s credibility.

When using the word “credibility”, I have firmly in mind the recent authorities in relation to credibility in particular B-M (Children: Findings of Fact) [2021] EWCA Civ 1371 where Peter Jackson LJ said:

“25.

No judge would consider it proper to reach a conclusion about a witness’s credibility based solely on the way that he or she gives evidence, at least in any normal circumstances.

The ordinary process of reasoning will draw the judge to consider a number of other matters, such as the consistency of the account with known facts, with previous accounts given by the witness, with other evidence, and with the overall probabilities.

However, in a case where the facts are not likely to be primarily found in contemporaneous documents the assessment of credibility can quite properly include the impression made upon the court by the witness, with due allowance being made for the pressures that may arise from the process of giving evidence.

Indeed in family cases, where the question is not only ‘what happened in the past?’ but also ‘what may happen in the future?’, a witness’s demeanour may offer important information to the court about what sort of a person the witness truly is, and consequently whether an account of past events or future intentions is likely to be reliable.”

More recently in Cazalet v Abu-Zalaf [2023] EWCA Civ 1065, the Court said at para.[62] that the judge’s assessment of the parties’ credit was “an important feature which should have fed into the judge’s determination alongside objective findings of fact.”

The judge made no detailed reference either to the evidence of the parents, which had apparently been given for half a day each, or to that given by the grandparents who routinely cared for W three mornings a week during term time. Other than the fact that there were no “red flags”, there is no consideration of the dynamics of this family or any picture of their day-to-day life or routines. Only when the finding of inflicted injury had been made, and the judge had turned his attention to identification of the perpetrator, does one have any limited understanding of the content of the evidence or the impression made by the family on the judge.

When one does consider the judge’s findings as to the credibility of the grandparents, it becomes immediately apparent that those findings should not only have had a place in the judge’s consideration as to the identification of any likely perpetrator, but also as to W’s likely presentation following the tibial fractures. It follows that their evidence should have formed part of the overall analysis prior to making the finding of inflicted injury.

The judge was of the view that both grandparents were generally doing their best to assist the court. So far as the grandmother was concerned, the judge rejected the local authority case that the grandparents should be in the pool of possible perpetrators. The judge dismissed, at para.[124], their submission that she was disingenuous or lacked credibility and he did not accept that she had not been “forthright with the court”. Further, the judge did not accept a submission that any inconsistencies in her evidence rendered her evidence unreliable. The judge said that whilst there was an element of “rose tinted spectacle” in relation to her account of family life and a desire to “reiterate the positives” there was no real possibility of the grandmother having inflicted the tibial injuries.

So far as the step-grandfather is concerned, these prolonged proceedings have had a devastating impact upon his health, none of which is recorded in the judgment. The judge did, however, state at para.[126] that he went into respiratory shock whilst giving his evidence and suffered a cardiac arrest between hearings. The judge believed him when he said that he had not caused the injuries and noted that his actions when W sustained the femoral fracture had been wholly appropriate.

Whilst there is no summary in the judgment of the factual evidence either the grandparents or parents gave, the Court has the benefit of the closing submissions of Ms Whitworth, counsel for W, which contains an excellent summary of the evidence given by each of the witnesses. These summaries are accepted by all parties to be an accurate record of the evidence.

The summary shows that each of the mother, father and step-grandfather changed W’s nappy during the fracture window and that whilst each had spent significant amounts of time with her, their evidence was that none of them had seen signs of distress or discomfort. If follows that the judge, having concluded that the evidence of the grandparents was reliable, had accepted, not only that they had not inflicted an injury on W, but also that they had not seen her exhibiting distress or discomfort when they were caring for her during the fracture window.

The failure to give a history on the part of the parents was regarded by the judge as highly significant and clearly weighed heavily in the balance leading to his conclusion at para.[132] that the parents “failure to give any history consistent with pain and distress in the aftermath of sustaining the fractures” and that “one or both of the parents is not telling the truth and doing their best to assist the court.”

Because the judge did not consider the evidence and credibility of the grandparents and their account of W’s presentation during the fracture window until after making his finding that there had been an inflicted injury, this important piece of evidence (namely that the grandparents gave honest evidence and that they had not seen signs of distress in W in the relevant period) was not put into the evidential equation when considering the significance of the parents’ inability either to give a history of how W came to have her injuries, or to note any specific distress or discomfort in the ensuing period. As a consequence, the judge did not have in mind evidence which should have served to remind him that there is no hard and fast rule that the carer of a young child who suffers an injury must invariably be able to explain when and how it happened.

The judge’s analysis of the parents’ presentation was even more condensed. The judge noted that the father worked nights at which time the mother was the primary carer, but that when not working he was fully involved. So far as the mother was concerned, the judge described her as a “highly anxious mother” but that her “mental health in terms of outward appearances appears to have been well managed”. (It should be noted that earlier in his judgment at para.[73] the judge had said at (d) that “There is no evidence of the mother’s mental health being unstable”.) The judge went on to speculate as to possible strains on the parents without reference to any evidence given by the parents or by any other witness to that effect. He said:

“130.

The mother and father each did not identify any issues with caring for [W]. I do wonder whether caring for [W] was not as easy as they are saying it was. The mother was undertaking a substantial amount of child care whilst at the same time having to manage her own anxiousness and her PTSD. The father was working nights, and which this was an established routine for him, I do wonder whether he was overly tired as a result of that and his understandable desire to be fully involved with caring for [W]. I do consider that each of the parents may be minimising the impact that caring for E was having on them both.

131.

In any event, the portrayal by the family of things being perfect, cannot be accurate given that [W] has sustained tibial fractures that I have found to be inflicted injuries caused by unreasonable care.”

This speculative explanation as to the possible strain the parents were under, and which he said may have led to one or other of them inflicting the injuries, was not only not put to the parents, but was in my view (notwithstanding that it is not a concrete finding) contrary to the guidance in Re G and B (Fact-Finding) [2009] EWCA Civ 10 at para.[16], where it is said that a judge should not “go “off-piste”, and […] make findings of fact which are not sought by the local authority”, without ensuring that such findings are “securely founded in the evidence” and that “the fairness of the fact finding process is not compromised”.

The judge in his extensive quotation from Peter Jackson J’s judgment in Re BR included this passage:

“8.

Each piece of evidence must be considered in the context of the whole. The medical evidence is important, and the court must assess it carefully, but it is not the only evidence. The evidence of the parents is of the utmost importance and the court must form a clear view of their reliability and credibility.”

This was undoubtedly a difficult and complex case which would have been challenging for any judge to navigate. The judge, focused as he was on the evidence of Dr Sharp and his understandable concern at the seemingly lack of any history or description of pain or distress exhibited by W following the tibial fracture, reached the conclusion that these were inflicted injuries. I am satisfied that a proper analysis of all the features of the case could only have led to the conclusion that the entirety of the evidence demonstrated that the burden on the local authority to satisfy the court that inflicted injury was more likely than not to be the cause of the tibial fractures had not been discharged. In those circumstances the Court concluded that it was neither necessary nor appropriate to remit the case for a retrial.

It follows that the threshold criteria under section 31(2) of the CA 1989 were not satisfied, it not having been established that W suffered significant harm attributable to the care given to her by either her mother or her father such care not being what it would be reasonable for them or either of them to have given to her. The interim care order has accordingly been discharged and the proceedings discontinued.

I agree.

I also agree.