Risk assessments in care proceedings: L-G and Re T

In this blog, I will consider two significant 2025 decisions of the Court of Appeal in the cases of L-G and Re T, which revisit some of the most fundamental considerations surrounding the application of s31 of the Children Act 1989 to the facts of every given case. These cases both emphasised the need for a structured assessment of what constitutes ‘significant harm’ for the purposes of threshold determinations, both by reference to the wording of the Children Act 1989 and its interpretation, and by the application of the law on a case-by-case basis according to the individual facts.

s31 of the Children Act is the statutory footing upon which any given local authority issues proceedings for a care or supervision order. Under section 31(2), it is clear that the court may ONLY make a care order or supervision order if it is satisfied:
(a) that the child concerned is suffering or is likely to suffer significant harm; and
(b) that the harm or likelihood of harm is attributable to
(i) the care given to the child, or likely to be given to him if the order were not made, not being what it would be reasonable to expect a parent to give to him; or
(ii) the child being beyond parental control.

The two sub-limbs under subsection b very rarely exist in isolation from one another, with those cases of the child being beyond parental control often resulting ultimately from the standard of care they have received.

A big question that arises from the wording of s31(2) is what is ‘significant harm’? s31(2) defines harm as ‘the ill-treatment or the impairment of health or development including, for example, impairment suffered from seeing or hearing the ill-treatment of another’. This is a very expansive definition that includes harm to the child not just from the treatment that they themselves receive but also from observing ill-treatment of others, such as other family members, friends and other connected persons. The definitions of ‘development’ and ‘health’ are likewise expansively defined, with ‘development’ including physical, intellectual, emotional, social or behavioural development, ‘health’ meaning physical or mental health’, and ‘ill-treatment’ defined to include sexual abuse and forms of ill-treatment which are not physical.

By virtue of the deliberately expansive definition of all of the given terms under s31 the court has the power to undertake a broad assessment of what constitutes ‘significant harm’ in care proceedings and to adapt its interpretation of the wording of s31 on a case-by-case basis. The powers of the court in this regard have been reiterated in both the cases to be considered.

L-G (Children: Risk Assessment) [2025] EWCA Civ 60

This case concerned two young children, O (aged 5) and Y (aged 1). Both children had always lived with their mother, who had provided them with an excellent standard of care and with whom they were said to be thriving. O’s parents had separated when she was young, and she spent time with her father by agreement of both parents.

Y’s father had a significant criminal history and had 4 older children in respect of whom care proceedings had been issued necessitating the children’s removal under interim orders. In those proceedings significant findings of fact were made against Y’s father, including of assault and threatening three of the older children with a weapon, which their mother knew about and failed to prevent. Ultimately Y’s father was thereafter charged with offences arising from his treatment of the four older children.

The local authority in these proceedings issued care proceedings seeking a supervision order in respect of both O and Y. An interim supervision order was made and the mother gave undertakings that contact between Y and his father would be limited and supervised. A month prior to the final hearing the mother said that she had separated from Y’s father but the trial judge found that there was reason to believe that that the mother did not fully appreciate the risk that Y’s father posed to both of the children and considered it likely that had social services not been involved then the mother would have continued her relationship with Y’s father. In light of this, the judge made final orders for both children to be separated from their mother with O being moved into the care of her father and Y placed with his maternal grandmother under a special guardianship order. Orders were made for the children to have significant contact with their mother but the order for separation was made, effective immediately.

The mother appealed the court’s decision on a number of bases, namely that there had been:
(1) a failure to make a proper risk assessment;
(2) a failure to make key findings of fact; and
(3) that there had been a procedural error in relying on incorrect information.

The appeal was granted, largely on the basis of the first limb of the appeal: failure to undertake a proper risk assessment.

Lord Justice Peter Jackson giving judgment found that:
“the risk of harm, important as it is, is one of a number of factors in the welfare checklist and it has to be carefully assessed, where it may be decisive. In a case where the outcome was considered by the judge and all the professionals to be finely balanced, a solid risk assessment was critical. The respondents broadly argue that the judge had the advantage of assessing the mother in evidence and the judgment as a whole covers all the necessary ground. I cannot accept this. It is clear that that the judge thought carefully about her decision, but the conclusion that the mother does not share her assessment of Y’s father as a dangerous man could not on its own be determinative. That was particularly so as this was a difficult welfare outcome for these children, under which they are separated from each other but have high levels of contact with their mother under uncertain and possibly disadvantageous arrangements for supervision for the indefinite future”

The judge thereafter referenced and endorsed the prior authority in the case of Re F (A Child: Placement Order: Proportionality) [2018] EWCA Civ 2761 that in assessing a risk of future harm that the court had rightly asked itself a series of questions:
(1) What type of harm may arise?
(2) What is the likelihood of it arising?
(3) What would be the severity of the consequences for the child if it arose?
(4) Can the chances of harm be reduced or the consequences be mitigated?

The court found also that there was a need to tailor these broad questions to the facts of the individual case in L-G and that the court need to ask of itself a number of further questions:
(1) What were the settings and circumstances in which Y’s father would have had a realistic opportunity of harming the children?
(2) What was the likelihood of those setting and circumstances arising and remaining undetected?
(3) If the mother was incapable of being instinctively protective, could the likelihood or consequences of harm be reduced by protective measures such as obligations being imposed on the mother by the court, by the vigilance of O’s father or the children’s grandfather, or by virtue of the statutory powers of the local authority acting under a supervision order?

Lord Justice Jackson therefore ultimately allowed the appeal, concluding that while the trial judge’s finding that the mother would not be protective on her own could not be challenged on appeal, that the assessment of risk did not turn on her oral evidence alone and that the court had needed to go on to consider the likely real-world consequences of her lack of insight. The judge had failed to address why the lists of protective measures provided by both the local authority and the mother’s representatives would have been ineffective to protect the children.
The takeaway from the case of L-G is therefore that using the guidelines in Re F as a starting point, the court must tailor and expand on its questioning as to the assessment of risk according to the facts of the individual case.

Re T (Children: Risk Assessment) [2025] EWCA Civ 93

The decision in L-G was thereafter referred to, endorsed and expanded upon by the court in the decision of Re T (Children: Risk Assessment), emphasising the need for structured judicial reasoning, thorough risk assessment, and proportionality of decision making in care proceedings.

This case concerned three children, aged 11, 9 and 2. Findings had been made in previous care proceedings concerning the two older children dating back to 2016 that the children’s mother had caused a skull fracture to A2, then aged 15 weeks, albeit the court made no finding that the injury had been deliberately inflicted as opposed to accidental and there were no other matters relied upon by the local authority to satisfy the threshold criteria.

The two children were ultimately returned to the care of their father under a supervision order. It later transpired however, that the parents had never truly separated and had taken the two children to Africa where they lived with the parents and their maternal grandmother. This information ultimately became known to social services in 2023, whereupon further care proceedings were issued in relation to those two children and their younger sibling A3 born in March 2023. Care orders were made in respect of the older children and a placement order made in respect of the younger child, despite the two older children having expressed to the trial judge that they had desperately wanted to go home and the children’s guardian describing the children as being the most distressed children she had ever come across in foster care.

The parents appealed against the final decision, arguing that the trial judge had failed to identify the type or degree of risk to which each child would be exposed or to consider questions of risk management, or indeed to make a link between the threshold findings and the children’s welfare. It was noted that that while the independent social worker assessment had been concluded negatively in light of concerns surrounding the parents’ ability to adhere to a safety plan, that there were numerous positives associated with the parents, including observed quality family time with the children. Counsel for the parents noted that the trial judge had in her decision-making failed to address the welfare checklists under the Children Act 1989 and the Adoption and Children Act 2002 in respect of the younger child and therefore argued that the judge had failed to take account of important matters such as the father’s ability to manage and mitigate any risk, the impact on the children of separation from their parents and, in the case of the younger child, the difficulty in finding of finding an adoptive placement.

Lord Justice Peter Jackson again gave judgment in this matter, noting that:
“the separation of a child from a family can only be approved after a process of rigorous reasoning. That is essential where there is a plain of adoption, but it is also necessary for any significant decision where the outcome is not obvious. A structured process is of real benefit for these important and often difficult decisions, as without it there is a greater chance of error, leading to children living unsafely at home or being kept unnecessarily care.”

The court noted the court’s decision in L-G and its reiteration of the guidance in Re F and noted that “the risk of harm, important as it is, is one of a number of factors in the welfare checklist and it has to be carefully assessed, particularly where it may be decisive”. The court observed that the Children Act 1989 provides a framework within which the court can assess whether a child has suffered or is likely to suffer ‘significant harm’ for the purposes of the threshold for intervention, and ‘harm’ for the purposes of the welfare assessment, referring to the definition of harm pursuant to s31(9) of the Children Act addressed above.

Lord Justice Jackson thereafter stated at paragraph 33 of his judgment that the court had to address a number of questions in relation to each of these children:
(1) What type of harm has arisen and might arise?
(2) How likely is it to arise?
(3) What would be the consequences for the child if it did?
(4) To what extent might the risks be reduced or managed?
(5) What other welfare considerations have to be taken into account?
(6) in consequence, which of the realistic plans best promotes the child’s welfare?
(7) if the preferred plan involves interference with the Article 8 rights of the child or of others, is that necessary and proportionate.

Applying and considering these factors to the facts of the case, the Court of Appeal found that the trial judge had not undertaken a sufficiently robust analysis pursuant to the test under Re F, noting that it must be remembered that risk assessment is about the realistic assessment of risk, not about the elimination of all risk.

The Court of Appeal identified significant flaws including the lack of an effective assessment of the type or degree of risk to which each child would be exposed; the lack of a proper welfare evaluation by reference to both of the welfare checklists under the Children Act 1989 and the Adoption and Children Act 2002; a lack of a side-by-side comparison of alternative and competing options for the children’s care; and a lack of consideration of the proportionality of court’s final decisions and the orders made pursuant to the same. The Court of Appeal thereby allowed the appeal and set aside the care orders and placement order in respect of the children, with the applications for care and placements redetermined if pursued.

The decisions in L-G and Re-T together provide new guidance to local authorities as to the need for robust risk assessments and holistic considerations in the court’s final decision-making in care proceedings, with there being a clear identified need to conduct a proportionality analysis in all cases and to properly link the risk assessment to the welfare stage and welfare analysis. These decisions are likely to be significant with respect to welfare decision-making in future care proceedings.