England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
FAMILY DIVISION
(Mr Justice Charles)
Neutral Citation [2003] EWCA Civ 839

		Royal Courts of Justice Strand, London, WC2A 2LL
		1st July 2003

Dame Elizabeth Butler-Sloss P. :

1. This appeal again raises the issue of the correct approach of the court to the attempts to identify the possible perpetrator(s) of non-accidental injuries to a baby or young child when, as so often is the case, there is no direct evidence of the infliction of the injuries. In the present case a baby boy, M, was 11 weeks old when he was taken to hospital on the 15^th September 2002 suffering from serious injuries. The view of the hospital was that the injuries were non-accidental and the local authority issued care proceedings. The respondents to the care proceedings were the parents, M (through his guardian) and, joined shortly before the hearing, Miss S, the night nanny, who was on duty during part of the period when the child was injured. The maternal grandmother was not a party. The proceedings, the subject of the appeal, involved the first part of a split trial designed to ascertain whether the section 31 Children Act 1989 threshold was reached. The substantive hearing took place before Charles J over 8 days in March 2003 and his judgment was handed down on the 16^th April. The judge held that the child had been injured non-accidentally both on the 14th/15^th September and on an earlier occasion and that, accordingly, the threshold conditions were met. He was unable to identify the perpetrator of either injury but did not exclude any of the parents or the maternal grandmother or the night nanny from being the possible perpetrator of the second incident on the 14^th/15^th September. The local authority and the night nanny are both appellants and the parents, the grandmother and M are respondents to this appeal. The second part of the split trial, fixed for the 17^th June 2003, has now been vacated. We reserved our decision on the appeal.

The Facts

2. M is now nearly one year old and was born on the 1^st July 2002. His parents had tried to have a child over several years and had engaged in IVF treatment. The pregnancy was difficult and the mother had an emergency caesarean section with complications. He was not an easy baby to manage; he was fractious and cried a lot and was difficult to settle after feeds. He was however thriving well. After M's birth his mother was his primary carer and the father played a very limited role in his care. The mother had a considerable amount of assistance partly from her mother and primarily from a Mrs B who had been a midwife and regularly cared for the child. The parents were in business together and the mother returned to work at the beginning of September.
3. On the 10^th September the parents, accompanied by Mrs B and her husband, went to London. Mr and Mrs B left on the 12^th September and the night nanny was employed for the following three nights, 12^th/13^th/ 14^th September. She had not previously worked for the parents. She was employed through an agency and arrived with the highest credentials. On the 13^th September the maternal grandmother looked after M during the day. On the 14^th September M went out with the parents and maternal grandmother. On their return about 5.30 pm he was fed, bathed and put to bed by the maternal grandmother who then looked after him to enable the parents to go out at about 8.15 pm. The night nanny arrived at about 10pm. Between 10pm and 7.30am M was in the care of the night nanny. The parents returned about 11.30pm and went to bed. In the morning the night nanny and the mother met in the kitchen around 7 or 7.15am and the night nanny left at about 7.30am. The mother went to look at the baby and as a result of what she saw she took the child upstairs to her mother. The parents and the grandmother considered, correctly, that he was unwell and the mother and the father took M to the Accident and Emergency Department of the hospital. The child was recorded as arriving at the hospital at 8.27am.

The injuries

4. On examination M was found to be suffering from

a) fracture to the right arm which was described as a transverse fracture through the proximal shaft of the right ulna, near the elbow joint;

b) bilateral retinal haemorrhages;

c) bilateral subdural haemorrhages; and

d) some damage to the brain shown on the MRI scan.

There was also a bruise to the left forearm which did not appear until the afternoon of the 15^th September. The baby had one or more fits on the 17^th September.

5. It was accepted by all parties before the judge that the injuries were non-accidental and that the fracture, the retinal haemorrhages and damage to the brain were caused by one incident which occurred between 6.30 pm on 14^th September and the time of the baby's admission to hospital on the morning of the 15^th September. The judge described them as 'the injuries'.

The earlier injury

6. There was an issue between the parties as to whether there had been one or two incidents causing subdural haemorrhages. The neuro-radiologists identified from the imaging separate collections in the sub-dural region of different appearances which suggested bleeds of different ages. The earlier subdural haematoma described by the judge as the 'possible earlier injury' would have occurred between mid-July and the 8^th September. During this earlier period on the 11^th August the baby was unwell and was taken to the emergency doctor and was admitted to the paediatric department of the local hospital. He was discharged home later that day and there were no further problems. On the 8^th September the baby was again unwell and the mother and father contacted NHS Direct for advice. The mother was advised to take him to hospital but the child recovered and it was not considered necessary to do so. Although either of these occasions was consistent with an earlier incident causing the earlier injury, each was equally consistent with a natural explanation. The judge held, on the balance of probabilities, that there had been an earlier injury but it was not possible to identify the occasion on which it happened. It was however agreed by the family that during this period M had been in the care of the mother, the father and the maternal grandmother. Mrs B also cared for M during that period. The night nanny was not employed by the parents before the 12^th September and could not therefore have been responsible for the earlier injury.

The judge's conclusions

7. The judge's overall conclusion was as follows:

"337. On the present evidence I have reached the following conclusions to the civil standard and applying the tests set out earlier:

(1) M was injured on two occasions.

(2) On both occasions the perpetrator and anyone observing the event would realise that the handling of the child was wholly inappropriate. On both occasions the child was shaken with or without soft impact. On the later occasion the shaking was accompanied by rough handling of the right arm and overall the assault on the child was more severe than on the first occasion.

(3) I am unable to identify the person who injured M on these two occasions.

(4) I have not excluded the mother the father or the maternal grandmother as possible perpetrators of both the earlier injury and the injuries on 14/15 September 2002 or the night nanny as a possible perpetrator of the later injuries.

338. As was accepted those findings establish the threshold conditions."

8. In coming to the conclusions in paragraphs (3) and (4) above, the judge applied the test of no possibility, stating (at paragraph 155 of his judgment):

"The approach I have adopted is to ask myself whether having regard to the protective function of the court I am satisfied that there is no possibility that the relevant person injured M. As appears earlier I accept that there are problems in this approach. Further in my view it is unlikely that the House of Lords will adopt it because if they were to adopt a similar approach I think that they would be likely to define what should be regarded as "no possibility" by some pragmatic limitation and thereby make it easier for a person who had an opportunity to injure the child to satisfy the test. But as I have pointed out a "no possibility test" is arguably the approach in Re B No 2 and it is one that focuses on child protection. I have therefore decided to adopt it whilst accepting that after the decisions of the House of Lords I will probably have to revisit the issue whether the pool of possible perpetrators can be narrowed."

9. He was asked by all the parties to narrow the pool of possible perpetrators and explained why he refused to do so (at paragraph 325)

"I confess that I have reached this conclusion with some reluctance because of the difficult positions it leaves both the family and the night nanny in and because as I have explained I doubt that the guidance from the House of Lords will support the test I have adopted. However in my judgment at this stage it would be wrong for me to apply a different test or approach and thus for example one based on (i) the existence of a real possibility, or (ii) an elimination process by reference to who I thought was the least likely perpetrator in a decreasing pool which would result in a smaller pool or the identification of the most likely perpetrator."

On Appeal

10. I gave permission to appeal to this Court and added the maternal grandmother as a party to enable her to be represented on the appeal. Before this Court, there was no dispute over the fact that the child was injured, and that there were two incidents of non-accidental injury. The section 31 threshold was therefore satisfied. The main issue was the test applied by the judge in his attempt to identify who might have caused the second set of injuries. Before this Court, all the parties agreed that the test applied by the judge was erroneous and eventually agreed on the formulation of the correct test. Although the judge clearly doubted the appropriateness of his test and considered that, in the light of the awaited decision of the House of Lords in re O and N (Minors); re B (Minors) [2003] UKHL 18 (3^rd April 2003) which had not then been handed down, he might have to revisit his formulation, he did not set out what he would have done if he had applied a different test. As a result of the application by the judge of the wrong test, all the parties before us agreed the appeal would have to be allowed. The next question was how this Court should deal with the consequences. I shall return to this question below.
11. Although the case cannot be heard within the intended time span, partly because of the illness of the judge but also due to the intervention of the appeal, as it happens, the delay is far less serious than it might have been. The child is in good health and has returned to the care of his parents in their own home. Until recently the paternal grandparents were living with them and supervising them. The local authority, who have an interim care order, have now agreed that it is no longer necessary for the parents to be supervised in the home. The issues which arise as to the future arrangements for the child, based on his welfare in accordance with section 1 of the Children Act, are, we are told, in relation to whether a protection package is necessary and if it is, what the protection package should comprise, based on the premise that the child will remain in the primary care of the parents. At one level a judicial decision on the attributable limb may not be as important in this case as it might be in other similar cases.

The Law

12. Section 31 of the Children Act 1989 requires a threshold to be satisfied in applications by a local authority in public law proceedings before the welfare of and arrangements for the child concerned can be considered by the court. Section 31(2) states

"(2)   A 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's being beyond parental control."

13. There are two parts to the threshold test, the harm or likelihood of harm and the attributable condition.
14. The House of Lords in re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563 looked at the first part of the threshold test. In that case the likely risk to the children was based on the unproved allegations of rape by the step-daughter against the father. Lord Nicholls of Birkenhead said at page 584

" In section 31(2) Parliament has stated the prerequisites which must exist before the court has power to make a care order. These prerequisites mark the boundary line drawn by Parliament between the differing interests. On one side are the interests of parents in caring for their own child, a course which prima facie is also in the interests of the child. On the other side there will be circumstances in which the interests of the child may dictate a need for his care to be entrusted to others. In section 31(2) Parliament has stated the minimum conditions which must be present before the court can look more widely at all the circumstances and decide whether the child's welfare requires that a local authority shall receive the child into their care and have parental responsibility for him. The court must be satisfied that the child is already suffering significant harm. Or the court must be satisfied that, looking ahead, although the child may not yet be suffering such harm, he or she is likely to do so in the future. The court may make a care order if, but only if, it is satisfied in one or other of these respects."

15. There is no doubt in the present appeal that section 31(2)(a) has been amply satisfied by the undisputed evidence of two occasions when the child was injured. The child suffered significant harm. The problem in many cases arises as to whether the child is 'likely to suffer significant harm', as in re H. Lord Nicholls said in his speech that any conclusions had to be based on evidence and not on suspicion. He applied, however, the test of real possibility of future risk at page 591

"I have indicated that unproved allegations of maltreatment cannot form the basis for a finding by the court that either limb of section 31(2)(a) is established. It is, of course, open to a court to conclude there is a real possibility that the child will suffer harm in the future although harm in the past has not been established. There will be cases where, although the alleged maltreatment itself is not proved, the evidence does establish a combination of profoundly worrying features affecting the care of the child within the family. In such cases it would be open to a court in appropriate circumstances to find that, although not satisfied the child is yet suffering significant harm, on the basis of such facts as are proved there is a likelihood that he will do so in the future." (emphasis added)

16. Lord Nicholls revisited the test of real possibility re O and N (Minors); re B (Minors) and said at paragraph 16 (page 591)

"The second limb of condition (a) requires the court to evaluate the chance that an event will occur in the future: the child 'is likely to suffer' significant harm. In re H (minors) (Sexual abuse: standard of proof) [1996] AC 563 the House considered the matters which, in this context, the court may take into account in assessing whether the child is likely to suffer significant harm. In the context of section 31(2)(a) 'likely' does not mean more probable than not. It means a real possibility, a possibility that cannot sensibly be ignored having regard to the nature and gravity of the feared harm in the particular case. This is a comparatively low level of risk." (emphasis added)

17. In Lancashire County Council v B [2000] 2 AC 147, the House of Lords considered the second limb of the threshold test, the attributable condition in section 31(2)(b). The facts have similarities with the present appeal. The seven month old child was the victim of two shaking incidents and the three possible perpetrators were the parents and the child minder. There was another child, of the child minder, who had not been injured. The House of Lords held in respect of the injured child that, although the local authority was unable to identify which of the carers had been responsible for the injuries, the threshold conditions were satisfied. Lord Nicholls said at page 166

"I consider that a permissible and preferable interpretation of section 31(2)(b)(i), between the two extremes, is as follows. The phrase "care given to the child" refers primarily to the care given to the child by a parent or parents or other primary carers. That is the norm. The matter stands differently in a case such as the present one, where care is shared and the court is unable to distinguish in a crucial respect between the care given by the parents or primary carers and the care given by other carers. Different considerations from the norm apply in a case of shared caring where the care given by one or other of the carers is proved to have been deficient, with the child suffering harm in consequence, but the court is unable to identify which of the carers provided the deficient care. In such a case, the phrase "care given to the child" is apt to embrace not merely the care given by the parents or other primary carers; it is apt to embrace the care given by any of the carers. Some such meaning has to be given to the phrase if the unacceptable consequences already mentioned are to be avoided. This interpretation achieves that necessary result while, at the same time, encroaching to the minimum extent on the general principles underpinning section 31(2). Parliament seems not to have foreseen this particular problem. The courts must therefore apply the statutory language to the unforeseen situation in the manner which best gives effect to the purposes the legislation was enacted to achieve.    

I recognise that the effect of this construction is that the attributable condition may be satisfied when there is no more than a possibility that the parents were responsible for inflicting the injuries which the child has undoubtedly suffered. That is a consequence which flows from giving the phrase, in the limited circumstances mentioned above, the wider meaning those circumstances require. I appreciate also that in such circumstances, when the court proceeds to the next stage and considers whether to exercise its discretionary power to make a care order or supervision order, the judge may be faced with a particularly difficult problem. The judge will not know which individual was responsible for inflicting the injuries. The child may suffer harm if left in a situation of risk with his parents. The child may also suffer harm if removed from parental care where, if the truth were known, the parents present no risk. Above all, I recognise that this interpretation of the attributable condition means that parents who may be wholly innocent, and whose care may not have fallen below that of a reasonable parent, will face the possibility of losing their child, with all the pain and distress this involves. That is a possibility, once the threshold conditions are satisfied, although by no means a certainty. It by no means follows that because the threshold conditions are satisfied the court will go on to make a care order. And it goes without saying that when considering how to exercise their discretionary powers in this type of case judges will keep firmly in mind that the parents have not been shown to be responsible for the child's injuries.    

I recognise all these difficulties. This is indeed a most unfortunate situation for everyone involved: the child, the parents, the child-minder, the local authority and the court. But, so far as the threshold conditions are concerned, the factor which seems to me to outweigh all others is the prospect that an unidentified, and unidentifiable, carer may inflict further injury on a child he or she has already severely damaged."

18. The third decision of the House of Lords relevant to the issue before this Court is re O and N (Minors); re B (Minors) (above), which involved two cases of injuries to young children. In the first case the one year old child died from very serious injuries and in the second the six month old child suffered a fractured skull and other fractures. In each case the possible perpetrators were the parents. Lord Nicholls described these cases as the 'uncertain perpetrator' type of case. Under the heading 'Past events and future forecasts' he set out the general rule that the occurrence of past events required to be proved to the requisite standard and if so proved was treated as definitely having happened and if not so proved, it is treated as not having happened. He said at paragraph 11

"But the general rule does not always apply. Questions of proof of a past event arise in widely varying contexts. Sometimes the law limits the matters the decision maker may take into account. When this occurs, the reason is legal policy, not the requirements of logic."

19. In paragraphs 12 and 13 he said

"12. The same broad principle is applicable when courts or tribunals peer into the future and assess the likelihood that a particular event will occur. This is an inherently imprecise exercise, so far as human conduct is concerned. In theory it is a different exercise from deciding whether somebody did or did not do something in the past. Whether or not an alleged event occurred in the past raises a question of proof. In truth, the event either happened or not. That is not so with a future forecast. The future has not happened, and future human conduct is never certain. But in practice, the past is often as uncertain as the future. The judge cannot know for certain what happened and can only assess the degree of likelihood that something happened. The same is true of the future. The decision maker has to assess the degree of likelihood that an inherently uncertain event will occur. The degree of likelihood—beyond reasonable doubt, more probable than not, real possibility and so on—required in any particular legal context raises a question of legal policy.

13. … The legal context may permit, or require, the decision maker to take into account a real possibility that a past event occurred, or even a mere possibility. Rationality does not require that only past events established on a balance of probabilities can be taken into account. Or the context may require otherwise. The range of matters the decision maker may take into account when carrying out this exercise depends upon the context. This, again, is a question of legal policy, not logic."

20. The House of Lords was particularly concerned with the practical implications of judges finding themselves hearing care applications but not knowing which individual was responsible for inflicting the injuries. Lord Nicholls said at paragraphs 27, 28, 31 and 32

"27. Here, as a matter of legal policy, the position seems to me straightforward. Quite simply, it would be grotesque if such a case had to proceed at the welfare stage on the footing that, because neither parent, considered individually, has been proved to be the perpetrator, therefore the child is not at risk from either of them. This would be grotesque because it would mean the court would proceed on the footing that neither parent represents a risk even though one or other of them was the perpetrator of the harm in question.

28. That would be a self-defeating interpretation of the legislation. It would mean that, in 'uncertain perpetrator' cases, the court decides that the threshold criteria are satisfied but then lacks the ability to proceed in a sensible way in the best interests of the child. The preferable interpretation of the legislation is that in such cases the court is able to proceed at the welfare stage on the footing that each of the possible perpetrators is, indeed, just that: a possible perpetrator. As Hale LJ said in re G (Care proceedings: split trials) [2001] 1 FLR 872, 882:

"the fact that a judge cannot always decide means that when one gets to the later hearing, the later hearing has to proceed on the basis that each is a possible perpetrator."

 This approach accords with the basic principle that in considering the requirements of the child's welfare the court will have regard to all the circumstances of the case.

31. In 'uncertain perpetrator' cases the correct approach must be that the judge conducting the disposal hearing will have regard, to whatever extent is appropriate, to the facts found by the judge at the preliminary hearing. Nowadays the same judge usually conducts both hearings, but this is not always so. When the facts found at the preliminary hearing leave open the possibility that a parent or other carer was a perpetrator of proved harm, it would not be right for that conclusion to be excluded from consideration at the disposal hearing as one of the matters to be taken into account. The importance to be attached to that possibility, as to every feature of the case, necessarily depends on the circumstances. But to exclude that possibility altogether from the matters the judge may consider would risk distorting the court's assessment of where, having regard to all the circumstances, the best interests of the child lie.

32. Similarly, and for the same reason, the judge at the disposal hearing will take into account any views expressed by the judge at the preliminary hearing on the likelihood that one carer was or was not the perpetrator, or a perpetrator, of the inflicted injuries. Depending on the circumstances, these views may be of considerable value in deciding the outcome of the application: for instance, whether the child should be rehabilitated with his mother."

21. Although Lord Nicholls was focussing on the welfare or disposal stage of the hearing, the passages in paragraphs 31 and 32 relate equally directly to the task of the judge trying the first, section 31(2) part of the case. In his observations about the likelihood of future risk and the likelihood that carers were possible perpetrators of past injuries to children he was, in my judgment, applying the same test of real possibility.

The test applied by the judge

22. I return now to the present appeal and the test applied by the trial judge. He was unable to apply the balance of probability test positively to identify the perpetrator, since there was no evidence available to meet that standard of proof. If there were such evidence, it would not be necessary to take the next step and apply either a 'no possibility' or a 'real possibility' test.
23. The test applied by the judge was the 'no possibility that'. In relation to the second incident there were only four potential people who might have injured M and on the judge's test all of them had to be possible perpetrators. In relation to the first incident, since the date was not known, a large number of people potentially would be in the group of possible perpetrators, except the night nanny.
24. In arriving at the 'no possibility' test, the judge (at paragraph 155, reproduced above) relied upon the judgment of Thorpe LJ in Re B (No 2) (above). I do not read paragraph 38 of Thorpe LJ's judgment as applying a test of no possibility. Viewed in context, I read that paragraph as applying a test of no real possibility. Consequently I do not agree with the trial judge on his reading of Re B (No 2).
25. In my view the test of no possibility is patently too wide and might encompass anyone who had even a fleeting contact with the child in circumstances in which there was the opportunity to cause injuries.
26. In these difficult and worrying cases where the court has, as Lord Nicholls has said, to recognise and have regard to the differing interests of the adults and the child, Parliament has provided a two limb threshold which requires to be satisfied before the court has the right to consider the welfare of the child. The first is met in this appeal since the child was injured and suffered significant harm. In relation to the second limb, the attributable condition, it seems to me that the two most likely outcomes in 'uncertain perpetrator' cases are as follows. The first is that there is sufficient evidence for the court positively to identify the perpetrator or perpetrators. Second, if there is not sufficient evidence to make such a finding, the court has to apply the test set out by Lord Nicholls as to whether there is a real possibility or likelihood that one or more of a number of people with access to the child might have caused the injury to the child. For this purpose, real possibility and likelihood can be treated as the same test. As Lord Nicholls pointed out in re O and N (Minors); re B (Minors) (above) the views and indications that the judge at the first part of a split trial may be able to set out may be of great assistance at the later stage of assessment and the provision of the protection package for the injured child. I would therefore formulate the test set out by Lord Nicholls as, "Is there a likelihood or real possibility that A or B or C was the perpetrator or a perpetrator of the inflicted injuries?". There may perhaps also be the third possibility that there is no indicator to help the court decide from whom the risk to the child may come, in which eventuality it would be very difficult for the local authority and for the court to assess where the child might be at most risk.
27. Accordingly, in applying the no possibility test, the judge was wrong and his assessment of the potential perpetrators cannot stand. The appeal must be allowed.

Consequences of allowing the appeal

28. Returning to the question of disposal of this appeal, in the light of the test which should be applied, there are a number of possible outcomes. We could remit the entire case for a rehearing in the High Court and at one stage during submissions we were asked by Mr Hershman QC for the parents to do so. There are many disadvantages to such a course. Since the case took 8 days with written evidence of 15 doctors and the oral evidence of 6 doctors, that was a suggestion of last resort. In any event Mr Hershman accepted that a rehearing applying the test of real possibility would not exclude the parents whom, he agreed, would remain possible perpetrators under the real possibility test. There is therefore no good reason to remit the case for a full rehearing.
29. Another possibility would be for this Court to remit the case to the trial judge, Charles J, to apply the correct test, without rehearing the evidence, and then deal with the disposal, welfare, part of the case. A serious objection to this course was was raised by Miss Hamilton QC for the maternal grandmother who had not been a party nor represented at the court below and had not had the opportunity to cross-examine the doctors, in particular, on the crucial issue of timing. There was some medical evidence that the injuries were more likely to have occurred after midnight, in which case the maternal grandmother would be excluded from consideration as a possible perpetrator. As Miss Hamilton pointed out, the maternal grandmother would be seriously prejudiced if the test were to be applied on the existing evidence. Although submissions were made that the maternal grandmother was receiving legal advice and was in a position to be represented at the hearing, it is clear to me that no-one seriously expected her to be considered and found to be the perpetrator, and her attendance at the hearing as a witness and not as a party was entirely understandable.
30. A further possible approach to this difficult situation would be to set aside the findings of the judge but not to substitute findings in this Court or remit to the trial judge, leaving the matter with no finding or indicators as to who the perpetrators might be. It may not be necessary for the immediate disposal of this case for the local authority to have such findings. This suggested approach however places the local authority in some difficulties. As long as there is an outsider, the night nanny, who is a possible perpetrator even if not so labelled, it may be difficult for the parents to focus on the need for protection within the family. It may also have an effect on the approach of the local authority to the issue of protection of any other child who may be born to the parents. It cannot be ignored that there was an earlier incident when the baby was shaken at a very few weeks old. It is also unfair to the night nanny whose career is already affected by the position she is in. It is unfair to the maternal grandmother who has, and may have, other grandchildren and may be considered a possible risk to them. The parents sought to keep the night nanny in the group of possible perpetrators but not the maternal grandmother. The local authority wished to exclude the night nanny and the maternal grandmother as well as the father. The guardian supported the exclusion of the night nanny and the maternal grandmother. The night nanny sought to keep the family, including the maternal grandmother, in the group, so long as she herself was in it. In my view, fairness to the night nanny and the grandmother requires this Court to look again at the position of each of them. It is clearly not necessary to revisit the position of the parents in the light of the pragmatic recognition by Mr Hershman that it would not exclude them as possible perpetrators. The local authority had submitted that the father should be excluded, primarily on the basis of the mother's evidence both as to the limited amount of time that the father cared for the child and that he had not been alone with the child over the relevant period on the 14^th/15^th September. That support for and exculpation of the father was derived entirely from the mother and, as the judge recognised, in this case the credibility of the parties could not be accepted at face value. If it were otherwise, the injuries to the child were inexplicable. The parents, at least at this stage, have to be treated together; they are connected and neither can exonerate the other, unless or until someone tells the truth about what actually happened on either occasion which caused the injuries to M. Both of them therefore have to be treated on the basis that there is a real possibility that one or both injured the child.

The night nanny

31. The night nanny is in a different position from the other parties, in that she can be positively excluded from the first injury. The judge's view of her written and oral evidence and her demeanour was that they did not warrant a conclusion that she had not been telling the truth. His general impression was favourable and in line with the accepted improbability of her having injured M which was in line with her training, her references and the evidence before him. That conclusion was fortified by a comparison with the evidence of other members of the family. At the end of his judgment he said

"326. I acknowledge that my finding based on authority that there was an earlier injury founds an argument that I should exclude the night nanny from the pool of possible perpetrators on the basis of that finding and the proposition that it is unlikely that M was injured on the second occasion by the night nanny and on the earlier occasion by one of the other possible perpetrators of the later injury.

327. However as I have pointed out both of the stepping stones in that argument contain elements of doubt and having regard to the general impression given by all the possible perpetrators and the severe difficulties in assessing credibility that exist in this case I have concluded that on the present evidence the test I have set myself to exclude the night nanny as a possible perpetrator of the later injuries is not satisfied."

32. As I understand his judgment, the stepping stones include the judge's earlier uncertainty about whether there had been an earlier injury. That question he had however already resolved in the affirmative, and it would therefore be impermissible to put it back into the equation.
33. The judge's assessment of the night nanny lends strength to the proposition that she would be unlikely to have caused the injuries on the 14^th/15^th September. That proposition might not stand alone but added to it is the important fact, recognised by the judge, that it was highly improbable, although possible, that two separate people would have shaken this baby and caused these injuries within the period of the first 11 weeks of M's life. This has been the view of the local authority and the guardian.
34. For my part, I am satisfied, for the reasons I have given above, that there is no real possibility that the night nanny caused the injuries to M on the second occasion.

The maternal grandmother

35. No-one suggested until shortly before the hearing that the grandmother might have caused the injuries on the second occasion. She was not specifically referred to by the night nanny when she became a party. The allegation was that a member of the family caused the injuries. Now that I have excluded the night nanny none of the remaining parties suggests that the maternal grandmother is a possible perpetrator. All the remaining parties ask that she be excluded. The absence of suggestion from any of the parties, particularly the local authority, that a non party might be the perpetrator, is not a complete bar to the court nonetheless finding that there was a real possibility that that non-party might have caused the injury. It is however undoubtedly a factor to which the court should give some weight. There are other factors to support the unanimous view of the remaining parties. There was medical evidence to show that the bruise which appeared during the latter part of the 15^th September indicated that the injury was inflicted not earlier than midnight. The medical evidence about the fit on the 17^th September pointed to a period more likely after midnight than earlier during the previous evening. Miss Hamilton made cogent points about procedural irregularity and procedural unfairness to which I have referred briefly above. Again by themselves irregularities would not be a proper reason for exclusion of a possible perpetrator if there were other evidence to support her inclusion, although they might well require in some cases a retrial. In the present case, however, looking at all the circumstances, I am satisfied that there is no real possibility that the maternal grandmother caused the injuries on the second occasion. She was also one of those within the judge's test for the first injury. I do not consider that, if she did not cause the second set of injuries, she was likely to have caused the first. I would therefore exclude her from being considered a possible perpetrator of either set of injuries to M.
36. The effect of excluding the night nanny and the maternal grandmother is to leave for the assessment of the local authority and of the judge dealing with the second part of the case, if it becomes necessary, the real possibility that either or both parents caused both sets of injuries to their child.

Identification of issues and split trials

37. I would just add that, in the absence of new evidence affecting the possible identifiable perpetrators, it would seem to me difficult to revisit a decision made on the first part of a split hearing at the second part, particularly if one possible perpetrator had not been a party at the first stage of the proceedings. It might also be very unjust to a person identified as a possible perpetrator who had not been a party in the first part of the proceedings, such as the grandmother, who had not had the opportunity to play an active part in the earlier part of the case. This leads me to two further reflections. First it is particularly important to identify at an early stage all the relevant issues and all those in respect of whom it might be said that there was a real possibility that he or she might have caused the injuries. Second, I think it is important to consider at a very early stage whether a split trial is the best way forward or whether it might, in many cases, be better to hear the whole case and then, if necessary, adjourn for further assessments. These reflections underline the enormous importance of the Judicial Case Management Protocol which is just about to be launched and which will apply to all public law cases. The Protocol will require careful case management of each care case from the earliest stage and should identify problems such as arose in this case.
38. I would allow the appeal.

Lord Justice Thorpe

39. I have had the advantage of reading in draft the judgment of the President. I am in complete agreement with it and with her reasoning.
40. I only add a brief separate judgment given the reliance which Charles J placed on a passage from my judgment in Re B (Children: Non Accidental Injury) [2002] EWCA Civ 902, [2002] 2 FCR 654 in coming to his conclusion that an individual was only to be exculpated if there was no possibility that he or she had inflicted the injury or abuse.
41. Charles J at paragraph 150 of his judgment said:

"However it seems to me that it is inherent in the judgments in Re B that the Court of Appeal in that case envisaged that at the welfare or disposal stage the judge should approach the decision as to what would best promote the welfare of the child on the basis that the mother presented a risk to the child. That risk is not quantified but paragraphs 38 and 44 of the judgment of Thorpe LJ indicate that unless it can be said to what he describes as a degree of heightened cogency that the mother could not possibly have injured the child (my emphasis) she should be treated as a risk, or as having a question mark attaching to her. This test creates problems. For example how can it be said on a balance of probabilities or otherwise that a person who had the opportunity to cause the relevant injury to the child could not possibly have done so. In any event it is a very high test to found a conclusion that for the future a person should be treated on that basis and thus as if he or she presented no risk to the child (or other children) and no question mark attaches to him or her."

42. In the Court of Appeal the issues in the case of Re B were essentially factual rather than legal. Given the expert medical evidence was it open to the judge to exonerate the mother of either inflicting any of over 90 injuries on her child or of failing to seek medical treatment for the child or of failing otherwise to protect her child during the course of a recently developed relationship with her cohabitee? The court held that the medical evidence precluded such an outcome. So the legal issues later considered by the House of Lords were not ventilated in this court and my observations in paragraph 38 of an extempore judgment must be read in that light and also in conjunction with paragraphs 14, 43 and 44.
43. In paragraph 38 I said:

"Of course in the case of a single possible perpetrator there is no doubt that in deciding whether or not he or she is responsible for a physical assault, the court must apply the civil standard, elevated in accordance with the guidance given by the House of Lords in the case of Re H and R [1996] AC 563. Of course there was evidence of sufficient cogency to lead to the conclusion that KR was a perpetrator. But was there cogent evidence that he was the sole perpetrator of all 94 injuries? In situations like this the trial judge in the Family Division is in a position similar to the position of judge and jury in criminal proceedings. There, when both parents stand in the dock charged with the infliction of injury on a baby, and when their defence is either each to blame the other or to offer no credible explanation, how is it possible to determine which is guilty? The same sort of dilemma faces the judge in a situation such as this, where it is incumbent upon him to apply the elevated civil standard of proof. A degree of heightened cogency is necessary to enable the judge to say that it could not possibly have been the mother."

44. I did not conceive that that passage would be read as meaning that one parent could only be eliminated if there was no possibility that he or she was the perpetrator. No possibility effectively means no opportunity. By contrast no real possibility allows a review of all relevant facts and circumstances including opportunity. That is precisely what I envisaged as a necessary prelude to exculpation. Of course the phrase emphasised by Charles J ('… it could not possibly have been the mother.') read in isolation could be said to point to a no possibility test. But I intended that phrase to mean no more than the mother's exculpation.
45. Some relationship between the standard of proof justifying inculpation and that justifying exculpation arises starkly in cases where the injuries were caused by one or other or both of two parents. Inculpation cannot legitimately result merely from elimination but only from a review of all relevant facts and circumstances in relation to each and the application of the standard of proof defined in the case of Re H and R..

Lord Justice Clarke.

46. I agree with both judgments.