(Family Court, Peter Jackson J, 23 May 2014)
Fact-finding hearing – Death of child – Previous care proceedings – No medical aetiology
The two children, aged 3 and one, were removed from the family home and placed in foster care following the death of the youngest child’s twin sister at the age of 16 weeks. The 3-year-old child had previously been subject to care proceedings when she sustained a skull fracture at 8 months old but proceedings were subsequently withdrawn and she returned home.
In these proceedings the evidence of a paediatrician given in the first proceedings that the mother’s account of the older child falling down the stairs could have accounted for the skull fracture was discounted and therefore these proceedings included an investigation into the child’s death and the 3-year-old’s injury.
The parents were a close, intelligent and successful couple and accounts of their parenting were positive. Those aspects of their general good character were relevant to an assessment of the probabilities of the local authority case that they had concealed the true mechanism of the death and injury sustained by their children but did not prevent the local authority from proving its case. It was submitted that the child died as a result of inadequate sleeping arrangements on the sofa whereby the child suffocated.
The post-mortem of the deceased child found no ascertainable cause of death. Sudden infant death could not be ruled out as a possible cause of death which was more likely in children such as this who were both twins and born prematurely.
Even making a full allowance for the extremely distressing circumstances surrounding the child’s death, the parents were not witnesses whose accounts could always be accepted without question.
The evidence in relation to the oldest child’s head injury pointed to it having been an accidental injury possibly related to inadequate supervision but that was not sufficient for local authority intervention.
The child’s death by suffocation while sleeping on the sofa was not more probably than not. The medical evidence allowed for such a conclusion but did not demand such a finding. Although the parents’ evidence had been unsatisfactory on that issue it did not add weight to the medical evidence. It was not possible to say exactly how or when the child died. In those circumstances any finding against the parents would be based on speculation or suspicion which would be impermissible and wrong. The sleeping arrangements for such small babies were unwise but did not cross the threshold for intervention. The local authority applications were dismissed and the children would be returned to the parents’ care.
The fully referenced, judicially approved judgment and headnote will appear in a forthcoming issue of Family Law Reports. A detailed summary and analysis of the case will appear in Family Law. __________________________________________________________________
IMPORTANT NOTICE This judgment was delivered in private. The judge has given leave for this version of the judgment to be published on condition that (irrespective of what is contained in the judgment) in any published version of the judgment the anonymity of the protected person and members of his family must be strictly preserved. All persons, including representatives of the media, must ensure that this condition is strictly complied with. Failure to do so will be a contempt of court.
Case No: LA13C00454
Neutral Citation Number:  EWHC 3 (Fam)
IN THE FAMILY COURT SITTING AT MANCHESTER
23 May 2014
THE HONOURABLE MR JUSTICE PETER JACKSON
- - - - - - - - - - - - - - - - - - - -
Lancashire County Council
The Children (by their Children’s Guardian)
M -and- F
- - - - - - - - - - - - - - - - - - - -
Susan Grocott QC and Julia Cheetham (instructed by Lancashire County Council) for the Applicant
Jane Cross QC and Carolyn Bland (instructed by JWR Law) for the Mother
Gillian Irving QC and Prudence Beever (instructed by Birchall Blackburn Law) for the Father
Samantha Bowcock (instructed by Forbes Solicitors) for the Children’s’ Guardian
Hearing dates: 6 – 21 May 2014 Judgment date: 23 May 2014
- - - - - - - - - - - - - - - - - - - - -
Mr Justice Peter Jackson:
 These care proceedings concern two young children, both girls. They are K, born in March 2011, who is 3 years old, and S, born in May 2013, who is aged just 1. Lancashire County Council issued the proceedings in September 2013 following the sudden death of S’s identical twin sister L on 11 September 2013 at the age of 16 weeks. Since then, K and S have been in foster care with a very high level of parental contact.
 This is the second time proceedings have been issued in relation to K. On 14 November 2011, when she was aged 8 months, her mother took her to hospital with a swollen head; this was investigated and it was found that she had a fractured skull. Proceedings were taken but in September 2012 they were withdrawn with the court’s approval and K, who had been in foster care for 10 months, went home. The present Children’s Guardian acted for K in those proceedings and the current social work team manager was also involved.
 In the original proceedings, the mother gave an account of K having fallen down some stairs at home 12 days before she was taken to hospital. A paediatrician accepted that this might explain the skull fracture and swelling and it was in that context that the proceedings were withdrawn. In these proceedings, that explanation has been discounted by the medical experts, and no party now suggests that such an incident could realistically have been responsible for K’s injury.
 This hearing has therefore been an investigation into the injury to K and the death of L. The written evidence is spread across a dozen files and oral evidence was given over the course of eight days. In addition to conventional written submissions, counsel have created several agreed documents. These include summaries of the medical evidence, of the sequence of events from the moment the parents sought medical attention for each child, and of the statements made by the parents to others. These helpful documents were possible because of planned breaks in the evidence when the court could not sit for other reasons. This has saved court time and clarified the issues. In this instance, the effect of the medical evidence was agreed before the parents gave their evidence.
 The local authority's case at the outset of the hearing was that the parents were responsible for ill-treatment of K and L either by way of inflicting injury upon them or from accidents that were concealed or arose from a lack of care. Having tested the evidence, the local authority no longer pursues allegations of inflicted injury, and in particular any allegation of assault or deliberate suffocation. However, it alleges that one or both of the parents knows how K’s head injury occurred, but that they are concealing it. It further alleges that L probably died as a result of her father lying on her when he was sleeping on the same sofa, or that he recklessly put her to sleep on the sofa in such a way that she rolled and could not breathe. 6.No criticism is made of the parenting of K during the year when she was at home between her first and second removal, nor of the parenting of S at any time.
 The mother accepts that the sleeping arrangements for the twins on the night L died were, in her words, "not the greatest idea". Apart from that, the parents dispute the local authority's case.
 In A Local Authority v (1) A Mother (2) A Father (3) L & M (Children, by their Children’s Guardian)  EWHC 1569 (Fam), Mr Justice Baker summarised the legal principles in a manner that I adopt and apply:
1.First, the burden of proof lies at all times with the local authority.
2.Secondly, the standard of proof is the balance of probabilities.
3.Third, findings of fact in these cases must be based on evidence, including inferences that can properly be drawn from the evidence and not on suspicion or speculation …
4.Fourthly, when considering cases of suspected child abuse, the court must take into account all the evidence and furthermore consider each piece of evidence in the context of all the other evidence. The court invariably surveys a wide canvas. A judge in these difficult cases must have regard to the relevance of each piece of evidence to other evidence and to exercise an overview of the totality of the evidence in order to come to the conclusion whether the case put forward by the local authority has been made out to the appropriate standard of proof.
5.Fifthly ... whilst appropriate attention must be paid to the opinion of medical experts, those opinions need to be considered in the context of all the other evidence. It is important to remember that the roles of the court and the expert are distinct and it is the court that is in the position to weigh up the expert evidence against its findings on the other evidence. It is the judge who makes the final decision.
6.Sixth, cases involving an allegation of non-accidental injury often involve a multidisciplinary analysis of the medical information conducted by a group of specialists, each bringing their own expertise to bear on the problem. The court must be careful to ensure that each expert keeps within the bounds of their own expertise and defers, where appropriate, to the expertise of others.
7.Seventh, the evidence of the parents and any other carers is of the utmost importance. It is essential that the court forms a clear assessment of their credibility and reliability.
8.Eighth, it is common for witnesses in these cases to tell lies in the course of the investigation and the hearing. The court must be careful to bear in mind that a witness may lie for many reasons, such as shame, misplaced loyalty, panic, fear and distress, and the fact that a witness has lied about some matters does not mean that he or she has lied about everything (see R v Lucas  QB 720).
9.Ninth, as observed by Dame Elizabeth Butler-Sloss in an earlier case
“The judge in care proceedings must never forget that today’s medical certainty may be discarded by the next generation of experts or that scientific research would throw a light into corners that are at present dark.”
This principle, inter alia, was drawn from the decision of the Court of Appeal in the criminal case of R v Cannings  EWCA 1 Crim. In that case a mother had been convicted of the murder of her two children who had simply stopped breathing. The mother’s two other children had experienced apparent life threatening events taking a similar form. The Court of Appeal Criminal Division quashed the convictions. There was no evidence other than repeated incidents of breathing having ceased. There was serious disagreement between experts as to the cause of death. There was fresh evidence as to hereditary factors pointing to a possible genetic cause. In those circumstances, the Court of Appeal held that it could not be said that a natural cause could be excluded as a reasonable possible explanation. In the course of his judgment, Judge LJ (as he then was) observed: “What may be unexplained today may be perfectly well understood tomorrow. Until then, any tendency to dogmatise should be met with an answering challenge.”
10.With regard to this latter point, recent case law has emphasised the importance of taking into account, to the extent that it is appropriate in any case, the possibility of the unknown cause. The possibility was articulated by Moses LJ in R v Henderson-Butler and Oyediran  EWCA Crim. 126 at paragraph 1: “Where the prosecution is able, by advancing an array of experts, to identify a non-accidental injury and the defence can identify no alternative cause, it is tempting to conclude that the prosecution has proved its case. Such a temptation must be resisted. In this, as in so many fields of medicine, the evidence may be insufficient to exclude, beyond reasonable doubt, an unknown cause. As Cannings teaches, even where, on examination of all the evidence, every possible known cause has been excluded, the cause may still remain unknown.”
11.In Re R (Care Proceedings: Causation)  EWHC 1715 (Fam), Hedley J, who had been part of the constitution of the Court of Appeal in the Henderson case, developed this point further. At paragraph 10, he observed: “A temptation there described is ever present in Family proceedings too and, in my judgment, should be as firmly resisted there as the courts are required to resist it in criminal law. In other words, there has to be factored into every case which concerns a discrete aetiology giving rise to significant harm, a consideration as to whether the cause is unknown. That affects neither the burden nor the standard of proof. It is simply a factor to be taken into account in deciding whether the causation advanced by the one shouldering the burden of proof is established on the balance of probabilities.
12.Finally, when seeking to identify the perpetrators of non-accidental injuries the test of whether a particular person is in the pool of possible perpetrators is whether there is a likelihood or a real possibility that he or she was the perpetrator. In order to make a finding that a particular person was the perpetrator of nonaccidental injury the court must be satisfied on a balance of probabilities. It is always desirable, where possible, for the perpetrator of non-accidental injury to be identified both in the public interest and in the interest of the child, although where it is impossible for a judge to find on the balance of probabilities, for example that Parent A rather than Parent B caused the injury, then neither can be excluded from the pool and the judge should not strain to do so.”
 To these matters, I would only add that in cases where repeated accounts are given of events surrounding injury and death, the court must think carefully about the significance or otherwise of any reported discrepancies. They may arise for a number of reasons. One possibility is of course that they are lies designed to hide culpability. Another is that they are lies told for other reasons. Further possibilities include faulty recollection or confusion at times of stress or when the importance of accuracy is not fully appreciated, or there may be inaccuracy or mistake in the record-keeping or recollection of the person hearing and relaying the account. The possible effects of delay and repeated questioning upon memory should also be considered, as should the effect on one person of hearing accounts given by others. As memory fades, a desire to iron out wrinkles may not be unnatural – a process that might inelegantly be described as "story-creep" may occur without any necessary inference of bad faith.
 In the present case, I bear all of these matters in mind. They are perhaps particularly relevant to the events concerning K, which arose in 2011 and seemed to have been resolved in 2012, only to re-emerge a year later.
 The children’s mother is aged 29 and comes from Lancashire. Their father, aged 28, is from overseas. They met in the summer of 2009 and married in February 2010. They work together in their own small business.
 The parents are a very close couple. They are intelligent, hard-working and aspirational. This is seen not only in the business, started by the mother alone at a very young age, but also in their home and garden, which has been transformed by their efforts. They have no criminal or antisocial characteristics, do not abuse drugs or alcohol, and there is no sign that they are given to outbreaks of temper. All observations confirm that they are capable and committed parents in normal circumstances. The children were planned and are much loved. K and S are happy youngsters who show every sign of good parenting. The quality of contact has been outstanding. Although the mother, for reasons that cannot be held against her, is currently estranged from her family, the couple has good social relationships and friendships with others, some of whom gave evidence, and they are well regarded in their community. The father's family, though abroad, is another source of support.
 The tragic events with which the court is concerned must be assessed within this context. The parents’ general good character is relevant to the assessment of probabilities, though it does not of course prevent the local authority from proving its case.
 Photographs of the home at various times show that in 2011 the living room had not yet been renovated and contained a marble fireplace and hearth, and that in 2013 the room was dominated by a huge, three-sided sofa.
The injury to K’s head
 The family first came to the attention of social services in 2011. On 14 November, the mother took her to the GP with a large swelling to the side of her head. On advice, K was taken to hospital, where tests were carried out and the parents were asked to return the following day for a CT scan. This was initially said to be normal and K was again sent home. However, on review of the scan a fracture was detected and K was returned to hospital on 16 November, where she remained until her discharge into foster care a few days later. A full skeletal survey was normal. K was noted to be a well-nourished infant who was developmentally advanced, was crawling and could pull herself up.
 The police and social services were informed and the parents were interviewed. Care proceedings began.
 Consultant opinions about K’s head injury have been given by:
Dr W, paediatric radiologist, in hospital at the time Dr Karl Johnson, paediatric radiologist, in the initial proceedings Dr Dawn Saunders, neuroradiologist, in the initial proceedings Dr Ian Mecrow, paediatrician, in both sets of proceedings Dr Neil Stoodley, paediatric neuroradiologist, in these proceedings Dr Jayaratnam Jayamohan, paediatric neurosurgeon, in these proceedings Professor Timothy David, paediatrician, in these proceedings
 The medical evidence can be summarised as follows:
1)K had a right-sided parietal skull fracture and a sub-galeal haematoma about 3” x 2” in size. This swelling was noted as non-tender and fluctuant. A sub-galeal haematoma is a collection of blood and tissue fluid beneath the lining of the scalp. It had a small amount of fresh blood in it which could not be dated accurately from a radiological perspective. It was not due to a leakage of cerebrospinal fluid.
2)The fracture and haematoma were caused by an impact to the head. It is likely that they occurred at the same time. The precise degree of force required is unknown.
3)The skull fracture was less than ten weeks old. It is the most common area for the skull to be fractured and its type and location do not help to discriminate between accidental or inflicted injury. 4)The true incidence of injuries of this nature is not known as not all children are presented for medical attention. The fracture could have been sustained in an unwitnessed fall. If K fell from standing height on to the floor it would be unusual but not impossible for her to have sustained this injury. Low level falls without collision with other objects rarely cause fractures.
5)The haematoma cannot be dated radiologically. From a clinical perspective, the scalp swelling would probably start within a couple of hours, reaching its peak within a couple of days. It would probably resolve within a maximum of ten days, with five to seven days being a reasonable average. This swelling was last noted on the fifth day after admission to hospital.
6)Consequently, an incident on 2 November would be extremely unlikely to result in a scalp swelling first seen on 14 November. The possibility was described by a number of witnesses as being fanciful.
7)The visibility of the swelling would depend on matters such as the competence and attentiveness of the carers and the length and texture of the child’s hair.
8)Skull fractures are almost always immediately painful but do not always cause ongoing pain. When the fracture occurred it is likely that K would have been distressed and that her carer would be aware that she may have had been injured. If K had an unwitnessed accident while unattended, she may have settled by herself. Thereafter, there may have been little in the way of symptoms to suggest to a carer that she had a fractured skull.
9)A subgaleal haematoma can be uncomfortable for a period of time but some can be completely painless. The child’s reaction could have varied depending on the degree of force used when the swelling was touched.
 The morning of Monday 14 November was an unusually busy time for the parents. A new kitchen was being delivered first thing and the normally tidy house was in some disarray. The father had got up very early to get things ready. When K woke, he fed her and gave her a shower. The mother got up, dressed K, and went out to collect some timber, leaving K with the father for a few minutes. On her return, the delivery men were bringing the kitchen into the house. When the delivery was complete, a process taking maybe half an hour, she left the father at home and went to open the shop, taking K with her. She ordered breakfast from a neighbouring cafe and describes noticing the swelling on K's head while Mr C, the cafe owner, who is a friend, was delivering it. She telephoned the father to ask if he had seen it. She made an appointment with the GP and from there was sent to the hospital.
 As to the exact moment when the swelling was first seen, there was some difference between the evidence of the mother and Mr C. He describes the mother noticing the swelling while she was holding K in her arms to feed her, while she initially described seeing it while K was playing on the floor. She now believes that Mr C may be right. I do not regard this as necessarily sinister. The gap between K being on the floor and being in her arms may have been a few seconds only, and in recalling events such a long time ago M may be engaging in what I have referred to above as "story-creep".
 On either 15 or 16 November (again in my view it does not matter which), the mother described an occasion at home when K had fallen off the flight of stairs that adjoins the sitting room. Ultimately, the date for this was given as 2 November. The mother and her friend Ms G were in the sitting room. From where they were sitting, they could see the hall at the foot of the stairs, but not the stairs themselves. K, who was by then cruising, managed to climb up and fall off one or more stairs.
 The mother describes an unfortunate but fairly mundane incident. She heard a thud and saw K on the floor at the bottom of the stairs. She picked her up. K looked momentarily shocked, but seemed fine after a few seconds and wanted to play. There was no need to take her to the doctor.
 Ms G, who would have had a slightly better view than the mother, gives an altogether more vivid account, involving her hearing several thuds, suggesting that K had fallen down much of the flight of stairs, and seeing her “cartwheeling” into sight and falling on to the floor. She agrees that K was shaken but not greatly affected and she says that the mother asked her not to tell the father what had happened, as she had left the stairgate open. The mother says she does not remember saying that, but I find that she did.
 The mother and Ms G made statements in the original proceedings, but were not called upon to give evidence at the time. Having now heard their evidence, I am satisfied that K did fall off one or two steps on 2 November, but that it was a relatively minor event that has nothing to do with the skull fracture. Ms G was a well-meaning but excitable witness, and I am afraid that her passionate belief that the parents are innocent has led her to overdramatize the event.
 Evidence was called about a conversation between Ms G and a friend who happened to be a social worker. Having heard the friend’s evidence, I am satisfied that that conversation has no bearing on the matter.
 It is unfortunate that the evidence about K’s fall was not evaluated promptly during the original proceedings. Had that happened, the medical experts, the parties and the court could have seen the incident for what it was. But at the time, Dr Mecrow considered the case to be difficult and gave the opinion on a fine balance that the fracture and swelling were days and even weeks old when K was taken to hospital. He was influenced by the absence of tenderness when the swelling was touched. This point of view prevailed at an experts meeting.
 The local authority accordingly considered that the medical evidence placed it in difficulties in proving its case. It applied to withdraw the proceedings in September 2012, presenting a full written argument that drew the court's attention to anomalous features. It pointed in particular to the absence of any report of K crying following the injury and to the fact that, if the injury occurred on 2 November, no swelling was observed for 12 days. No written analysis was presented to the court by or on behalf of the Children's Guardian, who made submissions through counsel.
 Her Honour Judge Watson granted the application, which had the support of the Guardian. She gave a short judgment, noting the state of the medical evidence, the positive features in the family, the disadvantages for K of further delay and the cost of continued proceedings.
 In my view, the court should have required a written analysis from the Children's Guardian before granting the local authority's application. This was a serious injury and the context was not clear. The factual basis for the hypothesis that a fall on the 2 November might be relevant was not investigated.
 What the court could not have known at that time is that Dr Mecrow now accepts that an old article he had relied upon (referring to the possibility of late-presenting scalp swelling arising from leakage of cerebrospinal fluid) does not apply in K’s case. Professor David also convincingly challenges the proposition that the swelling’s lack of tenderness indicates an older injury.
 I find that the Ks’ injury was the result of a single impact to the head and that this almost certainly occurred on the morning she was taken to hospital. Had the swelling been present earlier than that, it would have been seen. It is worth noting that three days earlier, K had spent the day with a friend of the family who would have seen any abnormality to her head, but did not. Nor does either parent describe seeing anything when bathing or dressing her at the start of the day itself.
 I concur with the stance of the local authority that the injury to K is not likely to have been inflicted by either parent. On the basis of the medical findings, Professor David does not contend for inflicted injury. That conclusion has not, in my view, been shifted by the wider evidence. I therefore find that the injury was likely to be the result of an accident. The question is whether the local authority has proved that one or both of the parents know(s) how it occurred.
 The following matters are relied upon as suggesting that the parents know more than they are saying:
•On their own account K was closely supervised in a small house and was only out of their sight for moments at a time. There were no obvious hazards in the home, and she was never seen to be distressed in the way that would be expected after a blow to the head.
•Accordingly, for this to be an unwitnessed accident, K would have to be in the small group of children who injure themselves in low-level falls, and also to be an unusual child who did not make her injury known to carers.
 Against this, these points are made on behalf of the parents:
•K was an advanced and active child for her age. The Guardian, who saw her at the time, describes her as exceptionally mobile.
•There is some evidence that K tolerates pain without protesting. She is said to have made little fuss when having milk teeth extracted on one occasion and when cutting her leg at nursery on another.
•The parents' response to discovering the injury was entirely appropriate.
•The parents' accounts of the events of the morning have generally been consistent.
•It is possible that an unwitnessed incident occurred during the disruption surrounding the delivery of the kitchen.
 A final evaluation of these possibilities depends in part on an assessment of the parents' credibility. Accordingly, I now turn to the evidence relating to the death of L.
The birth of the twins
 L (Twin 1) and S (Twin 2), were born on the 17 May 2013 at 32½ weeks gestation by elective caesarean section due to intra-uterine growth retardation of L. After their birth the twins were transferred to the neonatal unit before discharge home on 14 June.
 The parents were seen informally on several occasions by the Community Neonatal Team in preparation for discharge home. Various issues were discussed, including temperature control, feeding, sterilising, bathing, medications, cot death prevention, resuscitation, vaccinations and registering the twins at the GP. Both parents were amenable and co-operative.
 Following the twins’ discharge, the Community Neonatal Team made seven visits to the home and all observations were positive, in particular regarding the mother’s handling of the twins. Safe sleeping guidance had clearly been followed. The last visit was a joint visit on 28 June with the Health Visitor who was taking over the case.
 The Health Visitor visited on 3 July. There were three further visits before L’s death. On two of the visits, safe sleeping guidelines were discussed. No significant concerns were noted regarding L’s health and welfare or her parent’s capacity to meet her needs. On 31 July, home conditions were described as immaculate.
 The father’s sister visited from abroad for a month from mid-July.
 At the end of August, the Health Visitor received the records for K and noted what had happened in her case. The parents had not spoken of it.
 There is no suggestion that S has ever come to any harm and investigations have not shown her to have any underlying medical conditions.
The death of L
 At 7.35 a.m. on 11 September, the Mother made an emergency call saying that L had stopped breathing. An ambulance and rapid response vehicle arrived at 7.37 a.m.
 The paramedics could find no sign of life in L and took her immediately to hospital, where she was met by the crash team at 8.03 a.m. Following intensive efforts at resuscitation, L was pronounced dead at 8.30 am in the presence of her parents.
 The Lancashire protocol for cases of sudden death in infancy (SUDI) was immediately put into effect. The police and social services were immediately on the scene and the family home was visited and photographed that day. I commend the quality of the support for the family and of the gathering of information.
 The court has read or heard evidence from many who were involved on that day. Apart from the parents themselves, oral evidence was given by two paramedics (Mr L and Ms S), a trainee specialist registrar (Dr H), a consultant paediatrician (Dr D), a specialist SUDI nurse (Ms C), a police constable (PC H), the detective chief inspector leading the investigation (DCI H) and a detective sergeant (DS O).
 Mr L, the first paramedic on the scene, entered the house and found L lying lifeless on the floor. He did not notice any hypostasis or rigor mortis. In contrast, his colleague Ms S described L as being cold to touch and slightly stiff. Neither paramedic noticed any blood on or around L’s face. Mr L had a brief conversation with the mother, who told him that she had last heard the child crying at about 6 a.m. The father was standing in the doorway of the house, motionless and silent. I note in passing that some comment has been made about his unusual behaviour on this occasion. It is not relied upon by the local authority and I do not attach significance to it. His account of being in shock may offer an explanation.
 At hospital, Dr H recalls that L was cold to touch. He was told by the paramedics that the baby had been heard crying at 6.10 a.m., and was cuddled and put back to sleep in a rocker. The mother had gone upstairs to feed the other twin but was called by the father at 7.10 a.m. as something was not right.
 Dr H says that while resuscitation was continuing, the anaesthetist noted frothy, slightly bloodstained, secretions coming up the endotracheal tube. Dr H considered this to be a consequence of the efforts to resuscitate. He did not observe any blood in L’s nose or mouth. Had there been, he would have documented it in his notes, which are full.
 Dr D arrived in the department around 8.25 a.m. After he had examined L and consulted with the crash team, further resuscitation was considered futile and death was pronounced in the presence of the parents. At the date of death L’s corrected age was 8-9 weeks. His evidence in relation to the presence of blood was effectively the same as that of Dr H.
 The mother referred in her court statement to having seen blood coming from L’s nose and being wiped away during the resuscitation attempts. This is the only reference to blood as such. It was understandably an extremely distressing sight, but on the whole of the evidence I am satisfied that what she saw were frothy secretions with traces of blood in them, and not what has been described during the hearing as "frank blood".
 L’s body was examined within an hour by PC H and a colleague. The mother told PC H that the father had told her that he had been woken by the twins at 6 a.m. and that he had fed L and put her to sleep. At 7.15 a.m., he had checked on her and found something wrong, whereupon he had called the mother.
 Dr D spoke to the parents in a side room at 9.25 a.m. His notes, which were written up afterwards, give the parents' account of the events of the night. The notes do not include information about when L was last seen alive.
 At 11.30 a.m., the parents were spoken to by DI H, and Nurse C, with notes being taken by DS O and by a colleague of Nurse C. There are some discrepancies in the details of their respective accounts, but the broad picture, given mainly by the father, is that he last fed L at 1 a.m. and that she had woken crying at about 6 a.m. He had attempted to feed her and had put her in her rocker. He had then gone upstairs to wake the mother so that she could feed S and he could get some sleep. An hour later he was then woken by the mother when K started stirring. He then went downstairs, found that L was not breathing, and called the mother, who rang for an ambulance.
 The officers and the nurse were questioned on behalf of the father to the effect that he had not given these times. Having heard their evidence and read the two sets of notes, which coincide with each other, I am sure that they are a broadly accurate record.
 Later that afternoon the officers and Nurse C went to the family home. The property was cluttered and untidy. There was a cot in the bedroom which had clearly not been in use recently as it had clothes and other items in it. In the living room on the sofa there was a nearly empty baby bottle containing milk and another bottle containing an orange substance that looked like baby food. There were several cushions around the area where the father described the twins as having slept. It was evident that they had been sleeping on the sofa and in rockers, both positions being contrary to the guidance that the safest place for a child to sleep is on its back in its own cot.
 K and S were admitted to hospital for full investigation. No cause for concern was found. With the agreement of the parents, they were discharged into foster care.
 L’s death has been medically investigated by the following consultant specialists:
Dr Alison Armour, pathologist and Home Office Pathologist Dr Melanie Newbould, paediatric histopathologist Professor Tony Freemont, histopathologist Dr Peter Cooper, forensic pathologist and Home Office Pathologist Professor Timothy David, paediatrician The post mortem examination
 This was performed on 13 September by Dr Alison Armour and Dr Melanie Newbould.
 Dr Armour’s report noted intra-alveolar haemorrhage (bleeding into the lungs). The findings are present microscopically in every slide examined and vary in density from 20-30%, which is regarded as moderate. There was no evidence of established bronchopneumonia or bronchilitis. There was no evidence of inhalation or aspiration of stomach contents. Perl stain was negative. Hypostasis (pooling of blood after death) was present on L’s back. There were no laryngeal polyps, oedema or cysts in the upper airways to account for obstruction, nor any marks of violence or asphyxial changes such as petechial haemorrhage to support mechanical obstruction. No petechiae or dried blood in L’s nose were noted.
 Dr Newbould attempted to identify natural disease that may have led to L’s death. L’s weight was on the 50th centile after correction for prematurity. She was described as well-nourished and had undergone good growth following her discharge from hospital. Her stomach contained a small amount of green mucoid material only, a few flecks of curdled milk were identified.
 Sections of the bone from the wrist and ulna were sent to Professor Freemont for analysis and were found to be essentially normal. No evidence of severe bacterial infection was identified. There was no evidence of biochemical upset, such as renal failure, to suggest a chronic illness.
 Metabolic disease can sometimes cause sudden and unexpected death. The cause of death at any time of life can be undetectable at post mortem because it is caused by a physiological factor, such as cardiac arrhythmia or seizure disorder, neither of which is detectable. It is a cause that can never be excluded, but there are usually preceding symptoms, none of which were reported for L. The absence of symptoms and the results of the metabolic analysis of the post mortem tissue led Dr Newbould to conclude that there was no evidence that L had an underlying metabolic disease that might have led to her death.
 Dr Newbould identifies several environmental factors known to be associated with an increase in sudden and unexpected infant death in infancy. One is parental smoking and the other is an atypical sleeping environment. Sleeping on a settee, chair or with an adult in an adult-sized bed are all associated with a higher incidence.
 Following the post mortem, the cause of L’s death was given as unascertained. The parents have not been formally interviewed by the police, who found no basis for doing so.
The Core Assessment
 This was completed by the social worker Ms P on 30 October. It records that a few days before L’s death the mother had felt unwell and the twins were sleeping downstairs on the sofa rather than in their cot in the parents’ bedroom. Overall the assessment of the parents was very positive, though some concern was expressed about the state of the home and about sleeping and feeding practices. In its submissions, the local authority contends that the parents made poor decisions that amounted to a hazardous situation in the home. It argues that the untypical state of the home suggests factors at play that combined to compromise the care of children. The parents' resistance to criticism is noted.
 The social worker read through her assessment, which was slightly late, with the parents. It is agreed that at one point the father queried where she had got information about the events of the night of 11 September. The parents say that he specifically challenged the timings but that the mother told him to desist; Ms P does not agree. I do not find it necessary to prefer one account to the other: it is possible that each understood what was being said in their own way. What is however clear is that by early November, the parents had become aware of Dr Armour’s opinions.
L was almost certainly already dead when the paramedics first attended to her. The experts are unanimous that the cause of her death is medically unknown.
 In Dr Armour’s view the bleeding into the lungs remains unexplained. She notes that suffocation of a child with a soft object like a pillow can leave no sign whatsoever. She expressed concern at the rapidity of the child’s death, in that she was noted to be awake at 5.50 a.m. but moribund at 7.30 a.m. A further concern for Dr Armour was K’s unexplained skull fracture.
 In Dr Cooper’s opinion, there is nothing to prove or suggest that L died from trauma. Theoretically she may have died as a result of suffocation, either accidental or deliberate, as typically there are no positive findings. Dr Cooper agrees with Dr Newbould that the presence of intra-alveolar haemorrhage excludes neither a natural cause of death nor death due to upper airway obstruction. Dr Newbould opines that the presence of the intra-alveolar haemorrhage in the lung is a common terminal sign. Dr Cooper describes it as “the only slightly atypical post mortem finding”. They further advise that blood coming from L’s nose after a lengthy attempted resuscitation does not assist with a cause of death. If it was blood-stained frothy secretions that had come from L’s nose, Dr Armour has seen cases of admitted overlaying with this feature. However, she could not rule out resuscitation attempts being the cause.
 Dr Armour and Professor David consider that if frank blood had come from L’s nose it would be a very worrying sign.
 There are frequently no signs at post-mortem, for example petechiae or crush injuries, to establish cause of death as asphyxia.
 The absence of a history from the parents of L choking on a feed renders it unlikely that this was the cause of her death. Nothing was blocking L’s airways at post-mortem and there was no evidence of her having choked or having inhaled or aspirated the contents of her stomach.
 There is no evidence that L’s death was preceded by illness or a detectable natural process. However, heart arrhythmia or another undetectable cardiac abnormality, or prolonged QT syndrome cannot be excluded as the cause of death.
 Sudden infant death cannot be excluded as the cause of L’s death. There is a higher incidence of sudden infant death in babies who were born premature and in babies who are twins. It is very uncommon for a sudden infant death to occur in the presence of an adult carer.
 The time of death cannot be determined from the medical evidence. L having been last seen alive at about 3 am is more consistent with sudden infant death, but being alive at about 6 am and dying shortly thereafter cannot be excluded. Dr Armour attaches no significance to the paramedics’ varied observations of temperature and stiffness.
 The hypostasis on L’s body indicates her position after death and not at the time of death. It is consistent with her being found in her rocker.
 The findings at post-mortem do not exclude overlaying as the cause of death, but there is no evidence of blanching to L’s skin as a result of being crushed.
Statements by the parents
 The parents’ statements are fully set out in the document entitled "Sequence of Accounts", which it is unnecessary to repeat in detail. Essentially, their accounts were given on two specific dates. The first was on the morning of L's death and the second was in their court statements about 10 weeks later.
 Inevitably, there is a degree of variability in matters of detail. However, the descriptions of the time when L was last seen alive differ markedly between the two dates. On 11 September, as stated above, the account was that L had last been fed at 1 a.m. and was last seen alive at about 6 a.m. In November, the parents described the father attempting to feed L unsuccessfully at 3 a.m. before putting her into her rocker.
 In his court statement, the father (who had previously said that the twins fed around 1 a.m.) says that he placed the twins on the sofa sometime after the mother went to bed at between 9.30 p.m. and 11 p.m. At 3 a.m. L woke up crying. She did not need a nappy change. He made a bottle of milk which she refused and so he then added some baby food into the milk, but when he presented the bottle she would not open her mouth. He placed her in a baby rocker at the other end of the room so that her crying would not disturb S. He then went to sleep, waking at 6 a.m. to S crying. He left her crying on the couch and, without checking on L, went upstairs to get the mother to attend to S, asking her to wake him at 7 a.m. When she woke him, he went downstairs and noticed something wrong with L.
 The local authority argues that the parents have changed their account as a result of the concern expressed by Dr Armour about the rapidity of death. It points out that the earlier account is more consistent with the twins' normal feeding pattern, which involved bottles every four or five hours at most.
 The mother's case is that she was relying on the father for times and that she should not be criticised for information given when she was in deep shock. She says that it was only when she came to make a statement that she realised how important timings were. The father's case has varied between saying that he never said that L had been alive at 6 a.m. and saying that he cannot remember saying this. Both parents deny tailoring their evidence to meet Dr Armour’s concern.
The parents’ credibility
 I have carefully considered the accounts given by the parents in relation to each of the events, making full allowance for the extremely distressing circumstances that existed following L's death. Having done so, I regrettably do not find them to be witnesses whose accounts can always be accepted without question. I have already found that the account of the police officers and Nurse C is to be preferred to theirs, and that the parents did say that L had been alive at 6 a.m. or thereabouts. There is no reason why they should have made a mistake about this. Even at that time, it would have been obvious that a broadly accurate account was important, there being real worries about the health of L’s twin, S, and also that of K. I accordingly reject the parents' account on this aspect and the reason they offer for the different timing in their November statements.
 Having seen them give evidence, I am unable to differentiate between the mother and the father in terms of credibility. They are a very close couple. In relation to events on the night that L died, the father is in a better position to know what took place, but the mother was also around in the early morning to know for herself what was going on, and she has also had the opportunity to question the father. Painful though the subject was, the parents were very active in trying to seek explanations for L’s death and I do not accept that the mother would have allowed herself to remain under such a potentially important misunderstanding for over two months.
 This conclusion about the parents’ credibility does not automatically mean that they are culpable in relation to either event. The possible reasons for them to give inaccurate information must be assessed.
 In my judgment, a feature of potential relevance in this case is that the parents are acutely concerned to present themselves to the outside world as good parents. It is easy to understand how defenceless a young couple must feel in these circumstances, but these parents are notably reluctant to make concessions. Most people in their position, struck by such misfortune not once, but twice, would allow others to see their vulnerability. The mother’s repeated self-reproach at hospital ("I should have checked on her") may give a truer picture than her evidence in these proceedings, in which she firmly rejected any criticism. Her description of the sleeping arrangements as "not the greatest idea" was the furthest she was prepared to go. When one considers the obvious possibility that those sleeping arrangements might be linked in some way to L’s death, this was a strange and telling choice of words. The mother’s request to Ms G on 2 November that the father should not be told about K’s fall provides another example of this fear of criticism. As for the father, he did not accept any failures whatever. My impression is that this is couple who have lost so much that they find it very hard to accept any blemish on the image that they hold of themselves as excellent parents.
Conclusion in relation to K
 As already stated, the evidence establishes that this is likely to have been an accidental injury that occurred on the morning of admission to hospital. The question is whether the local authority has proved that the parents know how it happened. With some hesitation, I find that it has not done so and that the parents’ account of events can be accepted. Balancing the matters contained in paragraphs 33 and 34 above alongside my view of their credibility, I cannot conclude that it this more probable than not that they are concealing a known incident. The general probabilities of the situation are not underpinned by any specific feature of the evidence that disproves their account. Furthermore, I think it likely that K was not always as closely supervised as the parents suggest. The incident on 2 November is a case in point. A similar, but unwitnessed, fall by K against a hard object (to take an entirely hypothetical possibility, the marble fireplace and surround) on the morning of 14 November could have led to injury that escaped notice during the course of a hectic morning. This may mean that K, a highly active child, was being inadequately supervised, but that would not in these circumstances cross the threshold for intervention.
Conclusion in relation to L
 Again as stated, it is not suggested that L’s death was the result of any deliberate act by either parent. What is now alleged is that she died as a result of inadvertent suffocation arising from unsafe sleeping arrangements, and that the father may have covered this up by placing her in her rocker after death.
 Again, taking everything into account, I do not find this more probable than not. The medical evidence allows for, but does not demand, such a finding. The indicators for inadvertent suffocation are slight (a moderate degree of intra-alveolar haemorrhage whose significance is disputed between the pathologists) and there are a number of potential indicators that do not exist (facial or conjunctival petechiae, or blanching of the skin). The location of the hypostasis does not indicate that L was on her side or front after death, as might be the case if she had been lain on by her father or had her face in the sofa or cushions. The temptation to seek an explanation for L’s tragic death on the basis of such equivocal evidence must be resisted.
 For the father to have placed L in her rocker, knowing that she was moribund, and then gone to bed would be extremely out of character. There is no reliable evidence that he did this.
 The only other evidence of potential significance relates to the parents' changing accounts. What inferences are to be drawn from that? I acknowledge the possibility that they have put their heads together to cover up a case of overlaying, but I have in the end concluded that a more likely reason for the change in their story is to defend themselves against what was at the time an allegation extending to their having deliberately suffocated L. Accordingly, unsatisfactory though the parents' evidence has been on this issue, I do not find that it adds weight to the medical evidence. Specifically, I am not satisfied that the parents have changed their accounts in order to cover up what happened to L. The parents say that they do not know how L died and the local authority has not proved that they do.
 In the end, despite the close attention given to the issue, it is not possible to reliably identify the time of L’s death or to say which of the parents’ accounts of the last time she was seen alive is more likely to be true. One might assume that they had moved from an accurate account (6.a.m.) to an inaccurate one (3 a.m.) but it is in my view just as likely that they gave an inaccurate account of her being alive at 6 a.m. in order to shield themselves from any suggestion that they had neglected L, and later reverted to a more accurate account when they saw that the first account was leading them into difficulty. The irony, as Ms Irving QC puts it, is that the medical analysis is not determined by which account is correct.
 My conclusion, echoing Dr Cooper, is that we simply do not know when or why L died. Nor has it been proved that either of the parents knows. In such circumstances, a finding against the parents would be based on speculation or suspicion, which would be impermissible and wrong. The sleeping arrangements that were made for such small babies were obviously unwise, but that does not cross the threshold for intervention.
 I therefore find that the grounds for making care orders do not exist and I dismiss the local authority's applications. In consequence, the children will return to the care of their parents in a manner to be agreed between the parents and the social workers.
 I conclude by recording that the local authority has acted entirely properly in bringing these proceedings and in the manner in which it has scrutinised the evidence. On the facts of the case, it had no alternative, and the dismissal of its applications implies no criticism whatever.
 These proceedings are concluded.