Berkeley Lifford Hall Accountancy ServicesIQ Legal TrainingHousing Law WeekAlphabiolabs

Home > Articles > 2014 archive

From "Shaken Baby Syndrome" to "Non-Accidental Head Injury" – The Continuing Research and the Law

David Bedingfield of 4 Paper Buildings charts the recent history of scientific research into serious non-accidental head injuries suffered by babies and the response of the family and criminal courts in England and Wales.

David Bedingfield, barrister, 4 Paper Buildings

David Bedingfield, barrister, 4 Paper Buildings

Non-accidental head injury remains the most common cause of fatal child maltreatment. An analysis by the NSPCC released in November, 2013, puts the number of babies under one suffering from a serious non-accidental head injury in England and Wales each year at approximately 24 per 100,000.1   At least half of the survivors have significant neurological impairments.

Few doubt that most of these children were victims of parental abuse.

The difficulty, of course, is that no one doubts that in the very recent past physicians (and judges and jurors) wrongly found that parents or other carers had caused significant head injuries to children by shaking. Physicians believed that diagnosis of certain injuries to an infant's head almost inevitably meant that the child had been shaken by the child's last carer. Since at least the turn of this century, however, medical and bio-mechanical experts have begun to offer proof that this "certain diagnosis" was in fact all too uncertain.

The Medical Research
The history of the evolution of the scientific research – and the legal responses to that new research – is brilliantly set out by the U.S. academic Deborah Tuerkheimer in an article in the Washington University Law Review that should be mandatory reading for any child law practitioner handling cases involving allegations of shaken babies.2

Professor Tuerkheimer first sets out the research regarding infant head injuries that led to convictions for murder of children routinely being handed down by juries on both sides of the Atlantic, despite no eye-witness evidence of the crime and repeated protestations of innocence by heretofore blameless parents. All of these convictions were based solely on scientific evidence by experts, who explained the importance of what has now been called the "classic triad" of symptoms in shaken baby cases: 1) retinal haemorrhages (bleeding of the inside back surface of the eye); 2) subdural haemorrhages (bleeding between the hard outer layer and the spongy membranes that surround the brain); and 3) cerebral edema (swelling of the brain). The presence of those three signs had been considered – certainly prior to 2001 – as pathognomic – or exclusively characteristic – of SBS.

The presence of the "classic triad" has in the past been taken to mean that a baby had been shaken hard enough to produce what were conceptualised as whiplash forces.3 The application of ". . . rotational acceleration and deceleration forces to the infant's head causes the brain to rotate in the skull. Abrupt deceleration allows continuing brain rotation until bridging veins are stretched and ruptured, causing a thin layer of subdural haemorrhage on the surface of the brain."4The scientific underpinnings of SBS, as Professor Tuerkheimer notes, have crumbled over the past decade. The medical establishment now accepts that the term "shaken baby" is misleading, and non-accidental head injury, or Shaken Impact Syndrome, have been suggested as replacements.

Professor Tuerkheimer sets out the logical fallacy that underlies much pre-1999 research regarding Shaken Baby Syndrome: researchers selected cases by the presence of the very clinical findings and test results they sought to validate as diagnostic. Not surprisingly, as Professor Tuerkhimer notes, the studies tended to find their own case selection criteria pathognomonic of SBS. The circularity of the logic is represented by the equation: "SBS = SDH + RH [inclusion criteria], therefore SDH + RH = SBS."

The difficulties surrounding the "admissions" by carers that they shook the child are obvious: Did they shake the child after discovering the child was unconscious? How hard did they shake the child? How long? Did any symptoms precede the shaking?

Experts also began to consider the expertise of biomechanical engineers and pathologists. The shaking that had been thought to be required, these studies seemed to show, would certainly have injured the cervical spinal cord and neck. In many cases, no injuries to the neck or spinal cord were shown. The conclusion, at least by some experts, was that the child could not have been shaken to that degree if no neck injuries were shown.5Advocates in cases where no injuries are revealed to the neck and spinal column must question the experts carefully with regard to the degree of force that would be required to cause the classic triad. Some experts, it would seem, have now backtracked on their previous contention that only extremely rough handling could cause these injuries. It is also right that some new research showed that relatively short-distance falls may cause fatal head injuries that look like SBS-type injuries.6

In the mid to late 1990's, magnetic resonance imaging (MRI) revolutionized the field of radiology. MRI enabled a far more detailed assessment of the pattern and extent and timing of injuries than had its precursor, computed tomography (CT).  These new radiological findings revealed the presence of triad symptoms in cases where it had been accepted that the cause was not shaking: for example, in cases of accidental injury, or in cases where medical disorders such as autoimmune deficiency had been diagnosed. These conditions could be seen by less subtle and nuanced diagnosis to be NAI.

The importance of the Louise Woodward case also should not be underestimated.7The British au pair's case was perceived as one that revealed the divide in the scientific community. Scientists challenging the SBS dogma, according to Professor Tuerkheimer, ". . .  emerged as a significant force in terms of numbers as well as influence." (See 2009 Wash LR, p. 15.)  Defenders of the SBS diagnosis began in the face of these arguments to adapt their analysis in subtle but important ways. No longer were physicians willing to state with certainty that the constellation of symptoms must in every case indicate that a child was abused. In particular, new studies generated new explanations for the presence of subdural haematomas and retinal haemorrhages.8Most physicians by this time began to concede that the triad was not necessarily induced by shaking. A differential diagnosis must always be considered. This was a dramatic evolution in mainstream scientific thinking.

Professor Tuerkheimer points out other difficulties with the SBS diagnosis. In particular, it had been previously thought that the perpetrator of abuse was necessarily the person with the infant immediately prior to the loss of consciousness. Studies have since shown that children suffering fatal head injuries may be lucid from more than 72 hours before death.9 This research, of course, makes it difficult if not impossible to pinpoint the time of the injury and the identity of the perpetrator.

The Law's Response
As a result of these new medical advances, the Court of Appeal (Crim) in 2005 considered four cases on appeal. The cases are reported as  R v Lorraine Harris, Raymond Charles Rock, Alan Barry Cherry and Michael Faulder [2005] EWCA Crim 1980, [2006] Cr App R 5; [2008] 2 FLR 412, CA. Though the cases concern the criminal standard of proof and the verdicts of juries, the Court of Appeal's judgment is required reading for practitiners in family proceedings handling non-accidental head injury cases.

The Court of Appeal in Harris held that no longer would the classic triad "automatically or necessarily" lead to a conclusion that the infant had been shaken. The appellants had all appealed in separate cases against their convictions for either manslaughter or murder or inflicting grievous bodily harm. Each case involved SBS. In each case, at the time immediately before each of the victims became seriously unwell, each had been in the sole care of the respective defendant.  All had been convicted after jury trials. Each appealed on the basis of fresh evidence. The argument was essentially this: The previously accepted hypothesis in cases of SBS depended on findings of the classic triad: brain swelling and loss of brain function; subdural haemorrhages; retinal haemorrhages. Between 2000 and 2004 a team of doctors led by the neuropathologist Dr Geddes had produced three papers setting out the results of their research into the triad. In the third paper (known as "Geddes III") the team put forward a new hypothesis: This hypothesis (known as the "unified hypothesis") challenged the infallibility of the triad. The new hypothesis did not seek to show that the triad was inconsistent with NAHI, but it did seek to show that it was not diagnostic. At least prior to the trial, Dr Geddes and her team contended that it was likely that the brain swelling itself caused the subdural haemorrhages and retinal haemorrhages.

Dr Geddes accepted in cross-examination during the hearing, however, that the unified hypothesis was not likely correct, and that the brain swelling did not likely cause the subdural haemorrhages or retinal haemorrhages. Therefore the unified hypothesis was not a credible alternative explanation for the appearance of the classic triad.

But that was not the end of the Appellate Court's examination of the issue. The Court noted that there remained a body of medical opinion which did not accept that the classic triad was inevitably diagnostic of abuse. Common sense suggested that the more severe the injuries the more probable it was that they would have been caused by greater force than mere "rough handling." 

In the case involving the defendant Harris, her four-month old son had collapsed and died. The evidence was solely the classic triad, without other medical evidence of injury or rough handling. The new evidence, according to the Court of Appeal, threw doubt on the significance of the subdural haemorrhages that were found. It threw doubt on the evidence of injuries to the brain.  The finding of retinal haemorrhages was powerful supporting evidence of shaking, but on its own was not diagnostic of shaking. Accordingly, the conviction was quashed.

In the case of the defendant Rock, however, the triad did not stand alone. There was no dispute that the defendant had shaken his partner's 13-month-old daughter and no dispute that she had suffered an impact to the back of her head. Rock's explanation could not account for the injuries. The Court of Appeal did, however, reduce the defendant's conviction from murder to manslaughter.

The defendant Cherry was convicted of manslaughter after the death of the 21-month-old daughter of his partner. He contended the child fell from a chair some 6 to 8 inches off the floor. Two elements of the triad were present. There were two separate sites of scalp bruising. The conviction was upheld.

The defendant Faulder's seven-week-old son suffered a non-fatal injury to his head. The defendant contended the child fell from a chair. The prosecution first contended, based solely on the existence of the triad, that the child had been shaken. The Crown then changed its case to contend the child was likely beaten several different times, citing the number of bruises the child had. The Court of Appeal, noting the potentially credible alternatives put forward by the defence, held that the conviction was unsafe.

After Harris, the then Attorney General Lord Goldsmith conducted a seven-month review of 88 SBS cases, including guilty verdicts and guilty pleas. See The Rt Hon The Lord Goldsmith QC, The Review of Infant Death Cases: Addendum to Report Shaken Baby Syndrome (HMSO 2006).  The AG's review was criticised by Professor Tuerkheimer and others, primarily because Lord Goldsmith considered admissions by defendants as corroborative of the diagnosis. Lord Goldsmith also considered the presence of chronic subdural haematomas as corroborative evidence. Professor Tuerkheimer contends research shows that the presence of subdural haematomas to have limited value as corroborative evidence of shaking. Three of the cases reviewed by Lord Goldsmith (3.4% of the total reviewed) were seen to be unsafe and gave rise to referral to Criminal Court of Appeal. As Professor Tuerkheimer points out, Lord Goldsmith's systematic review and the Court of Appeal's decision in Harris appreciably altered the course of SBS prosecutions.10

In March, 2011, as a result of the review by Lord Goldsmith as well as the new research findings, the Crown Prosecution Service issued new guidance with regard to bringing what the CPS now refers to as "non-accidental head injury" cases:

  1. NAHI cases will usually be diagnosed by the triad of internal head injuries set out above;
  2. To prove a NAHI case, the Crown will usually require the triad of injuries PLUS supporting evidence;
  3. The "unified hypothesis" is a theory used by the defence experts to challenge the triad diagnosis; however the theory has not been endorsed by the Court of Appeal;
  4. CPS policy is to resist challenges to the triad diagnosis based on the unified hypothesis;
  5. The defence may also try to introduce bio-mechanical evidence; to do so, the defence would be required to demonstrate its relevance to the case;
  6. Expert evidence must be dealt with in accordance with the CPRs;
  7. The Strategy and Policy Directorate must be informed of NAHI cases in which these issues arise.

The Latest Research – a Middle Way Forward?
Practitioners should also be aware of new research by the Canadian pathologist Evan Matshes, published in July, 2011, by Academic Forensic Pathology, the journal of the US National Association of Medical Examiners. Matshes's research shows how death from shaking could in fact occur, not because of the traditional triad of injuries to the brain, but because of injuries to the child's spinal column that directly impacted the child's ability to breathe.  The new findings, in fact, split a lot of the difference between the warring camps on Shaken Baby Syndrome. In investigating the deaths of 35 children, Matshes did autopsies in a new way. The usual practice is to dissect only part of the spinal column. Matshes instead dissected the spine down through the neck and into the nerve roots. In the 12 babies whose history showed evidence of injury from hyperflexion – severe whiplash, from either shaking of from a car accident – he found bleeding in the nerve roots of the part of the spinal column at C3, C4 and C5. Matshes then compared this to the spinal columns of a group of 23 children for whom there had been no solid evidence of an injury from whiplash. (This group in the main died from smothering or Sudden Infant Death Syndrome.) Only one baby in this group had bleeding in the same C3, C4 and C5 region. That child's history, interestingly, made shaking a distinct possibility, though it was denied by the carer.

The key to the finding is that C3, C4 and C5 control the child's diaphragm. Babies depend on their diaphragms to breathe, much more so than older children or adults. As the New York Times reporter Emily Bazelon points out in her story reporting Dr Matshes's findings, Matshes's work, if proved correct, could be the missing piece of the puzzle: the causal mechanism that the biomechanical experiments have not accounted for.11The New York Times reporter approached four experts, two supporters of the traditional shaken baby diagnosis and two critics. They all agreed the paper, while based on a small sample, pointed to a new area worthy of more research. Dr Waney Squier, a paediatric neuropathologist who has often given evidence in the UK in these cases, told the Times reporter that "It is now terribly important to look at the neck." Squier also pointed to a methodological weakness in the paper: the study was not double blind. The researchers, in other words, knew the histories of the babies when they conducted their research.12 Pathologists doing autopsies, however, are ethically bound to know the subject's history. There can be no randomised controlled double-blinded studies in forensic pathology, for obvious reasons.

The paediatric neurosurgeon Norman Guthkelch, whose research 40 years ago was one of the most important early pieces of research that tended to connect head injuries in young children to violence, has agreed that Matshes's study is a vitally important new contribution to the debate. Dr Guthkelch told the US public broadcaster National Public Radio in the summer of 2011 that he worries that doctors and other medical experts are too quick to diagnose shaken baby syndrome, without considering other possibilities.13

ne difficulty with the study is that it is unlikely that in every case involving suspected NAI these types of autopsies can be undertaken. There are time-consuming extra steps involved. The spinal column has to be placed in formaldehyde for up to a month in order for the bone to soften before the pathologist can begin the testing process.

The Current Approach of Family Judges
Critics of the new research argue that shaking is still the most likely explanation for rhetinal haemorrhaging and subdural haematomas.14 That said, it is clear that judges look at all other possible explanations before making findings.15  All courts now consider the various differential diagnoses set out by Professor Timothy David in his seminal article: "Non-accidental Head Injury: The Evidence."16 Several differential diagnoses must now be considered: 1) congenital malformation; "metabolic disorders; 3) haematological diseases; 4) infectious diseases; 5) autoimmune conditions.

But more importantly, courts must consider, as Mr Justice Hedley did in the case of R (A Child) [2010] EWHC 1715 (Fam), whether the cause of the injury is simply "not presently known or understood". In that case, Hedley J cited Moses LJ in the criminal appeal of Henderson and others [2010] EWCA Crim 1269, CA:

"There remains a temptation to believe that it is always possible to identify the cause of injury to a child. Where the prosecution is able, by advancing an array of experts, to identify a non-accidental injury and the defence can identify no alternative causes, it is tempting to conclude that the prosecution has proved its case. Such 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."

In the case before Hedley J, there were three possible causes of the subdural haematomas found in the child: firstly, a peri-natal event, although there was nothing of note in this case to specifically point to this; secondly, a non-accidental, inflicted head injury; and thirdly, a cause that was not known or understood. Hedley J stated that a conclusion of "unknown etiology" was not a professional or a forensic failure; it simply recognises that there is much we do not know and that it is dangerous and wrong to infer non-accidental injury merely from the absence of any other understood mechanisms. After reviewing all of the evidence before him, Hedley J was able to make a finding that the cause of the subdural haematomas was one of unknown etiology. In that case, however, it must noted that there were no retinal haemorrhages, and no encepalapathy or brain disturbance. The child, however, suffered from a fracture of the leg. Hedley J, after examining the evidence, found that the fracture was caused by accident.

Another case that is mandatory reading for any advocate appearing in a head injury case is Baker J's careful judgment in the case of Re JS.17 This was ct-fa fact-fearing within care proceedings. The local authority sought findings that the head injuries suffered by the child were inflicted by his parents when the child was 15 weeks old.

Baker J heard evidence over 15 days from six leading experts in their fields. Baker J gave a summary of the up-to-date research in this area, in particular with regard to subdural haematomas and retinal haemorrhages. After considering the evidence, Baker J found that the father was responsible for the injuries by shaking the child, and that the mother had failed to protect the child from the father and therefore contributed to the harm suffered by the child.

Practitioners will take particular note of two reports cited by Baker J. The first, Intracranial Haemorrhage in Asymptomatic Neonates: Prevalence on MR Images and Relationship to Obstetric and Neonatal Risk Factors18, reviewed three studies (by Looney and others; Whitby and others, and Rooks and others) that showed  subdural haematomas occur much more frequently at birth than was previously recognised. The Rooks paper indicated that 46 per cent of neonates had subdural haematomas seen by MRI within 72 hours of delivery. Most of these subdural haematomas had resolved within one month. All had resolved by three months. In the trial before Baker J, the experts refused to rule out the possibility, however, of chronic bleeding that persisted past three months.

With regard to retinal haemorrhages, a report by the Welsh Child Protection Systematic Review Group was cited by the experts. This report, by Maguire and others, is entitled Which Clinical Features Distinguish Inflicted From Non-inflicted Brain Injury—A Systematic Review.19 The review showed that apnoea and retinal haemorrhages are present in a high percentage of inflicted brain injury cases. As the expert reported to Baker J (at para 62):

". . . non-abusive head injury was a rare cause of retinal haemorrhage and, when present, compared to those in abusive head trauma, was more frequently unilateral, fewer in number and restricted to the posterior pole. Retinal bleeding is much more likely to be found in cases of abusive head trauma with the retinal bleeding described as multi-layered, extensive and extending to the periphery, but. . . can occur in non-abusive head injury where they are more likely to be unilateral, non-extensive and restricted to the posterior pole."

Baker J was careful to note, however, that while the appearance of retinal haemorrhages may be associated with abusive or non-abusive head trauma, the appearance of the retinal haemorrhages is not diagnostic of a particular cause.

Practitioners involved in any fact-finding hearing where non-accidental injuries have been alleged should follow the 10 Legal Commandments given by Mr Justice Baker in the case of Re JS [2012] EWHC 1370, at paragraphs 36-45.

  1. The burden of proof lies with the local authority.
  2. The standard of proof is the balance of probabilities.
  3. Findings of fact must be based on evidence, not speculation or suspicion.20
  4. When considering cases of suspected child abuse, the court must consider each piece of evidence in the context of all other evidence. A judge must view the totality of the evidence in order to come to the conclusion whether the case has been made out to the appropriate standard of proof.21
  5. In serious non-accidental cases, including head injury cases, the opinion of medical experts must be considered in the context of all other evidence. The court must weigh up expert evidence against other evidence. There may be cases where a court determines that the weight of the evidence is at variance from that reached by medical experts.
  6. In assessing the expert evidence, the court is assessing the evidence of a group of specialists, each bringing a different expertise to bear on the issue. Each expert must keep within the bounds of his or her expertise. The expert must defer, where appropriate, to the expertise of others.22
  7. The evidence of the parents or other carers is of the utmost importance. Credibility and reliability are key issues.
  8. Witnesses often tell lies. The court must bear in mind that a witness may lie for many reasons, such as shame, misplaced loyalty, panic, fear and distress. The fact a witness has lied about some matters does not mean he or she has lied about everything.23
  9. A court must take into account the possibility that the cause of an injury or condition is simply unknown. This does not affect the burden or the standard of proof. It is simply a factor to be taken into account in deciding whether the causation advanced by the party holding the burden of proof is established on the balance of probabilities.24
  10. When seeking to identify the perpetrators of NAIs, the test regarding 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. To make a finding that a particular person was the perpetrator, the court must be satisfied on a balance of probabilities. It is desirable, where possible, for the perpetrator to be identified. But a judge should not strain to do so.25

Mrs Justice Theis's judgment in the case of London Borough of Islington v Al Alas and Wray [2012] EWHC 865 (Fam) must also be parsed carefully by any advocate appearing in an SBS case. The facts were as follows: The child Jayden Al Alas Wray was born on 7.3.2009 and died little more than four months later, on 25.7.2009. The local authority contended this was a straightforward SBS case: the child suffered from the classic triad of injuries, and also suffered from fractures and multiple sites and of varying ages. After a hearing that lasted some six weeks, Theis J conclusively found all allegations not proved. She found that all of the fractures were caused by rickets, which had been misdiagnosed by the child's treating physicians. The death was attributed to a constellation of benign causes:  1) severe vitamin D deficiency; 2) ongoing seizures leading to raised intracranial pressure, retinal haemorrhages, and subdural haematoma, culminating in hypoxic ischemia brain injury and death.

Theis J also found that there were significant intervening events between admission to hospital and the child's death that played a role in causation, including what the judge termed "sub-optimal" medical treatment by the University College London Hospital team.

Both Great Ormond Street and UCL, which had treated the child, attributed all his injuries to NAI. GOSH did not pick up radiological evident signs of rickets, and specifically excluded any underlying metabolic bone disease to account for the fractures. This clinical misjudgement led to the erroneous conclusion that the skull fracture occurred at the same time as the child's collapse and was therefore linked to the cause of death. As counsel who conducted the case for the mother stated in an article published in the June edition of Family Law, "GOSH could not have been more wrong."26

The Court, after hearing argument, decided it was not appropriate to shield the identity of the child and the treating physicians. All experts and treating physicians were therefore named in the judgment.

One factor that will always be considered by the court (under the category of "other evidence") is the reliability or credibility of the evidence given by those caring for the child at or near the time the injury is noticed. An example is the recent decision by Mr Justice Keehan. See A Local Authority v DB, RB and SM [2013] EWHC 4066 (Fam).

In that case the child, aged 19 months, suffered the following injuries, some combination of which were fatal:

a) a large space occupying right sided subdural haematoma, causing a midline shift;
b) widespread cerebral infarction with secondary brain swelling;
c) bilateral retinal haemorrhages, affecting all parts of the retina;
d) bilateral perimacular folds;
e) extensive haemorrhage into the orbital connective tissues;
f) extensive haemorrhage into the collagenous dual sheath of both optic nerves;
g) extensive haemorrhage in the subdural space around both optic nerves;
h) marked papilloedema in both eyes.

Keehan J also pointed out what was NOT present:

a) no skull fractures were identified;
b) no rib fractures or fractures of the long bones were seen;
c) there was no swelling or bruising on the right side of the head;
d) there were no suspicious marks or bruises found on the child's body;
e) there was no damage to the child's internal organs caused by a grip or a shake;
f) there was no contusion to the brain;
g) the subdural bleeding was unilateral and not bilateral and multi-focal; and
h) the child (19 months of age) was outside the usual age range at which children ordinarily sustain injury by shaking. 

One area of contention involved the question of whether an arteriovenous malformation in the child had bled spontaneously and had destroyed itself in the course of the haemorrhage. The paediatric neurosurgeon Mr Richards agreed that one "possible" but rare cause of the subdural bleeding in the child could be spontaneous bleeding of this sort, but stated in his written report that such cases are extremely rare. In the course of the experts' meeting, the forensic pathologist Dr Cary was reported to have said this: "I would just like to make the point though from earlier discussion I think it was felt that a spontaneous bleed was so unlikely to be completely unrealistic in these circumstances."

Mr Richards agreed. In Mr Richards' view, ". . . it [the spontaneous bleed from an arteriovenous malformation] was the only potential alternative [explanation] for acute disaster like this to trauma, and I think it is extremely highly unlikely to the point of being virtually implausible." (See para 76.)

Mr Richards explained that the vast majority of arteriovenous malformations bleed into the substance of the brain. No bleeding was found on post mortem exam of the substance of the child's brain. It was also extremely rare for an arteriovenous malformation to rupture spontaneously and extremely rare for the evidence of the existence of an arteriovenous malformation to disappear. It is also right, Mr Richards stated, that arteriovenous malformations would not cause the retinal bleeding found in the child's eyes.

In all fatal cases of suspected NAHI, an opthalmic pathologist will be asked to review the evidence. The Royal College of Paediatrics and Child Health and the Royal College of Ophthalmologists have recently issued updated guidance regarding the appearance of a child's retina after suspected abusive trauma.

Only in the last two years have paediatric and forensic pathologists begun providing to ophthalmic pathologists the orbital connective tissue surrounding a child's eyes. There is a difference between the anatomy and physiology of the eye compared with the orbital connective tissue, and it is clear much research remains to be done.

In the case before Keehan J, leading counsel for the father took the consultant ophthalmic pathologist Dr McCarthy through the various differential diagnoses that must be considered. Dr McCarthy in this case concluded that the evidence pointed (in his view conclusively) to an inflicted injury. The extent of the retinal haemorrhages seen in this child, the fact they were bilateral, the fact there were also haemorrhages in the optic and nerve sheath, in the orbits of the eye and in the orbital connective tissues – all of this taken together  meant the retinal haemorrhages were not likely caused by raised intra cranial pressure. In Dr McCarthy's view the retinal haemorrhages were caused by trauma.

But at least as important as the medical evidence for Keehan J in this case was his consideration of the evidence of the mother and father. Both were seen to be kind, loving and considerate people. This was not a family with a history of violence or drug abuse. The parents did not present with the sort of chaotic lifestyle that courts often see in NAHI cases. The contact between mother, father and the other children of the family was seen to be of a very high quality.

On the other hand, the mother clearly drank more alcohol than she was willing to acknowledge. And more important, the mother and father were willing to lie about that. It was also right that the father lied about his whereabouts on the night of the injury, and had given various and inconsistent accounts about how often he was up feeding the couple's new born child (not the child who suffered the fatal injuries). The father also gave inconsistent accounts of how the child looked when he discovered the child lying face down in the child's cot.  Both parents were found by the judge to have been untruthful about their movements on the day before and the day of the incident. The fact that five text messages between them had been deleted from their phones also raised questions in the judge's mind.

Keehan J gave himself the direction required of Crown Court judges when directing a jury in similar situations: the fact that a defendant has been found to have lied about something does not necessarily mean the defendant is lying about whether he or she committed the criminal act at issue. People may lie for a variety of reasons not connected with the alleged crime that was committed. Judges and juries must consider this when seeking to determine whether the defendant (or respondent) in fact caused the injuries complained of.

But for Keehan J, the parents' unsatisfactory evidence made the medical evidence even more compelling. It is right, as Hedley J pointed out in Re R (A Child), that courts must consider whether the medical evidence should simply be disregarded. Some head injuries in infants are simply unexplainable, and we are not at a point where the medical evidence can simply without any question rule out all possible causes save an inflicted trauma. But where the parents have been found to have lied about crucial matters, the court must consider carefully whether those lies are in fact positive evidence that each parent knows more than they are revealing. Keehan J, after making findings regarding the parents' evidence, had no difficulty in also finding that either the mother or the father caused the injury, that the injuries would have been obvious to the non-caring parent, and that both were colluding with each other to hide the truth from the court.

[1] See Statistics on non-accidental head injuries in infants.
[2] See 'The Next Innocence Project: Shaken Baby Syndrome and the Criminal Courts 87 Washington University Law Review p.1 (2009). An abstract is here. The findings are discussed in Professor Tuerkheimer's forthcoming work, "Flawed Convictions: 'Shaken Baby Syndrome' and the Inertia of Injustice" (Oxford University Press, April 2014). Details are here.
[3] See John Caffey, 'On the Theory and Practice of Shaking Infants' 124 American Journal of Diseases of Children 161, 172.
[4] See Harding,, 'Shaken Baby Syndrome' 328 British Medical Journal 720 (2004).
[5] See Bandak, 'Shaken Baby Syndrome: A Biomechanics Analysis of Injury Mechanism' 151 Forensic Sci Int'l 71 (2005).
[6] See Plunkett, 'Fatal Pediatric Head Injuries Caused by Short Distance Falls' 22 Am Journal Forensic Med and Pathology 1, 10 (2001). Note that many experts disagree with the study. See 328 Brit Med Journal 1316 (2004).
[7] See Commonwealth v Woodward, No CRIM 97-0433, 1997 WL 694119 (Mass Sup Ct Nov 10 1997). In that case Ms Woodward, a young woman working as an au pair for a couple in Massachusetts, telephoned emergency services to report that she could not rouse eight-month-old Matthew Espeen from his nap. Doctors found massive intracranial bleeding, brain swelling, and a retinal haemorrhage. Matthew died from his injuries. The prosecution relied almost entirely on medical evidence. The defence team challenged the evidence in ways this type of evidence had not previously been challenged. The defence lawyer Barry Scheck hired medical experts at the rate of thousands of dollars a day. One expert gave evidence that the fracture likely occurred three weeks previously. The fatal bleeding could have been unleashed by just a slight jar. The fatal haemorrhage, argued Scheck, was caused by a re-bleed of a chronic brain clot resulting from a previously undetected injury. The jury convicted Ms. Woodward of murder. The trial judge reduced the verdict to involuntary manslaughter and sentenced Ms. Woodward to time served, meaning she was released. In his order, the judge stated that one rational view of the evidence was that the re-bleed occurred upon "rough" handling by Ms. Woodward. She continued to deny that this occurred. See Rosenberg and Thomas, "I didn't do anything," Newsweek, Nov 10, 1997, http:
[8] See Cohen and Scheimberg: "Evidence of Occurrence of Intra-dural and Subdural Haemorrhage in the Perinatal and Neonatal Period in the Context of Hypoxic Ischemic Encephalopathy, 12 Pediatric Developmental Pathology 169 (2009) Lantz, et al, Perimacular Retinal Folds From Chjildhood Head Trauma, 328 Crit Med J 754 (2004).:
[9] Gilland, 'Interval Duration Between Injury and Severe Symptoms in Non accidental Head Trauma In Infants and Young Children' 43 Journal of Forensic Science 723 (1998); Arbogast, et al, 'Initial Neurologic Presentation in Young Children Sustaining Inflicted and Unintentional Fatal Head Injuries' 116 Pediatrics 1608,(2005); Robert Huntington, Letter, 'Symptoms Following Head Injury' 23 Am J Forensic Medicine and Pathology 105 (2002). Dr Huntington is a pathologist. He had testified prior to 2002 that it was likely a child would have been injured "within two hours" of being seen to be unwell. Dr Huntington, after considering the latest research regarding lucid intervals of children who had been abused, has now changed his opinion. See Tuerkheimer, supra, p. 19, footnote 115.
[10] See Sam Lister, Q and A: Shaken Baby Syndrome, Times On Line, 14.2.2006;
[11] See www.//6thfloor.blogs.nytimes.com2011/07/05new-evidence-on-shaking.
[12] Dr Waney Squier's work is this area has been questioned, however, by trial judges in the United Kingdom, in particular by Mrs. Justice King in the case of A Local Authority v S [2009] EWHC 2115 (Fam). In that case, the child was 13 weeks of age when presented to A and E with the classic triad of symptoms. King J was asked to make adverse findings regarding the evidence of Dr Squier. King J noted noted that Dr Squier regards findings of trauma as "very important." In the absence of evidence of trauma, Dr Squier contends that shaking alone "could" cause the triad, but the force required would likely be impossible for an ordinary person to generate. Dr Squier contends that choking causes hypoxia, which is then responsible for the subdural haemorrhages and retinal haemorrhages. Mrs. Justice King noted that Dr Squier's view is a legitimate one, and an appropriate line of research. But King J also said this: "All agreed that much remains unknown about SBS and the triad. It is essential, however, that Dr Squier and others engaged on such research avoid becoming a zealot with the consequence that scientific rigour is lost or sacrificed." (See Para 246 of King J's judgement.) Dr Squier's use of research articles in this instance was criticised by King J, who labelled it "disingenuous." King J therefore refused to accept Dr Squier's reasoning and rejected her conclusions, and found that the expert had permitted her convictions to lead her analysis. (See para 285.) Dr Squier responded to the judgment, and to a review of that judgment, in an open letter to Family Law Week which can be found here.
[13] See
[14] See Chadwick, "Annual Risk of Death Resulting from Short Falls Among Young Children: Less than 1 in 1 million" 121 Pediatrics 1213 (2008)
[15] See, for example, Lancashire County Council v D and E [2008] EWHC, 832, [2010] 2 FLR 196; R v Henderson, Butler and Oyediran [2010] EWCA Crim 1269, [2010] 1 FLR 547.
[16] See (2008) Journal of Paediatric Radiology 370-377.
[17] [2012] EWHC 1370, Fam.
[18] [2007] Radiology 242 (February, 2007).
[19] 94 Archives of Diseases in Childhood 2009, p. 860.
[20] See Munby LJ in Re A (A Child) (Fact-finding hearing: Speculation) [2011] EWCA Civ 12.
[21] See Butler-Sloss P in Re T [2004] EWCA Civ 558, [2004] 2 FLR 838, at para 33.
[22] See King J in Re S [2009] EWHC 2115 (Fam).
[23] This is a direction given to criminal juries. See R v Lucas [1981] QB 720.  It applies to all fact-finding decisions made by judges in all family proceedings.
[24] See Re R (Care Proceedings: Causation) [2011 EWHC 1715 (Fam) per Hedley J. See also R v Henderson and others [2010] EWCA Crim 1219, where the Criminal Court of Appeal noted again that it is not always possible to identify the cause of an injury to a child.
[25] See Re D (Children) [2009] EWCA Civ 472, [2009] 2 FLR 668; Re SB (Children) [2010] UKSC 17, [2010] 1 FLR 1161; North Yorkshire County Council v SA [2003] EWCA Civ 839, [2003] 2 FLR 849.
[26] See Jo Delahunty QC and Kate Purkiss, "The Vitamin D and rickets case" {2012} Fam Law 659.